&EPA
United States
Environmental Protection
Agency
Environmental Research
Laboratory
Narragansett Rl 02882
EPA-600 3-79-029
March 1979
Research and Development
Dredge Spoils and
Sewage Sludge in the
Trace Metal
Budget of
Estuarine and
Coastal Waters
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are
1 Environmental Health Effects Research
2. Environmental Protection Technology
3 Ecological Research
4 Environmental Monitoring
5 Socioeconomic Environmental Studies
6 Scientific and Technical Assessment Reports (STAR)
7 Interagency Energy-Environment Research and Development
8 "Special" Reports
9 Miscellaneous Reports
This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials. Problems are assessed for their long- and short-term influ-
ences. Investigations include formation, transport, and pathway studies to deter-
mine the fate of pollutants and their effects. This work provides the technical basis
for setting standards to minimize undesirable changes in living organisms in the
aquatic, terrestrial, and atmospheric environments.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/3-79-029
March 1979
DREDGE SPOILS AND SEWAGE SLUDGE IN THE TRACE METAL
BUDGET OF ESTUARINE AND COASTAL WATERS
by
H. James Simpson
Lamont-Doherty Geological Observatory
of Columbia University
Palisades, New York 10964
Grant No. R803113
Project Officer
Robert R. Payne
Environmental Research Laboratory
Narragansett, Rhode Island 02882
ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
NARRAGANSETT, RHODE ISLAND 02882
-------
DISCLAIMER
This report has been reviewed by the Environmental Research Laboratory,
Narragansett, U. S. Environmental Protection Agency, and approved for
publication. Approval does not signify that the contents necessarily reflect
the views and policies of the U. S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
ii
-------
FOREWORD
In the natural environment, few regions are as complex or as difficult to
understand as estuarine and coastal waters. In contrast to the open ocean
remote from upwelling areas or a fresh water system where individual chemical
cycles may be studied in a relatively constant matrix, coastal waters and
especially estuaries show temporal and spatial variability in almost every
measurable parameter. The extreme physical, chemical and biological changes
encountered in going from fresh water to sea water, and the transport dynamics
associated with river flow and tidal cycles almost demand a relatively broad
approach to the study of such systems if we are to have any hope of
understanding how to manage them.
In recent years, increased need for fresh water and interest in better
management of our environmental resources have helped focus attention on the
serious degradation of estuarine and coastal regions largely as a result of
urbanization. The impact of trace metals, domestic and industrial organic
residues, nutrients, and dredge spoils from estuaries on coastal waters has
become a major concern in environmental planning. In the case of the estuary
of the Hudson River, a major portion of the suspended particles are deposited
within the system only to be moved to the continental shelf by dredging. Thus
the strong coupling between an estuary and the adjacent coastal waters can be
perturbed by man's activities on a scale comparable to natural processes, even
in large systems. The following report examines some of these perturbations
and their relationships to natural cycles and processes. The impact of dredge
spoils and sewage sludge in the natural cycles of trace metals in estuarine
and coastal waters are examined for the Hudson River estuary and adjacent
region of the continental shelf through the use of a broad range of geochemical
approaches, based primarily on data collected on field samples.
Eric D. Schneider
Director, Environmental Research
Laboratory, Narragansett
111
-------
PREFACE
The research documented here was the result of the combined efforts of
a number of people, including several graduate students who were responsible
for a major portion of the field and laboratory work, and for the
interpretation and reporting of these studies. These current, or former,
students include Richard Bopp (Columbia University), Peter Bower (Queens
College and Columbia University), Bruce Deck (Columbia University),
Douglas Hammond (Columbia University), Gary Klinkhammer (University of
Rhode Island), and Curtis Olsen (Columbia University). Dr. Susan Williams
(now at Lederle Laboratories, Pearl River, New York) and Dr. Yuan-Hui Li
(Columbia University) were Co-Principal Investigators and Professor
Michael Bender (University of Rhode Island) was responsible, along with
Gary Klinkhammer, for water column measurements of trace metals in the
Hudson. Dr. Edward Catanzaro (Lamont-Doherty Geological Observatory)
and Professor William Corpe (Barnard College) also participated in this
study. The results of portions of our research have been reported in the
following abstracts and publications:
ABSTRACTS
Bopp, R.F., H.J. Simpson and C.R. Olsen, PCB's and Cs-137 in Sediments of
the Hudson Estuary (Abs.), Trans. Amer. Geophys. Union, 58, 407, 1977.
Hammond, D.E., H.J. Simpson and G. Mathieu, Distribution of Radon-222 in the
Delaware and Hudson Estuaries as an Indicator of Migration Rates of
Dissolved Species Across the Sediment-Water Interface (Abs.), Trans.
Amer. Geophys. Union, 57, 151, 1976.
Klinkhammer, G.P., M. Bender and H.J. Simpson, The Partitioning of Some
Trace Metals in the Hudson River Estuary (Abs.), Trans. Amer. Geophys.
Union, 57, 255, 1976.
Olsen, C.R., H.J. Simpson and R.M. Trier, Anthropogenic Radionuclides as
Tracers for Recent Sediment Deposition in the Hudson Estuary (Abs.),
Trans. Amer. Geophys. Union, 58, 406, 1977.
Olsen, C.R., H.J. Simpson and S.C. Williams, Sedimentation Rates in the
Hudson River Estuary (Abs.), Amer. Assoc. of Petroleum Geol. Bull.,
60, 703, 1976.
Simpson, H.J., T.H. Peng, C.R. Olsen and S.C. Williams, Radiocarbon Dating
of Estuarine Carbonates (Abs.), Amer. Assoc. of Petroleum Geol. Bull.,
60, 723, 1976.
-------
PUBLICATIONS
Hammond, D.E., H.J. Simpson and G. Mathieu, Methane and Radon-222 as
Tracers for Mechanisms of Exchange Across the Sediment-Water Interface
in the Hudson River Estuary, In: Marine Chemistry in the Coastal
Environment, ACS Symp. Ser., Vol. 18, edited by T. Church, 119-132,
American Chemical Society, Washington, D.C., 1975.
Hammond, D.E., Dissolved Gases and Kinetic Processes in the Hudson River
Estuary, Ph.D. Thesis, Columbia University, New York, New York, 1975, 161 pp
Hammond, D.E., H.J. Simpson and G. Mathieu, Radon-222 Distribution and
Transport Across the Sediment-Water Interface in the Hudson River
Estuary, Jour. Geophys. Res., 82, 3913-3920, 1977.
Olsen, C.R., H.J. Simpson, S.C. Williams, T.H. Peng and B.L. Deck, A
Geochemical Analysis of the Sediments and Sedimentary Structures in
the Hudson Estuary, Jour. Sedimentary Petrology, in press, 1978.
Simpson, H.J., D.E. Hammond, B.L. Deck and S.C. Williams, Nutrient Budgets
in the Hudson River Estuary, In: Marine Chemistry in the Coastal
Environment, ACS Symp. Ser., Vol. 18, edited by T. Church, 618-635, 1975.
Simpson, H.J., C.R. Olsen, R.F. Bopp, P.M. Bower, R.M. Trier and S.C. Williams,
Cesium-137 as a Tracer for Reactive Pollutants in Estuarine Sediments,
USA-USSR Symposium-Odessa, U.S. Environmental Protection Agency, in press,
1978.
Simpson, H.J., C.R. Olsen, S.C. Williams and R.M. Trier, Man-Made Radionuclides
and Sedimentation in the Hudson River Estuary, Science, 194, 179-183, 1976.
Simpson, H.J., S.C. Williams, C.R. Olsen and D.E. Hammond, Nutrient and
Particulate Matter Budgets in Urban Estuaries, In: Estuaries, Geophysics
and the Environment, Studies in Geophysics, 94-103, National Academy
of Sciences, Washington, D.C., 1977.
Williams, S.C., H.J. Simpson, C.R. Olsen and R.F. Bopp, Heavy Metals in
Hudson River Estuary Sediments, Marine Chemistry, in press, 1978.
MANUSCRIPTS SUBMITTED FOR PUBLICATION
Bower, P.M., H.J. Simpson, S.C. Williams and Y.H. Li, Trace Metals in the
Sediments of Foundry Cove, Cold Spring, New York.
Friedmann, M., S.C. Williams and H.J. Simpson, A New Enzymatic Method for
Analysis of Cellulose in Sediments.
Hammond, D.E. and H.J. Simpson, Methane Distribution and Sediment Flatulence
in the Hudson River Estuary.
Simpson, H.J. and D.E. Hammond, Application of One-Dimensional Models to the
Hudson River Estuary.
Simpson, H.J., T.H. Peng, C.R. Olsen and S.C. Williams, Radiocarbon
Geochemistry in the Estuary of the Hudson River.
v
-------
ABSTRACT
Many reactive pollutants, such as Zn, Cu, Pb, Cs-137, Pu-239,240
and PCB's appear to be transported and accumulated together in association
with fine-grained particles in the Hudson River estuary. Anthropogenic
increases of 3-6 times natural levels of Zn, Cu, and Pb were found for
Hudson sediments. Mobilization of Cd and Ni in the sediments of a small
embayment of the Hudson with -very high contamination levels appears to be
primarily' by resuspension of fine particles, although elevated concentrations
of Cd in pore waters were also observed. Radiocarbon measurements indicate
the predominant source of organic carbon in New York harbor sediments is
recent sewage and not petroleum hydrocarbon contamination. A new enzymatic
technique was developed to trace the distribution of cellulose, a
significant component of sewage sludge, in coastal sediments. Radon-222,
a natural radioactive gas dissolved in the Hudson, is supplied primarily
from the sediments at approximately twice the rate predicted by molecular
diffusion. Methane measurements provided additional information on the
flux of materials from sediments. The behavior of phosphate and trace
metals derived from sewage was examined on the basis of field data and the
use of simple models to examine management alternatives. The most reasonable
course appears to be completion of secondary sewage treatment plants in
New York City and major upgrading of primary treatment in New Jersey.
Tertiary treatment for nutrient removal does not appear to offer at present
the likelihood of significant improvements of receiving water quality in
the Hudson estuary. Discharge from the Hudson estuary appears to be the
dominant source of soluble metals to the adjacent coastal zone and if
soluble trace metal fluxes were the only criterion for placement of
discharge sites for dredge spoils and sewage sludge, the present site would
appear to be a reasonable one, since estuary discharge will probably
dominate soluble metal transport budgets whether or not dumping is continued
at that site.
VI
-------
CONTENTS
Foreword
Preface
Abstract
Figures
Tables
Acknowledgment
Section 1
Section 2
Section 3
Section A
Section 5
Section 6
Section 7
Introduction 1
Conclusion 2
Recommendations 4
Sources of Heavy Metals in Sediments of the 5
Hudson River Estuary
Introduction 5
Sampling and Analytical Methods 5
Results and Discussion 8
Summary and Conclusions 20
Heavy Metals in the Sediments of Foundry Cove 23
Introduction 23
Previous Work 23
Sample Collection and Analytical Procedures 25
for Sediment Metals
Analytical Data for Sediment Metals 27
Discussion of Sediment Metal Data 35
Pore Water Sampling - in situ Methods 37
Discussion of Interstitial Pore Water Results 40
Summary of Pore Water Results 60
Cesium-137 as a Tracer for Reactive 61
Pollutants in Estuary Sediments
Introduction 61
Cesium-137 as a Indicator of Recent Sediments 62
Cesium-137 and Other Anthropogenic Components 63
in Hudson Estuary Sediments
Cesium-137 as a Pollutant Tracer in Other 66
Aqueous Systems
Radiocarbon Geochemistry in the Estuary of 68
the Hudson River
Introduction 68
Hudson Estuary Morphology and Sedimentation 69
History
VII
-------
Sample Description and Preparation
Procedures for Hudson Radiocarbon
Measurements
Discussion
Conclusions
Section 8 A New Enzymatic Method for Analysis of
Cellulose in Sediments
Introduction
Methods
Results and Discussion
Section 9 Radon-222 as an Indicator of Transport
Rates from the Sediments to the Water Column
in the Hudson
Introduction
Analytical Methods
Results
Discussion
Conclusions
Section 10 Methane as an Indicator of Transport Processes
Between the Sediments and Water Column in the
Hudson
Introduction
Measurement Techniques
Results
Discussion
Conclusions
Section 11 Nutrients and Transport Models in the Hudson
Nutrients in Urban Estuaries
Sewage and Phosphate in the Hudson Estuary
Section 12 Water Column Trace Metals in the Hudson
Introduction
Sample Collection and Analytical Methods
Results
Discussion
Conclusions
Section 13 Summary of Hudson Field Research Results
References
Technical Report Data Sheet
72
77
84
87
87
88
95
100
100
101
104
109
122
129
129
130
131
134
146
147
147
149
170
170
170
171
171
187
189
194
207
vlii
-------
FIGURES
Number Page
1 Core Locations for Heavy Metal Analysis in the Hudson 6
Estuary
2 Core Locations for Heavy Metal Analysis Taken on the 7
Continental Shelf in the Hudson Shelf Canyon
3 Zinc vs. Copper Concentrations for Different Hudson Sediments 14
4 Zinc vs. Lead Concentrations for Different Hudson Sediments 14
5 Zinc vs. Copper (Expanded Scale) Concentrations Comparing 19
Hudson and Shelf Sediments
6 Location Map of Foundry Cove in Hudson River Drainage Basin 24
7 Core Location in Foundry Cove 26
8a Contours of Foundry Cove Surface Cadmium Values 28
8b Contours of Foundry Cove Nickel Values 28
9 Cd and Ni Concentration vs. Depth in Foundry Cove Core #15 29
10 Zn, Pb and Cu Concentration vs. Depth in Foundry Cove 30
Core #15
11 Cs Concentration vs. Depth in Foundry Cove Core #15
36
12a Sediment Pore Water Conductivity Measured in situ at Foundry 43
Cove Site CII
12b Pore Water Conductivity Measured in situ at Foundry Cove 44
Site CII
13 Temperature vs. Depth in Foundry Cove 45
14 pH Values vs. Depth for Foundry Cove "Peeper" Pore Water 46
Determinations
15 Chloride Concentrations for Foundry Cove "Peeper" Pore Water 48
Determinations
-------
Number Page
16 Reactive Silicate Concentrations vs. Depth for Foundry Cove 49
"Peeper" Pore Water Determinations
17 Reactive Phosphate Concentrations vs. Depth for Foundry Cove 50
"Peeper" Pore Water Determinations
18 Soluble Manganese Concentrations vs. Depth for Foundry Cove 52
"Peeper" Pore Water Determinations
19 Reactive Iron Concentrations vs. Depth for Foundry Cove 53
"Peepre" Pore Water Determinations
20 Pore Water Soluble Cadmium Concentration at Foundry Cove 54
Peeper Sites CI, CII and B
21 Co and Cs Concentration in the Sediment at Foundry Cove 55
Site B
22 Reactive Phosphate vs. Depth in Lower (mp 18) Hudson 57
Sediment Pore Waters
23a Reactive Silicate vs. Depth in Lower (mp 18) Hudson Sediment 58
Pore Waters
23b Chloride and Reactive Iron vs. Depth in Lower (mp 18) Hudson 59
Sediment Pore Waters
24 Location of Important Discharge Points of PCB, Cd and Ni, and 64
Cs in the Hudson Drainage Basin
25 ' Pu vs. Cs Activities in Hudson Sediments 65
26 PCB Concentration vs. Cs Activities for Locations in the 65
Hudson
27 Zn, Cu and Pb Concentrations vs. Cs Activities in 67
Hudson Sediments
28 Cd and Ni Concentrations vs. Cs Activity in Hudson Sediment 67
at Foundry Cove
14
29 Core Location for which C Data is Given 70
14 13
30 Apparent Initial Age &C from <5C in the Hudson 82
31 Subsurface Shell Dates - mp 18.5 in the Hudson 83
32 Subsurface Shell Dates - mp 22 in the Hudson 83
33 Sedimentation Rates in the Hudson 86
-------
Number Page
34 Absorbance at 420 nm vs. Glucose Concentration Added 92
35 Absorbance at 420 nm vs. Cellulose Reaction Time in Hudson 92
Sediments
36 Absorbance at 420 nm vs. Cellulose Added to Hudson 94
Sediments
37 Absorbance at 420 nm vs. Cellulose added to Hudson 94
Sediments (Low Range)
38 Location of the Sampling Points for Cellulose Analysis 97
39 Schematic of a Model Radon Profile in the Ocean 102
40 A Hudson Estuary Radon Profile (September 27, 1971) 102
40a The Hudson Estuary Showing Important Regions 103
41 Radon and Salinity vs. Depth (July 1972) 106
42 Radon Profiles in the Tappan Zee Region of the Hudson ]_Q7
43 Time Series of Radon Over August 1974 108
44 Radon Histogram for Samples Collected in the Hudson Estuary HQ
45 Radon vs. Depth Models in Hudson Sediments 119
46 The Biogeochemistry of Methane in Estuaries 128
47 Methane and Salinity vs. Depth Along the Hudson Estuary 132
48 Hudson Estuary Methane Distribution (August 16-18, 1973 and 133
March 2-3, 1974)
48a Methane vs. Salinity in the Lower Hudson 137
48b Surface Film Thickness vs. Wind Speed 136
48c Mass Transfer Coefficients for Various Gases vs. Diffusion 133
Coefficients (Relative to Radon)
49 Location Map of the Major Sewage Outfalls into the Lower 143
Hudson Estuary
50 Reactive Phosphate Concentrations vs. Salinity in the Lower 150
Hudson Estuary
51 Reactive Phosphate Concentrations vs. Mile Point in the 152
Lower Hudson Estuary
xi
-------
Number Page
52 Reactive Phosphate Concentrations vs. Salinity in the 152
Lower Hudson Estuary (Composite of Figure 50)
53 Schematic of Sewage Loading in the Hudson Estuary 153
54 Schematic of the Phosphate Budget in the Inner Harbor 154
55 Hudson Estuary Phosphate Fluxes (Box Model) 156
56 Major Estuaries in the Northeast U.S. With Average Flows 157
57 Predicted Phosphate Concentration from Model Calculations 162
58 Hudson River Fresh Water Flow (at Green Island Station) 163
59 Soluble Zn, Co, Mn, Ni, and Reactive Phosphate Concentration 172
vs. Salinity in the Hudson (April 1974 and October 1975)
60 Suspended Matter Concentration vs. Salinity and Mile Point 173
in the Hudson
61 Weight of Al vs. total Particulate Weight in Hudson Water 173
Samples
62 Metal to Al Weight Ratios of Particulates in Hudson Samples 182
Plotted vs. Salinity
63 Total Mn vs. Salinity for Hudson Water Samples 185
64 Mn Abundances in Surface and Subsurface Particulate Samples 186
vs. Mile Point
65 Partitioning of Five Trace Metals in the Hudson Estuary 188
XII
-------
TABLES
Number Page
1 Hudson Estuary Sediment Composition 9
2 New York Harbor Sediment Composition 10
3 New York Bight Sediment Composition ]_]_
4 Comparison of Metal Concentration by Several Analytical 14
Techniques
5 Trace Metal End Members in Hudson Sediments 16
6 Normalization Procedures for Comparing Heavy Metals in 18
Sandy Sediments with Fine-Grained Sediment
7 Natural Abundances of Heavy Metals 21
8 Foundry Cove Trace Metals - Core #15 31
9 Foundry Cove Trace Metals - Core #'s 6 and 10 32
10 Foundry Cove Trace Metals - Surface Sediments 33
11 Foundry Cove Radionuclide Data 34
12 Foundry Cove Peeper Designations 41
13 C Measurements of Inorganic Carbon in Water Samples from 73
the Hudson River Drainage Basin
14 C-L4 Measurements of Organic Carbon in Surface Sediments from 75
the Hudson Estuary
15 c^ Dates of Carbonate Shells from Hudson Sediments 76
16 Cellulose Content of Hudson River Estuary and New York Bight 96
Sediments
17 Activity of Mobile Radon in Sediment 105
18 Estuary Geometry and Median Radon Concentrations 111
19 Radon (dpm/1) in Streams (mp 0-91) with Drainage Area > 100 km2 112
xiii
-------
Number _
20 Radon Budget for Hudson Water Column 113
21 Comparison of Diffusive, Observed and Turbated Fluxes 114
22 Summary of Rn Measurements 123
23 Methane Distribution in Hudson Estuary Sediments 135
24 Equations for Dissolution of a Rising Bubble 141
25 Environmental Conditions and Methane Averages 142
26 Comparison of Flatulant Production Rate and Burial Rate of 145
Organic Carbon
27 One-Dimensional Model Parameters for Tidal Hudson 158
28 Segmented Model Observed Salinities and Phosphate Concentrations 161
29 Comparison of Phosphate Behavior in the Lower Hudson Estuary 164
and Lake Erie
30 Nutrient Data 166
31 Concentrations of Dissolved Trace Metals in the Hudson River 174
Estuary
32 Mile Point Locations, Depths, Salinities and the Particulate 177
Concentrations of Six Metals for Samples Collected from the
Hudson Estuary
33 Mile Point Locations, Depths, Salinities and the Metal Abundances 179
of Six Metals in Particulate Matter Samples Collected from the
Hudson Estuary
34 Comparison Between Ranges of "Soluble" Metal Concentrations Found 181
in the Hudson Estuary, Narragansett Bay, Rhode Island, and the
Sargasso Sea
35 Reported Metal Concentrations of Sewage Being Discharged into 184
the Lower Hudson Estuary
36 Soluble Metal Fluxes to the New York Bight Apex 192
xiv
-------
ACKNOWLEDGMENTS
Organization, preparation, and editing of this report were primarily
accomplished by the efforts of Karen Antlitz and Bruce Deck, in addition to
their normal administrative and research activities at Lamont-Doherty
Geological Observatory.
xv
-------
SECTION 1
INTRODUCTION
Sediments dredged from urban and industrial harbor areas and sludges from
sewage treatment plants are continuously generated in large quantities by
our coastal cities. Disposal of these materials presents a number of
difficult management problems. In the New York City metropolitan region,
current practice involves discharge of both dredge spoils and sewage sludge to
coastal waters within a few tens of kilometers of shore. One concern is that
toxic pollutants such as heavy metals will be released from the waste solids
to solution in the water column in amounts sufficient to substantially affect
the biota of the impacted coastal waters. Another consideration is that many
pollutants are associated with fine particles which could be easily trans-
ported by currents away from the disposal sites possibly to accumulate in
other areas.
Much of the research on mobilization of toxic contaminants has employed
laboratory simulations, the results of which are then extrapolated to field
situations. We have concentrated on analysis of field samples to help provide
a framework into which laboratory experiments of other investigations can be
integrated. The field area of our research, the estuary of the Hudson River,
is one of the most important components of the estuarine-coastal water system
of the New York City metropolitan region. This region has one of the largest
dredge spoil and sewage sludge marine disposal operations in the world, and
is currently the focus of several major field research programs primarily in
the New York Bight. Our research in the Hudson is complementary to that In
the Bight, and will help provide critical information for development of the
best management approach for disposal of dredge spoils and sewage sludge in
the New York areas, and other areas as well.
One of our primary objectives has been to estimate the rate of releases
of dissolved heavy metals from polluted sediments and to examine the
importance of these releases relative to other heavy metal fluxes and cycles
In estuarine and coastal waters. We have examined the distribution of
several metals in the sediments, pore waters and water column of the Hudson
estuary. To interpret these data in terms of processes and rates of trans-
port, we have also measured a number of additional parameters in both the
sediments and water column of the Hudson. New analytical techniques and
sampling approaches were developed and used to obtain some of the data, and
these innovations should prove useful in studies by other investigators. We
have attempted to suggest tentative conclusions directly related to management
objectives whenever these seemed warranted, even when more extensive
research is obviously needed to reach really firm conclusions.
-------
SECTION 2
CONCLUSIONS
1. Trace metal (Zn, Cu and Pb) concentrations in recent Hudson sediments
are 3-6 times background levels for fine-grained sediments, and are comparable
to observed metal concentrations of sediments in the New York Bight sewage
sludge and dredge spoil dumping area.
2. Foundry Cove, a small embayment in the Hudson 60 km north of New York
City, appears to be the site of most intense metal contamination in the Hudson,
with Cd and Ni concentrations of up to a few per cent by weight. Mobilization
of Cd in this area is apparently primarily via resuspension of fine particles,
although the fluxes of soluble Cd from these sediments appears measurable with
additional field studies. This site offers an unusual opportunity for further
understanding of Cd and Ni mobilization from contaminated sediments in both
fresh water and saline environments.
3. Many reactive pollutants in Hudson sediments appear to be transported
and accumulated together in association with fine-grained particles. This
covariation of Zn, Cu, Pb, Cd, Ni, PCB's, Pu-239,240 and Cs-137 greatly
simplifies the task of mapping the distribution of contaminated sediments in
the Hudson.
4. Through radiocarbon measurements we have found the organic carbon of
New York harbor sediments to be predominantly recent sewage, with fossil fuel
carbon making up a considerably smaller proportion of the pollutant carbon than
in contaminated sediments from other areas in the Hudson. The hydrocarbon
pollutant levels in the harbor sediments are substantial, however.
5. We have developed a new analytical technique for cellulose, utilizing
cellulase enzymes to hydrolyze cellulose to glucose. This technique is quite
sensitive and specific and has been exploited here for studies of the source
of high molecular weight organic matter in shelf sediments.
6. The flux of a soluble, chemically-inert natural tracer, Rn-222, from
Hudson sediments to the water column is approximately two times the rate
predicted assuming only molecular diffusion rates in the sediment interstitial
waters. Thus the transport of some types of materials from sediments to the
water column can be estimated reasonably well from simple diffusion models;
the integrated effect of a number of processes which might enhance the flux of
materials to the water column is roughly comparable to that of molecular
diffusion.
-------
7. Dissolved methane in the Hudson appears to be supplied, at least in
part, by partial solution of bubbles produced in the sediments and released to
the water column. The variation of dissolved methane concentration in Hudson
sediments is very large, and may be a valuable indicator of the mobility of
some trace metals in sediments.
8. Phosphate is supplied to the Hudson in such large amounts by sewage
discharge, and is transported out of the harbor so rapidly by estuarine
circulation, that it approximates a conservative tracer of sewage in the lower
Hudson. This provides a valuable indicator of discharge rates of soluble
trace metals to the New York Bight by estuarine circulation. Tertiary treat-
ment of New York City area sewage for phosphate removal does not appear
justified in light of our present information. Completion of secondary
treatment plants and major upgrading of primary treatment plants, especially
in New Jersey appears to be a more sensible course to pursue.
9. A major fraction of the trace metals discharged to New York harbor
in sewage leave the Hudson through the Narrows in solution. To the first
approximation the soluble metal levels in the lower Hudson can be described
in terms of three end member mixing between fresh Hudson water, New York Bight
sea water and sewage.
10. The discharge of soluble metals from the Hudson estuary appears to be
the dominant source of soluble metals to the apex of the New York Bight.
Release of soluble metals from dumping area sediments does not seem to be a
significant source compared to estuarine discharge. If the only criteria for
the placement of a site for discharge of dredge spoils and sewage sludges was
consideration of soluble trace metal fluxes, the present site would appear to
be a reasonable one, since soluble metal transport budgets will probably be
dominated by estuarine discharge whether or not dumping is continued at that
site.
-------
SECTION 3
RECOMMENDATIONS
1. We suggest that studies of the impact of specific activities such as
dredge spoil or sewage sludge disposal in coastal waters require integrated
field research efforts which exploit a wider range of geochemical tracers and
other approaches than are commonly employed. Research of wide scope is
necessary' to develop the understanding necessary to place the impact of
specific activities in perspective.
2. Although more efforts should be devoted to the understanding of trace
metal transport pathways in polluted estuarine and coastal waters, it is at
least equally important to intensify studies of the environmental pathways and
significance of pollutants such a PCB's and other toxic compounds, especially
for the New York City area.
-------
SECTION 4
SOURCES OF HEAVY METALS IN SEDIMENTS OF THE HUDSON RIVER ESTUARY
INTRODUCTION
Heavy metals discharged into estuaries and coastal waters with domestic
and industrial wastes are often present as particulates, or have strong
affinities for solid phases in the receiving waters (Morel et_ al_. , 1975).
Present distributions of metals in sediments can serve as an indication of
the time history and extent of pollutant discharge in specific areas (Chow
ej^ _al. , 1973; Bruland et_ al_. , 1974; Erlenkeuser et_ al. , 1974; de Groot et al. ,
1976). To interpret the present metal content of sediments, several factors
must be considered including the "natural" preindustrial metal content of
sediments in the area in question. Components such as clay minerals, organics
and iron and manganese oxides are important phases in both the natural and
pollutant heavy metal content of sediments. Other phases such as quartz and
calcium carbonate, which tend to occur in relatively large particles, are
usually less important in binding heavy metals, and thus serve primarily to
"dilute" the heavy metal bearing phases which dominate the fine particle size
fraction. In environments where particle size distributions and mineralogy are
heterogeneous, substantial variability in heavy metal concentrations would
thus be expected, even in the absence of significant local sources of metal
contamination (de Groot et_ a^. , 1976). The pattern of sediment accumulation in
environments such as estuaries or coastal waters is generally sufficiently
complicated to make it difficult to establish the time history of accumulation
at any particular sampling site.
We have analyzed sediments from the Hudson estuary and adjacent coastal
area for zinc (Zn), copper (Cu), lead (Pb), manganese (Mn) and iron (Fe), as
well as man-made radionuclides such as cesium-137s to establish the present
distribution of heavy metals in a system subject to substantial pollutant
loading. We have related the present sediment metal concentrations to
those typical of the Hudson prior to industrial pollutant discharge, and
to the patterns of recent sediment accumulation.
SAMPLING AND ANALYTICAL METHODS
Cores were taken from near the upstream limit of saline water intrusion
in the Hudson to the middle of New York harbor (Figure 1) where salinities
are usually about two thirds of deep ocean water salinities. Large grab
samples of the upper 10 centimeters of sediment in the coastal waters
adjacent to the Hudson were collected out to beyond the edge of the
continental shelf (Figure 2). Most of the cores were collected in plastic
liners with a six centimeter diameter gravity corer, and were usually about
-------
-4I°00
-40°30'
74°00
73°30
Figure 1. Location map for cores collected in the Hudson River Estuary.
Sites where background levels of heavy metals were observed at
some depth in the core are indicated with triangles. Locations
in the Hudson are generally described in terms of the number of
miles (mile point = mp) upstream of the southern tip of Manhattan
Island.
-------
4I°N
HUDSON CANYON
STATIONS
75°W
Figure 2. Location map for cores collected on the continental shelf. Sites are indicated with
negative numbers, indicating the number of miles downstream from the origin of the mile
point reference system. The sampling sites generally follow the narrow band of fine-grained
sediments in the Hudson Shelf Valley and Hudson Submarine Canyon.
-------
0.5 meters in length. One of the cores from nineteen miles upstream of the
southern tip of Manhattan Island (mile point 19) was a 6 meter piston core
taken from one of Columbia University's oceanographic vessels (R/V VEMA).
Another of the cores (mp 24) which had a total length of about 12 meters was
taken in sections by hand drilling in a marsh which currently is under water
during only very high tides.
Sediment samples were air dried, ground in a mortar and pestle and passed
through a 250 y nylon sieve. Samples for heavy metal determinations were
dried at 105°C to constant weight and stored in a dessicator, while those
for measurement by gamma spectrometry were sealed in teflon-lined aluminum
cans and counted with a large volume lithium-drifted germanium detector
and multichannel analyzer.
Metals were leached from sediments by fuming samples with concentrated
HC1/HNO (1:3) for several hours, followed by treatment with 70% HC1C>4 until
the solid phase was essentially pure white. Final solutions of acid were made
up to approximately 5% in HCl after driving off most of the HC10,. Sediment
suspensions were filtered and washed through a Gooch crucible with preweighed
glass fiber filters. Filtrates were diluted and analyzed by flame atomic
absorption spectrometry and the white sediment residue was reweighed after
drying. The difference in sediment weight resulting from treatment with
strong acids is reported as "loss on leaching" (LOL). The difference in
weight for fresh sediment samples dried at 105°C to constant weight and
then heated to 500°C for a number of hours is reported as "loss on ignition"
(LIG). Quartz was determined by monitoring the a-g transition in a differen-
tial scanning calorimeter, potassium from the gamma ray emission peak for K
at 1.46 Mev and ^-^Cs from the gamma ray emission peak at 0.662 Mev.
RESULTS AND DISCUSSION
Heavy metal data and related measurements are reported in Tables 1, 2 and
3 for 56 samples of sediment and one of sewage sludge from a large New York
City sewage treatment plant. The data are listed beginning at the upstream end
of the sampling locations in the Hudson, with the data in Table 2 from New York
harbor and in Table 3 from the sediments of the continental shelf. Samples froi
the Hudson estuary were generally fine-grained silts relatively uniform in
physical properties, while several of those from the shelf were predominantly
sand despite our attempts to sample the narrow zone of fine-grained sediment
between the mouth of the Hudson Estuary and the Hudson Submarine Canyon.
We chose the analytical scheme outlined above for Hudson sediments after
trying a number of procedures. The reported data for Zn, Cu, Pb and Mn are
probably good estimates of the total metal concentration in our samples, but
the reported values for Fe may be systematically low (15-20%). We have
analyzed a number of samples after total dissolution with HF and found the
procedure to be somewhat less convenient and reproducible than the strong acid
leaching method employed here. Data reported in Table 4 provide some comparison
of metal concentrations in a Hudson sample which has been analyzed by three
separate procedures. The sample (mp 43) used for the intercomparison of
analytical techniques is typical of recent Hudson Estuary sediments upstream
of New York harbor. We have also analysed seven replicates of a sample typical
-------
Table 1
Hudson Estuary Sediment Composition
Location
(Mile Point)
56
54
53
44
43
25
24
19
a) Cs
Depth
(cm)
0-5
5-10
0-5
5-10
10-15
15-20
20-25
25-30
30-35
35-40
40-45
45-50
50-55
0-5
5-10
10-15
15-20
20-25
0-7
7-14
14-21
0-10
0-6
6-12
12-16
16-21
0-3
633-635
1175-1177
0-10
22-29
83-85
150-153
240-241
325-328
430-438
545-548
Cs-137a
(pCi/g)
0.9
0.0
2.5
1.8
0.2
0.0
0.0
0.0
2.7
2.1
0.7
0.0
0.0
0.6
0.0
2.7
1.2
0.8
0.3
0.1
( )
( )
-137 was measured by
Zn
(Ug/g)
190
100
315
320
305
225
125
97
90
80
84
80
82
290
295
245
170
87
125
96
65
315
190
205
220
170
230
82
105
48
81
83
82
68
77
77
81
gamma
were analyzed and found to be
Cu
58
26
88
86
80
65
31
24
18
17
18
17
17
110
97
83
55
18
29
27
26
107
99
115
105
89
115
23
26
35
16
20
20
14
20
21
15
Pb
(Mg/g)
66
22
150
130
120
110
49
33
25
20
20
23
23
175
130
105
74
13
45
45
37
140
92
92
98
72
175
22
24
55
39
24
28
28
24
27
21
spectrometry (photo
free of
Mn
(mg/g)
1.70
2.40
0.75
0.75
0.73
0.66
0.57
0.57
0.55
0.51
0.51
0.52
0.52
1.15
1.05
.92
.77
.61
0.53
0.65
0.44
1.39
0.45
0.51
0.52
0.48
2.45
0.52
0.48
0.32
1.15
0.84
1.00
0.96
0.84
0.90
0.83
peak at
Fe
CO
3.3
3.6
3.7
3.8
3.7
3.5
3.4
3.5
3.5
2.8
3.4
3.3
3.5
2.4
2.9
3.0
2.7
2.2
2.3
2.3
1.8
3.3
2.6
2.5
2.8
2.6
3.1
3.2
4.0
0.7
3.2
3.6
3.8
2.9
3.5
3.4
3.1
0.662
Cs-137; those indicated
LIGb
6.6
7.0
9.6
9.4
9.6
8.3
7.1
7.2
6.8
6.9
6.9
7.3
7.5
8.4
7.7
7.4
6.3
6.0
3. '5
4.6
3.2
4.6
4.9
5.6
4.5
38.5
6.8
7.5
2.6
7.1
6.6
6.0
4.7
4.7
4.9
4.5
MEV) .
LOLC
19
20
25
25
24
23
24
21
21
20
21
20
21
20
13
18
16
14
14
15
10
14
16
16
14
50
20
24
9
23
24
24
18
23
23
19
Samples
K<
2.2
2.4
2.2
2.5
2.4
2.2
2.4
2.2
2.2
2.2
2.3
2.3
2.0
1.9
2.1
2.1
2.2
2.3
2.0
2.2
1.7
2.4
2.0
2.4
2.2
2,4
2.3
2.2
Fine
Quartz Fraction
(%) (% < 63u)
25
24
25
24 97
29 98
30 98
31 96
31 96
29 96
30 96
32 97
95
57
26
24
27
35
27
29
37 90
reported as 0,0
with dashed lines (
} were assumed
to be free of Cs-137.
(b) Loss on Ignition (LIG) indicates weight loss upon heating from 105°C to 500°C.
(c) Loss on Leaching (LOL) indicates weight loss of a sample dried at 105°c after treatment with
strong acids (KC1, tniCy HC104).
40
(d) Potassium (K) was measured by gamna counting of the K peak at 1.46 KEV.
(e) Quartz was measured by monitoring the a-8 transition with a differential scanning ccvlorimeter.
-------
TABLE 2
o
Location
(Mile Point)
2
0
-2
Depth
(cm)
0-5
20-25
50-55
0-5
18-25
45-53
60-65
0-5
12-20
25-30
35-40
45-50
55-60
Cs-137a
(pCi/g)
1.2
0.0
0.0
0.7
1.5
0.7
0.0
0.4
0.6
0.4
0.6
0.6
0.9
LI »_- W -L >— ' J-
Zn
(yg/g)
345
235
245
260
215
225
53
337
434
459
399
472
557
IX lid J_ LIU i.
Cu
(yg/g)
225
145
180
180
200
285
12
248
344
348
294
338
416
iJCU-LlUell
Pb
(yg/g)
830
165
245
140
200
175
<28
202
271
253
247
286
345
L UUIUJJUS
Mn
(mg/g)
0.31
0.29
0.26
0.26
0.30
0.33
0.56
0.55
0.50
0.60
0.47
0.53
0.46
-LLJLUll
Fe
(%)
3.7
3.3
3.2
3.3
4.0
3.3
3.6
3.3
3.3
4.2
3.6
3.5
3.7
LIGb
(%)
9.2
6.5
7.8
8.5
9.8
8.8
4.9
7.5
9.4
9.4
10.0
11.0
LOL°
(%)
26
22
23
24
28
24
23
22
25
27
23
d e
K Quartz
(%) (%)
2.1
2.1
1.9
2.1
2.3
2.0
2.0
1.8
2.1
2.2
1.9
1.9
1.9
Fine
Fraction
(% <63y)
a, 95
% 95
^ 95
^ 95
a) Cs-137 was measured by gamma spectrometry (photo peak at 0.662 MEV). Samples reported as 0.0 were
analyzed and found to be free of Cs-137; those indicated with dashed lines ( - ) were assumed to be
free of Cs-137.
b) Loss on Ignition (LIG) indicates weight loss upon heating from 105°C to 500°C.
c) Loss on Leaching (LOL) indicates weight loss of a sample dried at 105°C after treatment with strong
acids (HC1, HNO HC10 ).
,40
d) Potassium (K) was measured by gamma counting of the K " peak at 1.46 MEV.
e) Quartz was measured by monitoring the a-$ transition with a differential scanning calorimeter.
-------
TABLE 3
New York Bight Sediment Composition
Location Depth Cs-137a Zn Cu Pb Mn Fe LIG LOLC
(Mile Point) (cm) (pCi/g) (ug/g) (yg/g) (yg/g) (mg/g) (%) (%) (%)
Fine
d e
K Quartz Fraction
Sewage
Sludge
-38
-67
-117
-136
-147
-157
0-10
0-10
0-10
0-10
0-10
0-10
0.1
0.2
0.1
0.0
0.0
0.0
0.1
692
110
38
17
79
90
69
1440
27
3
2
19
25
20
375
67
< 20
< 20
< 20
< 20
30
0.20
0.18
0.10
0.28
0.28
0.30
0.32
1.1
2.1
1.0
0.7
2.6
2.6
2.6
72.4
5.0
2.5
1.3
9.3
10.2
10.8
84
14
.7
6
28
30
33
0.7 1
1.5
1.5
1.3
2.3
2.3
2.1
33
17
5
87
98
93
a) Cs-137 was measured by gamma spectrometry (photo peak at 0.662 MEV). Samples reported as 0.0 were
analyzed and found to be free of Cs-137; those indicated with dashed lines (-) were assumed to be
free of Cs-137.
b) Loss on Ignition (LIG) indicates weight loss upon heating from 105°C to 500°C.
c) Loss on Leaching (LOL) indicates weight loss of a sample dried at 105°C after treatment with
strong acids (HC1, HNO , HC1O 1
40
d) Potassium (K) was measured by gamma counting of the K peak at 1.46 MEV.
e) Quartz was measured by monitoring the a-3 transition with a differential scanning calorimeter.
-------
TABLE 4
Comparison of Metal Concentration by Several Analytical Techniques
Zn
(yg/g)
336
327
302
294
315
(12%)
(570)
255
305
Cu
(yg/g)
ill
103
105
107
107
(9%)
79
108
102
Pb
(yg/g)
141
145
144
131
140
(14%)
144
114
_
Mg
(mg/g)
1.42
1.30
1.25
1.57
1.39
(18%)
1.27
1.09
1.39
Fe
(%)
3.5
3.3
2,8
3.5
3.3
(25%)
4.0
3.5
4.3
mp 43a
mp 43a
mp 43a
mp 43a
Average mp 43
2a
mp 43b
mp 43
mp 43°
a) Each of these samples was analyzed by the technique used for all of the
data reported in Tables 1, 2 and 3. They were analyzed on different days
with groups of other samples.
b) These samples were analyzed by a total dissolution technique using HF.
Erratic Zn values were common in a number of other samples we analyzed
by total dissolution.
c) Reported analytical data for a blind determination of sample composition
of a commercial instrument company using emission spectroscopy.
12
-------
of preindustrial Hudson sediments (mp 19, 525 cm) which has substantially
lower concentrations of Zn, Cu and Pb than the mp 43 sample. Statistical
variation (2a) in that data set was: Zn 8%, Cu 13%, Pb 39%, Mn 3% and Fe 6%.
The uncertainties of Pb and Cu were higher than for the sample at mp 43
because we were approaching the detection limits by our procedures for those
two metals (Pb: average 23 ppm, detection limit ^ 10 ppm; Cu: average 19 ppm,
detection limit ^ 4 ppm).
During the past several years we have developed some understanding of
the general^ccumulation patterns of recent sediments in the Hudson through
the use of Cs (Simpson eit al. , 1976; Olsen et_ al. , 1977). This tracer has
been added to the Hudson estuary from two sources: (1) global fallout from
atmospheric nuclear weapons testing over the last two decades with peak
deliveries from rainfall during the years 1963-1965; (2) low level releases
from a nuclear power reactor at Indian Point (mp 43) over the last decade, with
peak discharges in the years 1971-1972. Enough Cs is associated with
particles in the Hudson to provide a readily measureable tracer of sediment
accumulation rates in areas of high deposition, and of transport along the axis
of the Hudson downstream of the,reactor site, when used in conjunction with two
other radionuclides, Cs and Co which have been supplied to the Hudson
almost exclusively by reactor releases. Cesium-137, which has a half life
of about 30 years, is an unequivocal indicator of sediment which has been
in contact with the water column within the last two decades, and primarily
serves here as a "label" for recent (last two decades) sediments. Using this
tracer we have found that accumulation rates of fine-grained sediments in
the Hudson range over approximately two orders of magnitude, with the
dominant deposition zone being New York harbor. We have found cores with
Cs to nearly 3 meters below the sediment surface, and accumulation rates
of 10-20 cm per year typical of large areas of the harbor. Marginal coves in
the lower salinity reaches of the estuary (mp 45 - mp 56) have sediment
accumulation rates of a few cm per year (Wrenn e^t al., 1971; Simpson et al. ,
1976) while most of the total area of the estuary has little net accumulation
of recent fine-grained sediments labeled with Cs (less than a cm per year).
137
From Tables 1 and 2 it is clear that sediments containing Cs are
significantly higher in the metals Zn, Cu and Pb. Two of the reported cores
(mp 54 and mp 19) penetrate well into sediments with relatively low and constant
concentrations of Zn, Cu and Pb which appear to be typical of pre-industrial
sediments in the Hudson. One of the cores (mp 19) contains many subsurface
layers of estuarine carbonate shells, four of which we have analyzed for C
and found apparent ages of 1000-2000 radiocarbon years. Interpretation of
radiocarbon data from estuarine carbonates is not unambiguous, but we are
confident that the lower half of this core (3-6 meters), which lies below the
four radiocarbon dated carbonate layers in the upper 3 meters, is representative
of preindustrial fine-grained sediments in the Hudson.
When data from Tables 1 and 2 are plotted in terms of the concentrations
of Zn and Cu, (Figure 3) the samples fall into three relatively distinct groups.
Approximately one third of the Hudson samples, from seven of the eleven coring
sites contain similar Zn and Cu concentrations below some level in each
of the cores. This can be seen most clearly in three cores (mp 56, mp 24 and
mp 19) two of which were the longest collected. Similar "baseline" values
were found throughout the salinity range in the estuary, and none of the
13
-------
600-
500^
400-
e
Q.
Q.
200-
100
Figure 3.
recent harbor
sediments
recent Hudson
sediments
600-
500-
400-
Q.
Q.
300-
200-
00-
old Hudson sediments
recent harbor
sediments
recent Hudson
sediments
old Hudson sediments
00
400
100
200
300
400
200 300
Cu(ppm)->- Pb(ppm)—>-
Plot of zinc and copper concentrations in Hudson sediments, with individual samples indicated
by one of four symbols. Dots indicate old sediments, free of any cesium-137 or elevated metal
levels. Triangles indicate recent harbor sediments (mp 2 to mp -2) containing 137-Cs, and x's
indicate recent estuary sediments upstream of the harbor containing 137-Cs. Shelf samples are
indicated with S's.
Figure 4. Plot of zinc and lead concentrations in Hudson sediments.
are the same as for Figure 3.
The format and conventions
-------
estuary samples in the group with Zn concentrations of ^ 80 ppm and Cu
concentrations of ^ 20 ppm contain Cs. We consider these samples to
be typical of preindustrial fine-grained sediments throughout the Hudson.
-j O ~1
The remainder of the samples (all of which contain Cs and thus are
recent sediments) can be represented as members of two groups. One group
has a number of samples with Zn concentrations of ^ 300 ppm and Cu
concentrations of ^ 100 ppm, as well as others which fall along a mixing
line between these values and preindustrial sediment Zn and Cu values.
All of these samples were collected upstream of New York harbor (Table 1) and
appear to be representative of the level of heavy metals in recent Hudson
sediments upstream of New York City resulting from many diffuse sources of
Zn and Cu. One of the upstream sites for which we are reporting data (mp 54)
is a small cove which is grossly contaminated with Cd and Ni effluent from
a battery factory (Kneip et al., 1974). Bower (1976) has found Zn, Cu
and Pb concentrations in this cove to be unrelated to the local metal
contamination of Cd and Ni (up to ^ 0.1% by weight of both metals) and to be
typical of other recent estuarine sediments in the Hudson in terms of these
three metals. The third group of samples all from New York harbor (Table 2),
fall off of the trend line for recent sediments upstream of the harbor (Figure
3) and extend up to considerably higher concentrations 'of both zinc (^ 550 ppm)
and copper (^ 400 ppm). Metal contamination added directly to New York harbor
sediments appears to be richer in Cu relative to Zn than the diffuse recent
contamination of sediments transported down the Hudson toward the harbor. The
sewage sludge sample we analyzed had very high Cu (^ 1400 ppm) relative to Zn
(^ 700 ppm) , supporting the suggestion of relatively high Cu proportions in the
metal discharge to the harbor (Table 3).
A similar distribution of data from Tables 1 and 2 is produced if Zn is
plotted against Pb (Figure 4) rather than Cu, although the relative
proportions of contaminant Pb and Zn appear more similar in both recent
harbor sediments and those upstream of the harbor than was suggested for
Cu and Zn. Approximate concentrations for Zn, Cu and Pb in the three
"end member" types of fine grained Hudson sediments indicated on Figures 3
and 4 are listed in Table 5. The recent harbor sediments do not have metal
contents which tend to cluster around a small concentration range and thus
an end member composition for recent harbor sediments is more artificial than
for preindustrial or for recent sediments upstream of the harbor, but a
representative high concentration sample is included for comparison purposes.
Heavy metal data have been used to delineate the extent of spreading of
wastes such as sewage sludge and dredge spoils dumped in the coastal area
about 15 km outside of the mouth of the Hudson Estuary (Gross et al., 1971;
Gross, 1972; Carmody et _al., 1973). The sensitivity of such tracers is
relatively high because the dumped wastes are rich in heavy metals, and because
most of the coastal sediments off New York City are sandy and thus have low
natural levels of heavy metals. Three of the coastal sediment samples reported
here (Table 3) from the vicinity of the shelf break (mp -136, mp -147 and mp
-157) have Zn and Cu concentrations similar to preindustrial Hudson sediments
(Figure 3). These samples are very fine-grained and are from an area far
enough off shore to be reasonably free of much recent metal contamination. The
metal data for the other three shelf samples in Table 3 (mp -38, mp -67 and
mp -117) are more difficult to interpret. All three samples are relatively
15
-------
TABLE 5
Trace Metal End Members in Hudson Sediments
Zinc Copper Lead
(ppm) (ppm) (ppm)
1. Old Hudson sediments 80 20 25
2. Recent Hudson sediments 300 100 135
3. Recent Harbor sediments 550 400 350
4. (Recent - Old) Hudson sediments 220 80 110
t#2 - #1]
5. Recent (Harbor - Hudson) sediments 250 300 215
[#3 - #2]
16
-------
low in Zn, Cu and Pb but all three have relatively high proportions of sand
(> 63 u) which would tend to dilute the metal concentrations. Conceptually, it
is logical to "correct" metal concentration data in samples with a high
percentage of sand to provide a more representative comparison with fine-grained
sediments and thus to perhaps obtain an indication of the source of the metals
in a particular sample of sandy sediment. The most satisfactory procedure for
developing such a correction is not clear, however. A number of possibilities
are reasonable to suggest: (1) assume the metals in sediments are primarily
bound in organic phases, and multiply the observed concentration of metals in
sandy sediments by the ratio of organic content in fine-grained sediments to
that of organic matter in the sandy sediments; (2) assume that particles below
a certain size (such as 63 y) are the only ones important in binding heavy
metals (de Groot et_ a.1. , 1976), and multiply metal data from sandy sediments
by the ratio of the proportion of sample weight less than 63 y particles
in typical fine-grained sediments to that in the sandy sediments; (3) assume
quartz or other metal poor phase is not significant in binding metals and
calculate metal concentrations on a "quartz free" basis (Thomas, 1972; Thomas,
1973). There are many other possible approaches, such as normalizing to a
constant surface area, but no single procedure appears conceptually superior.
We have tried several normalization procedures for metal data on the
three sandy shelf samples mentioned above (Table 6). All of the procedures
increase the observed metal concentrations but the amount of increase
varied from one approach to another. The "corrected" values based on weight
loss on ignition (LIG), weight loss on acid leaching (LOL) and the Fe content
of sediments were all reasonably similar, and the "corrected" values are
plotted in Figure 5, which is an expanded version of the preindustrial sediment
portion of Figure 3. Two of the sandy shelf samples (mp -67 and mp -117) appear
to be relatively free of recent metal contamination, whereas the shelf sample
(mp -38) closest to the estuary mouth and to the area of dredge spoil and
sewage sludge discharge (^ mp -25) appears to have a significant proportion of
recent contamination. The "corrections" suggested are obviously crude and
simple minded, but are relatively convenient to apply from observational data
on sediment properties, and suggest reasonable conclusions on the basis of the
few shelf samples discussed here.
A previous study of the distribution of cation exchange capacity,
and organic material in Hudson sediments (McCrone, 1967) indicated that the
estuarine silts of the Hudson had relatively uniform properties (when compared
with variations found on the adjacent continental shelf) over an extended
reach upstream of ^ mp 22. We have not tried to apply any normalization
scheme to metal data from the estuary sediments reported in Tables 1 and 2.
Some variations in grain size proportions and organic content do occur within
the estuary samples, as indicated by the reported values for LIG and LOL.
These variations probably contribute to some of the spread in Zn, Cu and Pb
concentrations observed in Figures 3 and 4.
Concentrations of Fe are relatively constant in Hudson sediments, but
Mn has considerable variation. Except for the surface sample at mp 24,
which is really more of a marsh soil than an estuarine sediment sample, the
range of Mn to Fe ratios is 0.008 to 0.067. At locations with significant
variation down the core, the highest values of Mn/Fe are usually near the
surface. This observation is consistent with upward diffusion of reduced Mn
17
-------
TABLE 6
Normalization Procedures for Comparing Heavy Metals in Sandy Sediments
with Fine-Grained Sediment
Sediment Parameter
Weight fraction < 63 V-
Normalization Factor
Weight loss on ignition (LIG)
Normalization Factor
Weight loss on acid leaching (LOL)
Normalization Factor
Weight fraction iron
Normalization Factor
Fine-Grained
Hudson and/or
Shelf Sediment
Shelf Sediments
mp -38 mp -67 mp -117
^100%
1
8%
1
25%
1
3.5%
1
33%
3.0
5%
1.6
14%
1.8
2.1%
1.7
17%
5.5
2.5%
3.2
7%
3.6
1.0%
3.5
5%
20
1.3%
6.2
6%
4.2
0.7%
5.0
18
-------
175-
150
125-
£
Q.
Q.
100-
75-
50-
25-
S (-38)
S (-67)
:-38)
S
(-117)
«»• 8 •
*^e
D (-67)
D (-117]
25
50
Cu (ppm)
75
100
125
Figure -5. Expanded plot of a portion of Figure 3 to illustrate the effect of
transforming observed metal concentrations in sandy shelf (mp -38,
mp -67 and mp -117) sediments by the use of normalizing factors
for comparison with fine-grained sediments. Normalized shelf
sediment metal concentrations are indicated as S's in the figure.
19
-------
followed by precipitation of oxidized Mn near the sediment-water interface.
Although we have no direct evidence that this process actually takes place in
Hudson River sediments, the mobility of Mh(II) in reducing sediments is well-
known (Lynn and Bonatti, 1965; Li et al., 1969; Bender, 1971; Elderfield, 1976).
Manganese is quite reactive in the transition zone between river water and
sea water, and a number of studies of Mn in estuaries have been made (Lowman
et. al. , 1966; Windom _e_t al. , 1971; Evans and Cutshall, 1973; Graham et al. ,
T97lTf Evans j|t_ al. , 1977). A recent study in Narragansett Bay (Graham et al. ,
1976) which Summarized the behavior of Mn in estuarine environments, suggested
that Mn is desorbed from particles as salinity increases, but slowly oxidized
to Mn(IV). The time constant suggested for Mn oxidation in Narragansett Bay at
salinities of '^ two-thirds of sea water was on the order of two days, whereas
the desorption time constant was much shorter. In the Hudson, lower values
of Mn/Fe especially in the upper sediment layers are more common in the harbor
and shelf sediments, compared with sites in the Hudson where fresh or low
salinity waters are found (mp 40-60). This suggests that desorption of Mn from
suspended particles and accumulation of these particles in New York harbor is
one of the major processes affecting the behavior of Mn in this estuary.
Precipitation of Mn within the Hudson Estuary does not appear to be nearly as
important as in Narragansett Bay. If Mn precipitation is primarily the result
of slow oxidation of Mn(II), the short residence time of dissolved components
in the Hudson during high river discharge (^ 5 days), and heavy sewage loading
from the New York City area (^ 100 m /sec) may explain the lack of evidence for
this process within the Hudson Estuary. A similar conclusion has been reached
by Klinkhammer (1977) on the basis of an extensive survey including both
dissolved and suspended Mn phases in the Hudson.
SUMMARY AND CONCLUSIONS
Concentrations of Zn, Cu and Pb in Hudson sediments indicate those metals
can be considered in terms of three general sources of comparable magnitude:
(1) preindustrial natural sources of weathering; (2) diffuse recent
contamination throughout the estuary; and (3) sewage, industrial effluent
and urban runoff, reaching the harbor from the New York City area.
The preindustrial levels of Zn, Cu and Pb in Hudson sediments are
comparable to those reported for average shale (Turekian and Wedepohl,
1961) and average continental crust (Taylor, 1964) except for relatively
low Cu in the Hudson (Table 7). Our data for Hudson sediments is very
similar to background values suggested for Ottawa River sediments (Table 7,
Oliver, 1973). A previous study of sediments from an engineering boring at
about mp 60, an area which is now usually fresh water (Owens et_ al., 1974),
reported heavy metal data on seven "estuarine" silt samples which averaged as
follows: Zn -^ 80 yg/g, Mn -^ 1.6 mg/g, Fe '^ 4.2%. Except for Fe, where
our values are ^ 20% lower, these data are quite consistent with our values
from a large number of cores throughout the range of salinity.
137
The distribution of Cs in Hudson sediments corresponds well with that
of sediments with recent metal contamination, and provides a good indication
of the pattern of pollutant metal accumulation. The depth to which pollutant
metals are found in Hudson sediments is usually very similar to the depth
distribution of Cs.
20
-------
TABLE 7
Natural Abundances of Heavy Metals
Zn Cu Pb
Hudson Sediments (this study) 80 20 25
(preindustrial)
Average Shale 95 45 20
(Turekian and Wedepohl, 1961)
Average Continental Crust 70 55 12
(Taylor, 1964)
Ottawa River Sediments - Background 84 28 26
(Oliver, 1973)
21
-------
The metal concentration of shelf sediments consisting of appreciable
fractions of sand-sized particles can be compared with fine-grained estuarine
sediments by normalizing the sandy sediment metal data on the basis of empirical
measurements of sediment properties. Similar conclusions about the amount of
recent metal contamination in sandy shelf sediments were reached using several
normalization parameters including the fraction of weight loss upon heating
from 105°C to 500°C, the fraction of weight loss due to leaching with strong
oxidizing acids, and the Fe concentration in the sediment.
The amount of Fe in fine-grained Hudson estuary sediments is reasonably
constant, but the concentration of Mn varies over approximately an order of
magnitude, with greater values near the sediment surface in low salinity and
fresh water areas and lesser values deep in cores and in the higher salinity
zones such as New York harbor.
The harbor sediments are relatively enriched in Cu, and a first order
budget for Cu released to the harbor can be estimated. Assuming a mean concen-
tration of 250 ppm Cu in harbor sediments, and 100 ppm Cu for recent fine-
grained sediments delivered to the harbor from upstream, the net increase is
150 ppm Cu. Approximately 4 x 10 tons per year (dry weight) of dredge spoils
are discharged in the coastal water off New York City (Gross, 1972). We do not
have any direct information on the average metal composition of the dredge
spoils which would be significantly influenced by the proportion of the total
which was sandy sediment removed from areas such as lower New York Bay.
Most of the dredge spoils do appear to be derived from New York harbor in
areas of fine particle deposition (Panuzio, 1965). If we assume half
(this estimate is probably a reasonable minimum value) of the total mass of
dredge spoils consists of recent fine-grained harbor sediment, then ^ 300 tons
of Cu per year added to the harbor sediments are being removed from the harbor
by dredging. From the data of Klein et_ al. (1974) a total delivery of ^ 2
tons/day of Cu to New York harbor can be estimated. Thus based on an assumption
that at least half of the dredge spoils consist of recent harbor sediments, the
current removal rate of Cu by dredging appears to be at least half of the
loading rate of Cu. For Zn, since ^ 5 tons/day are added (Klein jit_ _aJL. , 1975)
the fraction (^ 15-20%) going into the harbor sediments appears considerably
smaller than for Cu. A similar suggestion about the relative behavior of Zn
and Cu added to New York harbor has been made by Klinkhammer (1977) on the
basis of dissolved phase distributions of Zn and Cu in the Hudson.
22
-------
SECTION 5
HEAVY METALS IN THE SEDIMENTS AND PORE WATERS OF FOUNDRY COVE
INTRODUCTION
Sediments of rivers and estuaries in urban areas frequently have
increased heavy metal concentrations due to the discharge of industrial wastes
and domestic sewage. Usually metal contamination from many sources is super-
imposed, producing an increase of a number of "trace" elements over "back-
ground" levels which varies in magnitude with the strength of the pollution
source and total area of dispersion. In most cases metals such as zinc,
copper and lead - derived from a number of distributed sources - are found in
sediments of urban waterways at levels of a few times pre-industrial concen-
trations, up to perhaps an order of magnitude higher. In some situations,
metals derived from single sources form a "halo" of very high concentrations,
decreasing away from discrete discharge sites. Foundry Cove, a small embayment
of the Hudson River (about 60 km north of New York City) is a site where very
high concentrations (up to 5% by weight) of cadmium (Cd) and nickel (Ni) in
sediments are found as the result of waste discharge from a battery factory
(Figure 6).
Foundry Cove has received alkaline discharges of approximately equal
amounts of Cd and Ni (Cd/Ni 'v 1.8 by weight) since the construction in the
early 1950's of a battery fabrication facility operated for the U.S. Department
of Defense in the town of Cold Spring, New York. The cove is located in a
relatively pristine reach of the Hudson, with substantial areas of park land
nearby, and is remote from other industrial metal discharges of any
significance. Foundry Cove has a total surface area of approximately 0.5 km
extending eastward from the main channel of the Hudson River. The cove is
shallow (1-3 meters at low tide) and divided into two segments by a railroad
causeway with a channel of about 25 meters width connecting the two areas.
The Hudson River is tidal for more than 200 km upstream of New York City, with
semi-diurnal stage changes of 1-2 meters, producing relatively vigorous tidal
current flushing of the outer part of Foundry Cove, and also the inner cove
through the connecting channel. Foundry Cove and the adjacent main channel of
the Hudson contain fresh water during much of the year. Saline water - up to
3-6°/oo - is present most years during low fresh-water flow months, especially
during years of lower than average river discharge.
PREVIOUS WORK
Beginning about 1971, several studies were made of the distribution of
Cd and Ni in Foundry Cove (Kneip et_ al_. , 1974; Buehler and Hirshfield, 1974).
Much of this research was related to a court-ordered dredging operation which
23
-------
HUDSON ESTUARY
Figure 6. Location map of Foundry Cove in the Hudson River Drainage Basin,
24
-------
involved the most heavily contaminated zone of sediments (up to 5.0 x 10 ppm
for both Cd and Ni ) in the eastern end of the inner cove. Most of the
published data is from the highly contaminated portion of Foundry Cove on the
landward side of the railway. Surveys of sediment Cd and Ni in the inner
cove indicate significant differences in the distribution pattern before and
after dredging, with the primary change being a shift in location and a
broadening of the zone of highest concentration (Kneip, 1975). From reported
data we estimate that the total area of the zone of Cd concentrations greater
than 10,000 ppm increased from about 5% of the inner cove to about 10%. The
quantity of Cd in the inner cove prior to dredging has been estimated to be
about 25 tons (Kneip, 1974). The discharge rate of Cd from the inner cove
has been estimated to be 0.2-2 tons/year based on water column data over a
tidal cycle in the channel connecting the inner cove with the outer cove
(Kneip, 1975).
SAMPLE COLLECTION AND ANALYTICAL PROCEDURES FOR SEDIMENT METALS
We collected gravity cores at 15 sites in Foundry Cove, 13 of which
were in the outer cove (Figure 7). The cores, which were about 6 cm in
diameter and up to 50 cm in length, were sectioned into 5 cm intervals at the
field site, transported to the lab and frozen. Following thawing, the
samples were dried overnight at 100°C and pulverized with a mortar and pestle.
Aliquots of about 2 grams of the total sample weight of 80 grams were leached
with 20 ml of cone. HNO_ and 5 ml of cone. HC1 at room temperature for 30
minutes. The samples were heated slowly until most of the organic matter was
destroyed, and then reduced in volume on a hot plate to ^- 10 ml. After
cooling, 20 ml of 70% HC10, was added and the sample slowly evaporated to
near-dryness. After cooling, 10 ml of cone. HC1 and ^ 40 ml of distilled
water were added to the white residue of sediment. Then the sediment was
filtered, the residue rinsed, and the final filtrate volume brought to 100 ml
in a volumetric flask.
Analysis of the resultant acid solutions was performed by flame atomic
absorption spectrometry (Perkin-Elmer 303), using mixed-metal standards for
Cd, Ni, Zn, Cu, Pb, Co, Mn and Fe. The residual sediment, which consisted
of a fine white powder, was dried and weighed and the fraction of weight loss
of the original sediment sample is reported as "loss on leaching" (LOL)
by strong acids. Separate dry sediment samples were heated overnight to
500°C and the weight loss, which is related to the organic content of the
sediments, is reported as "loss on ignition'7 (LIG) .
Small aliquots (^' 40 mg) of dry sediment were analyzed for quartz
content by monitoring the a-g transition in a differential scanning calori-
meter a 1 atmosphere and 573°C. After the frozen sediment samples were
thawed, dried, and ground to a fine powder, and before subsampling and
treatment with strong acids for metal analysis, they were placed in 100 cc
aluminum cans and counted on a lithium-drifted germanium (Ge[Li]) gamma ray
detector. The spectrum of gamma ray emission from each sample allows the
measurement of several naturally occurring rad^onuclides, including K and a
number of radioactive daughters of U and Th. Anthropogenic radj^nuclides
from global fallout ( Cs) and from a nuclear power reactor ( Cs, Cs
and Co) located about 15 km downstream of Foundry Cove were also measured.
25
-------
Foundry Cove Sampling Stations
NJ
ON
Sewer
Line
Inner Cove
Outer Cove
0 0.5
Kilometers
Hudson River
meter
depth
Figure 7. Sampling locations for cores from Foundry Cove with "peeper" sites marked.
-------
Several of the samples have also been analyzed for plutonium isotopes by
alpha spectrometry ( ' pu).
ANALYTICAL DATA FOR SEDIMENT METALS
Concentrations of metals in Foundry Cove sediments are reported in
Tables 8, 9 and 10, while the radionuclide data are reported in Table 11.
Cd contents in samples analyzed here ranged between 'x. 1 ppm, which is the
same order of magnitude as average shale (0.3 ppm, Turekian and Wedepohl,
1961) and ^ 900 ppm. Ni contents ranged between 'x/ 30 ppm, which is somewhat
less than average shale (68 ppm), and 'x. 300 ppm. In general, Cd and Ni had
comparable concentrations with the concentrations of Cd (Figure 8a) and
Ni (Figure 8b) in surface sediments (0-5 cm) of the outer cove decreasing
away from the connecting channel to the inner cove. Approximate contours of
the more highly contaminated inner cove surface sediments are included in
Figures,8 and 9 based on data reported elsewhere (Kneip, 1975).
In the outer cove, the sediments contaminated with significant quantities
of anthropogenic Cd and Ni are confined to the upper 15 cm in the three
cores analyzed up to now (Tables 8, 9; Figure 9). In all three cores
concentrations of Cd and Ni as a function of depth in the sediment are
similar. In the case of Cd, the surface sediments of the outer cove are more
than two orders of magnitude above "background" levels while Ni is about one
order of magnitude greater. In the inner cove surface sediments Cd and Ni
concentrations up to two orders of magnitude higher than our data in the
outer cove have been reported (Kneip, 1975).
The concentrations of Zn, Cu and Pb in Foundry Cove surface sediments
(Tables 8, 9, 10) are approximately 3-6 times higher than background levels
reached at the bottom of the cores. The depth profiles of these three metals
are similar and show elevated levels to depths approximately twice that for
Cd and Ni (Tables 8, 9; Figure 10). There is no horizontal gradient in
surface sediment concentrations of Zn, Cu and Pb, indicating that the
discharges from the battery factory have not significantly increased the
levels of these metals in Foundry Cove sediments.
Concentrations of Fe and Mn are also reported in Tables 8, 9 and 10.
These metals do not appear to have been significantly increased by localized
pollutant discharges. Manganese in the upper 20 cm of the core samples is
slightly higher than deeper in the cores. The cause of this increase may be
diffusion of Mn from lower in the core, as has been observed in deep sea
sediments, or it could be due to other processes such as a generalized
recent increase in particulate Mn concentrations, as for Zn, Cu and Pb.
The surface sediments in Foundry Cove contain activities of Cs
comparable to continental soils (Hardy, 1974). There are measurable amounts
Of cs and Co derived from the reactor located about 15 km downstream,
k £ j-vp dominant source of Cs in Foundry Cove sediments is fallout.
239,240pu concentrations in the sediments are also comparable to fallout
levels typical of soils.
In general, the distribution of Cs in the outer cove sediments is
similar to that of Cd and Ni. Several cores have Cd and Cs almost down
27
-------
0 0.5
Kilometers
Cd
Figure 8a. Contours of cadmium in surface sediments of Foundry Cove. Data
for 103 and 104 contours from Kneip, 1974; Kneip, 1975; and
Kneip et_ al. , 1974.
Figure 8b. Contours of nickel in surface sediments of Foundry Cove. Data
for 103 and 104 contours from Kneip, 1974; Kneip, 1975; and
Kneip et al., 1974.
28
-------
0-5
Cd-Ni #15
200
ppm
300
Figure 9. Depth distribution of cadmium and nickel in core #15.
29
-------
Zn-Pb-Cu #15
I
4
D
40
120
ED Zn
O Pb
A Cu
I i I i
160 200 240 280
ppm—>
320
Figure 10. Depth distribution of zinc, lead and copper in core #15,
30
-------
Depth
(cm)
0-5
5-10
10-15
15-20
20-25
25-30
30-35
35-40
40-45
45-50
50-55
Cd
(ppm)
258
246
147
7
14
4
2
2
2
1
2
Ni
(ppm)
325
284
120
44
49
36
31
30
33
31
32
TABLE 8
Foundry Cove Trace Metals -
Zn Cu Pb Co
(ppm) (ppm) (ppm) (ppm)
316
321
305
224
124
97
90
80
84
80
82
88
86
80
65
31
24
18
17
18
17
17
147
132
121
109
49
33
25
20
20
23
23
23
24
20
21
20
20
16
16
16
15
15
Core #15
Mn
(ppm)
752
751
729
656
571
567
546
513
510
518
524
Fe
(%)
3.7
3.8
3.7
3.5
3.4
3.5
3.5
2.8
3.4
3.3
3.5
Quartz
(%)
25
24
25
24
29
30
31
31
29
30
32
LIGa
(%)
9.6
9.4
9.6
8.8
7.1
7.2
6.8
6.9
6.9
7.3
7.5
LOL
(%)
25
25
24
23
24
21
21
20
21
20
21
a) LIG - Weight loss on ignition from 105°C to 500°C.
b) LOL - Weight loss on leaching with strong oxidizing acids.
-------
TABLE 9
Foundry Cove Trace Metals - Core tt's 6 and 10
Depth
(cm)
Core No .
0-5
5-10
10-15
15-20
20-25
25-30
30-35
Core No .
0-5
5-10
10-15
15-20
20-25
25-30
30-35
35-40
Cd
(ppm)
6
214
261
79
4
4
2
2
10
249
69
7
4
2
-
2
2
Ni
(ppm)
152
184
115
44
42
31
30
356
119
59
56
42
-
40
40
Zn
(ppm)
358
342
314
235
219
119
79
309
307
274
238
168
-
106
107
Cu
(ppm)
104
100
87
67
59
28
15
85
85
82
70
45
-
26
26
Pb
(ppm)
179
174
124
109
97
60
26
129
133
125
111
87
-
43
31
Co
(ppm)
25
28
22
19
19
15
14
23
20
20
19
18
"
17
17
Mn
(ppm)
920
1090
790
590
630
440
420
740
710
720
610
530
-
550
560
Fe
3.9
4.2
4.1
3.6
3.5
2.9
3.2
4.0
3.8
3.7
3.5
3.5
-
3.5
3.4
LIGa
9.0
9.1
8.8
8.7
7.9
5.4
4.5
8.3
8.3
8.2
8.1
6.8
-
5.6
-
LOL
28
28
27
25
25
19
18
24
25
24
24
23
-
26
23
a) LIG - Weight loss on ignition from 105°C to 500°C.
b) LOL - Weight loss on leaching with strong oxidizing acids.
-------
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Cd
(ppm)
417
4
170
89
118
214
115
214
301
249
51
153
5
908
258
Ni
(ppm)
211
31
179
122
139
152
108
200
429
356
105
137
28
325
325
TABLE 10
Foundry Cove Trace Metals - Surface Sediments
Zn Cu Pb Co Mn
(ppm) (ppm) (ppm) (ppm) (ppm)
320
76
323
369
388
358
247
323
324
309
328
379
82
362
316
97
16
86
99
84
104
95
88
83
85
83
100
14
94
88
169
24
134
165
119
179
132
125
130
129
118
173
25
158
147
29
14
22
21
22
25
22
20
23
23
22
25
13
33
23
710
440
750
770
710
920
780
940
830
740
920
1040
320
690
750
Fe LIGS
(%) (%)
9.3
5.0
11.5
8.6
8.5
3.8
8.5
7.9
7.1
4.0
14.9
8.9
5.0
7.8
3.7 9.6
LOLb
(%)
30
19
29
25
28
28
28
26
25
24
26
27
16
33
25
a) LIG - Weight loss on ignition from 105°C to 500°C.
b) LOL - Weight loss on leaching with strong oxidizing acids.
-------
TABLE 11
Foundry Cove Radionuclide Data
Core
No.
1
2
3
5
6
7
8
9
10
10
11
12
13
14
15
15
15
15
15
15
Depth
(cm)
0-5
0-5
0-5
0-5
0-5
0-5
0-5
0-5
0-5
0-5
0-5
0-5
0-5
0-5
0-5
5-10
10-15
15-20
20-25
50-55
137
Cs
(pCi/kg)
2135
71
1840
2220
2785
670
1610
1725
2250
2320
1485
2135
42
1835
2475
1825
210
26
35
9
+ 96
+ 44
+ 115
+ 135
+_ 135
+_ 83
+ 105
+ 66
+ 81
+ 55
+ 54
+ 135
+ 35
+ 69
+ 63
+_ 68
+ 17
+ 17
+ 23
+ 14
134^
Cs
(pCi/kg)
8
1
17
2
118
83
37
10
14
34
43
15
21
98
17
- 8
-15
0
7
+_ 24
+ 30
+ 32
+ 32
± 43
+ 39
+ 32
+ 17
± 10
-
+ 14
+ 36
+ 23
± 21
+ 26
+ 32
+ 10
+ 11
+ 12
+ 9
60
Co
(pCi/kg)
103
8
69
32
257
66
11
34
42
66
148
72
96
69
19
15
30
7
13
+ 43
+ 49
+ 64
+ 60
+ 84
+ 70
+ 55
+ 37
± 17
-
+ 29
± 72
± 41
+ 47
+ 27
+ 29
+ 17
+ 25
± 27
+_ 20
239,240
Pu
(PCi/kg)
15
15
15
0-5
5-10
10-15
76
57
6
+ 7
+ 5
+ 1
34
-------
to background levels at the surface, apparently due either to dredging or
tidal current scouring of the recent sediments. Cd and Cs also have
similar distributions with depth in the sediments (Figures 9 and 11) .
DISCUSSION OF SEDIMENT METAL DATA
The data reported here indicate that significant quantities of Cd and Ni
from the battery factory discharge are present in the outer portion of
Foundry Cove. Estimates of the sediment burden in the outer cove (Bower, 1976)
based on the analytical data reported here are about 2-3 tons of Cd and
about the same amount for Ni. These quantities are 5-10% of the total burden
of Cd and Ni in Foundry Cove sediments (25-50 tons) estimated (Bower, 1976)
from data on the inner cove sediments reported elsewhere (Kneip, 1975).
The distribution of Cd and Ni in Foundry Cove sediments are quite
similar,,indicating comparable transport pathways for these two metals. The
most reasonable explanation is that particulate phase transport has been
dominant in establishing the present distribution of both metals in the
sediments of the outer cove.
The depth distribution of Cd and Cs in the outer cove sediments are
very similar. There are no major differences in grain size, mineralogy, or
organic content in the Foundry Cove sediments sampled here as indicated by
the data for quartz, weight loss on ignition, and weight loss by leaching with
strong acids. Particulates less than 62 microns accounted for between 85%
and 99% of the total mass of each sample., with the median value about 96%.
The similarity in depth distribution of Cs and Cd is somewhat surprising
since the sources of these metals are obviously quite different. However,
both are recent (last few decades), and could be expected to bind strongly
to fine-grained, organic-rich sediments.
There is some indication of slight differential movement of Cd and Ni
in Foundry Cove, if data from the outer cove from this study are considered
together with those for the inner cove from other sources (Kneip et al.,
1974; Kneip, 1975). Cd appears to be slightly less mobile than Ni, as
indicated by a possible decrease in Cd/Ni ratios away from the discharge
site. This trend is only barely suggested by the data, and is clearly of
minor importance in the overall transport of Cd and Ni in Foundry Cove.
The levels of Cd and Ni in Foundry Cove sediments appear to be definitely
a potential local health hazard, especially if indigenous organisms such as
crabs are consumed in any significant quantities by sport fisherman. As has
been pointed out elsewhere (Kneip ejt a^. , 1974, Kneip, 1975) the court-
required dredging in 1972-1973 did not eliminate Cd and Ni contaminated
sediments from the cove, and could possibly have enhanced the spread of these
metals within the system by mobilizing heavily-contaminated fine-grained
particj.es during dredging. We do not suggest, however, that renewed dredging
would be the best policy for Foundry Cove. Until a better solution is found,
the most prudent course to follow would appear to be to discourage human
consumption of organisms such as crabs from the cove.
35
-------
l37Cs for Core Station #15
5-
E
o
10-
Q.
CD
Q
I
5-
20-
50-
l
1000
2000
pCi/kg—>
i
3000
Figure 11. Depth distribution of " Cs in Foundry Cove, core #15.
-------
PORE WATER SAMPLING - IN SITU METHODS
We have devoted considerable effort in the last year to the development
of a relatively new tool for sampling sediment interstitial water (pore
waters). This tool, called a pore water "peeper" is based on equilibrating
water-filled chambers with pore waters through a permeable membrane which
prevents the passage of sediment particles. The peeper was conceived by
Ray Hesslein, a former graduate student at Lamont-Doherty, and tested at
LDGO and at the Experimental Lakes Area, in Ontario, Canada. The peepers
used in our studies were designed with 4.5 cm chambers (5.5 cm x .9 cm x 1 cm)
milled at 1 cm in^grvals. For all but one of the peeper profiles run in
1976, Spectropore 2 dialysis membrane (cellulose base) was used. This
material has an effective molecular weight cutoff of 12000 to 14000 MWU and
a wall thickness of 25 microns. This is similar to the type prescribed by
Hesslein (1976) for which he determined a half time for equilibration of
a peeper cell with a well-mixed outer reservoir of about 120 minutes. On the
basis of measurements made in lake sediments (porosity of > 90%) and through
numerical modeling he predicted 92 to 96% equilibration after 7 days. In
the sediments of the lower Hudson River (mile point 18), the dialysis
membrane was partly destroyed in one week, apparently by bacterial activity,
while farther upstream (rnp 54), minimal damage occurred to the membrane over
a two week period. An attempt was made with,one of the peeper samplers to use
Nuclepore polycarbonate membrane of 0.2 micron pore size as a substitute for
the dialysis membrane. The half time for 0.2 micron Nuclepore with a well-mixed
outer reservoir is similar (80 min.) to that of the dialysis material. In
other experiments in which the inner cell as well as the outer reservoir
were continuously mixed, equilibration half times were only 20 to 30 minutes.
for either material, indicating that diffusion through the membrane is a
minor factor in the overall rate of equilibration. Lab experiments measuring
the flux of chloride and metal radioisotopes were made by Dennis Adler, a
graduate student at Lamont-Doherty, who attempted to fit the results to a
numerical model describing the diffusion and transport within both the
sediments and pore water sampler. These experiments indicated that after
one week 70-80% equilibration should be attained if the diffusion occurs
only perpendicular to the sampler (one dimensional diffusion from a planar
source). If these results are extrapolated to the actual diffusion geometry
of the sampler in the sediment, which is closer to radial symmetry, the
degree of equilibration is predicted to be 85-90%.
In preparing the peeper for implacement, the plexiglass shell and cover
plate were soaked in dilute nitric acid overnight. Open cell polyurethane
foam which was used as a gasket to seal the membrane tightly was washed in hot
water and dried. The peeper cells were filled with distilled water and the
dialysis membrane placed on top so as to exclude air bubbles (for one of
the peepers [C III] the reloading was done in the field and only a distilled
water rinse was possible, also no attempt was made to exclude oxygen dissolved
in the cells; as later shown, the data does not differ significantly from the
other profiles and it is reasonable to assume that oxygen introduced into the
pore water from the peeper cells is rapidly consumed by organic oxidation and
that the effect on the chemistry of the surrounding pore water is small and
short lived). The foam gasket and cover plate were placed on top and the
entire assembly held together and compressed with nylon screws. The peeper
37
-------
was next transferred to an enclosure and purged with nitrogen. This was
repeated at later times to remove air that had diffused out of the cells and
any that might leak into the container. The length of time the peeper remained
in the container varied but was greater than 2 hours and should have allowed
for more than 50% oxygen removal from the cells. The peepers were transported
to Foundry Cove in their containers and were removed just before emplacement in the
sediments. When inserting the peepers into the sediments, care was taken to
insure that at least two cells remained above the sediment-water interface.
The length of time the peepers remained in the sediment was originally chosen
as two weeks to allow the cells to approach reasonably close to equilibrium
with the pore waters, but deterioration of the membrane, presumably due to
microbiological activity, required a decrease to one week equilibration time.
The peeper was removed from the sediments at the end of the equilibration
period by pulling on a rope attached to a surface float. Most of the adhered
mud was washed off in the surface water and the peeper transferred in a
horizontal position to the boat. The upper cover plate was covered with plastic
kitchen wrap to decrease evaporation and gas exchange and the peeper was brought
to shore. (The last peeper [C VI] was wrapped in the plastic and then replaced
in the air tight enclosure and purged with nitrogen followed by a 1.5 hour trip
back to the lab where it was sampled.) After removing the cover plate and foam
material, the membrane was washed clean of particles with a light stream of
distilled water and a fine brush. Samples for each of the tests were then
transferred to test tubes or serum bottles by syringe (gas samples) or plastic
tiped micropipettes. The dissolved gases CH and CO were sampled first
followed by sulfide, iron, phosphate, ammonia, metals, silicate, specific
conductance, chloride, and pH. The volume of water in each cell was limited
to 4.5 ml so that each test could only be performed approximately every 3
cells. With the help of several persons, the entire 80 cell peeper could be
sampled in under an hour. The analytical techniques used followed closely
standard wet chemical seawater or freshwater procedures with reduction in
size of all samples and reagents so the final test volume was between 4 and
7 ml. The references and procedure modifications are given below:
Specific Conductance
Two types of determinations were made using this parameter. A special
micro cell consisting of two platinum electrodes in a .5 ml volume teflon
sample holder was made. This cell was standardized with .01 N KCl solution
and against a 1.0 constant cell in the field and determinations made on
samples of the peeper pore water. Temperature correction was based on the
measured air temperature. Relative conductivity measurements were made in the
sediments directly by a long PVC rod which had two platimum electrodes mounted
on the side near the end. The open water cell constant was determined before
and after each run by comparison to overlying water conductance measured with
a standard cell. Temperature correction was obtained by measuring the
temperature profile in the sediment at the same time with a thermistor
mounted between the electrodes. In both the determinations, a Leed and
Northrup #4959 conductance bridge at 1000 cps was used.
Temperature
In situ measurements of the sediment temperature were made using a Fenwald
38
-------
glass bead thermistor mounted between the conductivity electrodes on the
sediment probe. The resistance was measured with an AC Wheatstone
conductivity bridge and the temperature determined by substitution into a
cubic equation relating temperature and resistance over the range 5°-30°C
using a high accuracy 0.1°C thermometer.
pH
pH measurements were made in the field with a Beckman model G pH meter
using microreference and glass electrodes. Standard pH 7 buffer was used
as a reference. One ml of sample plus wash was required.
Chloride
Chloride was determined by electrochemical titration with silver using
an American Instrument Co. Chloride Titrator; .5 ml of sample and 1.5 ml
dilution water were titrated with 2 ml conditioning reagent. Standards were
prepared from sodium chloride.
Silicate
The silicate test approximates that of Strickland and Parsons (1968) for
seawater scaled down by a factor of 10. Distilled water blanks were carried
into the field and standards of sodium silicofluoride in distilled water
were used. The ammonium molybdate reagent could not be added before the
sample so it was added rapidly to the plastic test tube containing the .2 ml
sample and 2.3 mis of dilution water at the start of the sample preparation
procedure.
Phosphate
The phosphate test is identical to that of Strickland and Parsons (1968)
for seawater with a 1:10 reduction in sample and reagent volume. Acid rinsed
test tubes were used.
Iron
Iron was determined using a wet chemical method for "reactive" iron
following Stookey (1970)- Total iron was determined after reduction for 15 min.
in aqueous hydroxylamine at boiling water temperatures. Boyle and Edmond (1977)
have found that the procedure approximates total dissolved iron as determined
by direct injection graphite tube atomic absorption spectrophotometry.
Manganese and Cadmium
For both Mn and Cd determinations direct injection graphite tube atomic
absorption spectrophotometry was employed. One ml samples were taken in the
field from the peeper cells using plastic tipped Finn pipettes and placed in
acid washed (8N HC1) glass test tubes. Samples for Mn had .02 ml 5N HC1
added while those to be analyzed for Cd had .02 ml 5N HNO added. Reagents
used were chosen for low blanks. Standards were diluted from Fisher 1000 ppm
solutions.
39
-------
Methane and Carbon Dioxide
A 2 ml sample was injected into a helium-purged 10 ml serum bottle
which previously had 0.05 ml concentrated H2SO added. Gas samples of 1 ml were
withdrawn and analyzed by gas chromatography using Porapack Q and 5A columns
with thermal conductivity (for CO ) and flame ionization (for CH4) detectors.
Sulfide
The sulfide test was a modification of the Methylene Blue comparison
test for total dissolved sulfide given in Standard Methods (13th Ed., 1971)
section 228B. The sample size was reduced to 2 ml which was added to tubes
containing 1.5 ml distilled water and 0.05 ml zinc acetate solution in the
field: 0.25 ml of the ammine solution 1 was used with 0.05 ml of ferric
chloride and 0.8 ml of ammonium phosphate. The color produced was measured
with one cm cells at 660 nm against distilled water blanks and standards made
from sodium sulfide in nitrogen purged water.
Ammonia
The ammonia procedure follows that of Solozono (1969) run at a scale
reduction of 1:10. The samples were either processed immediately after
returning from the field or "fixed" by the addition of the phenol reagent.
Special cleaning and care was used to reduce the blank and avoid contamination.
All ammonia determinations were run by Dennis Adler.
DISCUSSION OF INTERSTITIAL PORE WATER RESULTS
The development of the pore water peeper as a tool provides a new
approach to obtaining samples of sediment pore waters over closely spaced
intervals. As with any new technique, however, great care must be taken to
insure that the resultant data is valid. We have tried lab tracer experiments
and model calculations to determine the rate of equilibration of the sampling
chambers with interstitial waters. We have also attempted to compare the
results of a peeper profile in Narragansett Bay to those of a core taken and
centrifuged by Mike Bender's group at the University of Rhode Island. Apparently,
lateral homogeneities over a small horizontal distance (^ 1.3 meters) of the
Narragansett Bay sediments contributed substantially to disagreement between
two peepers 1 meter apart and a box core taken between them. The best
information that we currently have on the operating characteristics of our
sampler comes from careful examination of field data from Foundry Cove and
by comparison to data from other areas collected by different methods.
In the summer of 1976 nine separate peepers were placed in Foundry Cove
at three separate sites (B, C I and C II) as shown in Figure 7. Table 12
gives the times of emplacement and removal for each peeper and information
on procedures used for each run. Site B was located about 20 m from the
mouth of the channel through which the battery wastes had entered the cove.
This site was next to the dredged area and is thought to represent originally
deposited material containing about 2000-5000 ppm Cd. Two peeper profiles
were made here about 20' apart (B I and B II). Special procedures were used
with B II to determine soluble pore water metal concentrations in the highly
40
-------
TABLE 12
Foundry Cove
(mp 54)
Site Peeper Designation
B B I
B II
C I C I
C II C II
C III
C IV(a,b)
C V
C VI
Hudson River
(mp 18) ALP A
ALP I
ALP II
Dates
052076
062476
052076
061076
062476
071576
072276
080576
08 74
062775
072876
Duration
(days)
14
14
14
14
14
7
6
(Nuclepore)
6.5
7
14
7
(Nuclepore)
41
-------
contaminated sediments. C I is the only peeper profile made at site C I.
Later profiles were made at site C II due to currents near the mouth of the
causeway and sediment type (river-derived mud). The majority of peeper
profiles were made at site C II, which represents a 200 square meter area on
the western side of the large shallow area which makes up the central portion
of the cove. Peepers C II - C VI were positioned at this site. Cd and Ni
concentrations in the surface sediments are between 100 and 400 ppm.
One important piece of information that must be determined before the pore
water data is closely examined is the degree of horizontal homogeneity in the
Foundry Cove sediments. At site C II, iri situ conductivity measurements were
made for each peeper profile using a plastic probe with platinum electrodes.
Results of the sediment conductivity probe are shown in Figure .12a. The solid
line represents an approximate fit of the data, while the dotted line marks
the upper 15 cm of peeper C V's curve showing salt intrusion from the
overlying water. The similarity of the curves from different locations is
clear. The data points below 25 cm show only at +_ 10% spread while those
above show a somewhat greater range, which is not unexpected since the
sediment above 25 cm contains local vertical irregularities related to
biological and physical processes at the surface. A closer examination of
individual peeper curves shows that the +_ 10% zone is not the result of
random noise in each profile. The major ions supporting the specific
conductance are HCO , Cl (from measurements) and Na, K and Ca (ratio
unknown) to maintain charge balance. Since the concentration of these major
ions swamps the minor ion effects the latteral distribution of the major ions
can be assumed to be fairly uniform within sediments of the same type in the
cove if the conductivities at each site are the same.
Figure 12b shows the peeper pore water conductivity data as collected.
The construction of the micro-conductivity cell and operator error caused
large calibration shifts during several of the peeper runs. Only two of the
profiles may be accurate as plotted although several others can be made to
approach these if assumptions on the nature of the error are made. The
erroneous profiles are included to show the relatively small scatter of points
within a single profile indicating that dilution of individual peeper samples
did not occur during removal or the washing of the membrane.
The temperature profiles determined by the sediment probe are shown in
Figure 13. Differences in the profiles reflect unequal heat inputs possibly
caused by differences in the amount of sediment exposure during tidal cycles.
The slopes of the profiles are nearly constant at approximately 7.5°C per
meter.
Figure 14 shows the pH values determined on the peeper pore waters. All
of the profiles show the same general shape as approximated by the solid
line; initial values at the overlying water value of 7.3 decreasing to 6.6
at 10 cm, and remaining constant at 6.6 to 6.5 cm where a slight increase
occurs. The only profile not following this trend is C VI which was not
sampled until 1.5 hours after removal. The pH shift of about 0.25 units
could be explained by about 20% loss of the E CO .
42
-------
0
D
I5H
20.
25H
30H
CM
40^
45H
50H
55H
65
100 200 300
CONDUCTIVITY XMHOS/CM
400
I
500 600 700 800
I I I I
900 1000 100
I I I
SED. PROBE
CI -CVI
70-
Figure 12 a.
In situ specific conductance in yMHO /cm of Foundry Cove
sediments with depth. Measured with platinum electrodes
and standardized against aqueous solutions. The solid line
is an approximate fit of the data.
43
-------
CONDUCTIVITY / MHO%M 25° c
200 300 400 500 600 700 800 900 1000 1100 1200 1300 WOO
5-
10-
15-
20-
25-
D30-|
E
p
J 35-
H
M
40-
45-
50-
55-
60-
65-
70-1
PEEPER COND.
CI-CVI
Figure 12b.
s 3
Specific conductance in yMHO /cm of "peeper" pore water
samples vs. depth in sediment. Measured with platinum
electrodes in micro-conductivity cell and standardized
against .01 N KC1.
44
-------
TEMPERATURE C°
9 10 II 12 |3 14 15 16 17 18 19 20 21 22 23 24 25 26
Figure 13. In situ sediment temperatures vs. sediment depth measured with
glass bead thermistor and wheatstone bridge.
45
-------
pH
5-
10-
15-
20-
25-
D
P
T
CM
40-
45-
50-
55-
60-
65-
70-
PH
Cn
Cm
ao
6.5 7.0 7.5 8.0
Figure 14. pH of "peeper" pore water samples vs. depth in sediments.
Measurements made in the field with Beckman model G pH meter
and microelectroded pair, except CVI (+).
46
-------
The pH values in Foundry Cove sediments are generally lower than those
of surface marine sediments (see Narragansett Bay data) but are in good
agreement with reducing lake sediment values.
Figure 15 shows peeper pore water chloride values in ppm. The spread
of data from the approximated fit represented by the solid line is about
+_ 10%. Some of the variation can be explained by considering the nature of
the individual profiles comprising the figure. The profile from peeper C I
shows the lowest chloride values at any depth which might be expected since
C I was at a different site than C II - C VI. The high chloride values
between 5 and 40 cm are from profile C VI and result from continued
diffusion downward of saline water in contact with the surface at the time of
C V (dotted line). These observations on the minor variability of sediment
chloride concentrations must be taken into consideration in any discussion
on the degree of equilibration attained. From the comparison of chloride
profiles C I, C II and C III which represent two week equilibration times
and those of C IV, C V and C VI which were only equlibrated 6-7 days, it can
be clearly shown that no significant differences exist in chloride
concentrations between the two sets. This implies that equilibration of a
peeper cell with its surrounding is probably greater than 90% complete for
conservative materials after 7 days.
Molybdate reactive silicate values are shown in Figure 16. Again, there
is a +_ 15% spread around the fitted approximate curve with considerable
overlap in individual curves. (The 7 high points between 8 cm and 26 cm are
from peeper C IVa and probably represent poor data.) The striking difference
between this property and that of chloride is in the shape of the curve.
The chloride profiles show an almost uniform linearly increasing concentration
with depth while those for silicate indicate the importance of exchange
reactions, the presence of and/or dissolution in maintaining a uniform
concentration of 750 y M/l. The equilibration reactions are complex and involve
clays, organic silica and quartz (Garrel and Christi, 1965) but they appear to
be the same in marine sediments also. An examination of data on cores from
Narragansett Bay shows that the 750 y M/l concentration is often reached but
only rarely are higher values found. The general bow shape of the Foundry
Cove curve is also found in many of the ocean cores (Fanning and Pilson,
1974; Schink e_t al_. , 1975).
The molybdate reactive phosphate values from the peeper pore water samples
are shown on Figure 17. The values observed at site C II are enclosed by the
heavy lines while the light line represents the profile for peeper C IV.
The large spread in values is probably partly due to particulate contamination.
A study of the individual profiles shows similarity of shape but slightly
different concentrations for the different locations. The concentration
differences may reflect horizontal inhomogeneities in the phosphate
distribution while the similarity in shape between profiles suggests vertical
similarity in diagenetic and chemical changes with depth from location to
location.
Phosphate values observed in the Foundry Cove sediments are only 1/4 the
values found in Hudson River sediments at mp 18 and probably reflect the
absence of sewage inputs into the cove. In comparison to cores in
Narragansett Bay reported by the URI group, the unpolluted Jamestown N cores
47
-------
CHLORIDE PPM
20 40 60 80 100 120 140 160 180 200 220 2AQ
i i i i i i i i i i_ i i
Figure 15. Chloride concentration vs. depth in sediment of "peeper"
pore water samples in ppm. Measured by silver ion electrochemical
titration. The solid line is an approximate fit of the data.
48
-------
Si02/M
100 200 300 400 500 600 700 800 900
Or—' 1—• 1 1 1—J 1 .1.1.1.1,1
5-
10-
15-
20-
25-
D3°
P
J 35-
H
CM
40-
45-
50-
55-
60-
65-
70-
Si02
Gl-CYL
Figure 16. Reactive silicate concentration in yM/1 vs. sediment depth.
The solid line is an approximate fit of the data.
49
-------
P04 >uM
10 2p 30 40 5(3 60 70 ep 9p |QQ ||Q 120
10-
15-
20-
25-
D
E
P
T
H
30-
35-
CM
40-
45-
50-
55-
60-
65-
70-4
PO,
CI x
crt +
cm A
CH»
Figure 17. Molybdate reactive phosphate concentration in yM/1 vs.
sediment depth. The heavy solid lines represent the limits of
concentrations found at site CII. The light line is the profile
from "peeper" CIV.
50
-------
have concentrations similar to those of Foundry Cove (50-100 y M) while
values in the more polluted Rumstick Neck and Sabin Point cores are
approximately 4 and 10 times higher respectively.
The manganese and iron data are presented as Figure 18 and 19. With
these metals as with phosphate, horizontal inhomogeneities in the sediment
could possibly control the variation of the values: Considerably lower values
were found for the peeper C I Mn profile than those of peepers C II - C IV
taken at site C II. The general uniformity of the Mn values at 2.5 to 3.5
ppm and those of iron around 16 to 22 ppm over nearly the total length of
the peeper suggest that the reducing conditions necessary for the mobilization
of these metals are attained at or above the 10 cm depth and remain constant
to at least 70 cm. Comparison of Fe and Mn in Foundry Cove sediments which
have sulfide levels < 30 yM, to the sulfide-rich marine sediments of Narragansett
Bay (> 1000 yM sulfide) would not be valid. However, a comparison to low
sulfide lake iron (18 to 20 ppm) and manganese (1 to 3 ppm) , suggesting that
similar chemical equilibria are controlling the values.
The profile for peeper C V on Figure 19 does not agree with the other iron
data and presents a major problem in interpretation of all the iron data.
This peeper was the only Foundry Cove peeper to use Nuclepore (polycarbonate)
filter membrane rather than cellulose-based dialysis membrane. The other
parameters do not display any variation resulting from type of material used,
but the probability exists that one or the other or both of the sets of values
for iron are incorrect. Unfortunately, Mn data was not run for peeper C V
which would have told us whether or not any other metals were similarly
affected. Clearly, more work must be done to understand this problem.
The primary reason for choosing Foundry Cove as the site of our major
effort in peeper studies was the high concentrations of cadmium and nickel
found in the sediments. The concentration of dissolved Cd in the pore waters
must be known accurately if the flux of the metal into the overlying water is
to be determined. Cadmium profiles are shown for three locations in Foundry
Cove (Site C I, C II and B I). Approximate metal concentrations in the
sediment at the three locations are 200 ppm, 400 ppm and 5000 ppm Cd (dry
weight) respectively. The Cd profiles obtained for peeper C I (site C I)
peepers C II, C III and C IV (site C II) and those for peepers B I and B II
(site B I) are shown as Figure 20. A major problem in sampling procedures
was discovered during the B I run. With the extremely low pore water
concentrations of soluble Cd relative to the sediment, inclusion of even a
very small particle of sediment in the sample causes massive contamination
of the sample. This is the most likely explanation for profile B I and
probably accounts for the large number of data points with values over
100 ppb. Since peeper C I was sampled at the same time as B I, the same
type of contamination could be expected. Due to the lower sediment concentra-
tion at the C I site and a better washing of the membrane surface before
sampling, the number of contaminated samples was less. To overcome this
problem a special peeper was used for B II. This peeper had a convex curved
front and did not have a plastic cover plate. This peeper was transported
back to the lab where the foam and membrane were removed together thus
decreasing the possibility of any particles entering the cells.
51
-------
MN PPM
.5 ID 15 2.0 25 10 3.5 4.0 4.5
Figure 18. Total manganese concentration of "peeper" pore water
samples in ppm vs. depth in sediment by graphite furnace
A.A. The solid line represents limits to the data
observed at site CII
52
-------
PPM
0
8 10 12 14 16 18 20 22 24
Figure 19.
"Colorimetric" iron concentration in ppm of "peeper" pore water
samples. The method approximates total iron concentration.
The heavy solid lines indicate the range of concentrations found
at site CI and CII. The light line is the profile for "peeper"
CIV. 53
-------
Ip
CADMIUM PPb
, 1,0 20 3,0 40 5,0 6,0 7.0
90
°'
5
10-
15
20-
25
30
D35
P
T
i | AQ
CM
45-
50
55
60
65
70
.
9
X O-4 -
•
X 0
•
O X
*
r
JC O
4
e
«o
«,- 0
•
JfO
*
4
1 o jf
±
i o -*r
*
OJT
-to
-P x
1
K>
K> A'
0+ U1
*
0
A
t >r
h
0 •
j-
o Jr
•CI
« ^S1
+ CSI
A
O
A
O
A
3
O
0 A
A *w
0 A
O A
° V
0 A
**
o *— *
O A
a>
0 A
0
A 0-*
A »*5
A *tlJ
A *-*
0 A
A
A «-*
100
A 0
A *a.
IOOO
•/TO
v»«
° OBI
ABE
o
0
Figure 20. Cadmium concentration of "peeper" pore water samples with depth.
The samples from peepers CI, CII, CIV and CVI are on the left.
Samples from peepers BI and BII (near plant outfall) are
presented center to right. All concentrations are in ppb.
(The large number of points on the right margin represent off
scale values from "peeper" BI where particulate contamination
of the "peeper" water sample resulted in contamination.)
54
-------
Figure 21. Profiles of cobalt-60 and cesium-137 radioactivities with depth
in the sediment at Foundry^ove site CII (middle of inner cove).
Concentrations in pico(10 ) curies per kg dry sediment.
55
-------
(Acid was added to the chambers before the samples were removed to redissolve
any material that might have left solution during the 3 hour time period
before sampling.) As can be seen from the graph, none of the points for
peeper B II have values greater than 110 ppb, and the scatter of the data in
the profile has been reduced. While the error with the sampling and
chemical analysis may be as great as +_ 30%, the approximate shape of the
profile is clear. Low values of 10-20 ppb are found in the top 10 cm rapidly
increasing to 70-90 ppb in the 10 to 30 cm region, after which values again
drop. This profile is in sharp contrast to the values from peepers C I
through C VI where the majority of values are below 10 ppb and close to
3.0 +_ 2 ppb. This factor of 10 difference is approximately equal to the
difference in total sediment Cd concentrations. This difference is not in
agreement with the findings of Bondietti et al. (1973) who found approximately
equal pore water solution levels (^ 10 ppb) in high (35,000 ppm) and low
(2100 ppm) Cd sediments in the cove.
Figure 21 shows the distribution of Co and Cs in pico curies per
kilogram of dry sediment at peeper site C II. The main injection of these
isotopes to the Hudson from weapons testing ( Cs) and reactor releases
( Cs and Co) occurred over the last 20 years and their present distribution
gives some indication of the depth to which mixing (turbation) may have occurred
and/or the rate of sediment accumulation at the site. We have not used the
term bioturbation since, while this may be important, we have observed that
considerable thicknesses (10-20 cm) of sediment can become frozen to the
bottom of the overlying ice during low tides in the winter with resulting
movement and mixing when the ice melts. The profiles indicate that mixing to a
depth of 20 cm may have occurred at site CII since the late 1950's (cores taken
in other parts of the cove show similar profiles including the one taken at site
B) . If the input of Cd-Ni wastes began in the mid-1950's, contamination of the
sediment might be expected to 25 or 30 cm. Cadmium data from peeper profile
B II shows high values to a sediment depth of 35 to 40 cm. At present we don't
have any definite explanation of why the pore water data indicate deeper
contamination with Cd than the sediment Cs data, but one possibility is
compression of the sediment column during coring and extrusion of the sediment
cores.
Several other tests were run on pore water samples that are included
as figures. Ammonium concentrations in Foundry Cove sediments are between
10 and 500 y M with approximately linear profiles from the surface to
70 cm. Positive sulfide concentrations have been randomly found but never
have the values exceeded 15 y 11. Methane profiles are variable and
reflect bubble formation and gas loss during sampling. The methane
concentrations are high (200+ y M) which are comparable to those found for
Narragansett Bay by the URI group. T, CO measurements of Foundry Cove
peepers have been made and are in the range of 1 to 10 m M/l. Other trace
metals (Cu, Zn and Ni) were run on a few sets of samples but technical and
blank problems cast some suspicion on their accuracy.
In conjunction with the development of the peeper, several profiles were
measured in the lower Hudson estuary off Alpine, New Jersey (mp 18) during
1974-1976. The profiles of PO , SiO and Cl with depth are presented in
Figures 22, 23a and 23 b respectively. In comparison to Foundry Cove, the Alpine
site is distinctly brackish for the majority of the year with salinities of
5°/00 to 10°/00. The overlying water contains 6 to 10 times the soluble
56
-------
100 200 300 400 500
10
h- x
E
o
Q.
UJ
O
15
20
25
30
35
40 h
X
X
• 8/74
D 062775
X 072876
n
a
Figure 22. "Peeper" pore water molybdate reactive phosphate values in
yM/1 vs. depth in Hudson sediments off Alpine, New Jersey.
57
-------
600 700
x
o
Figure 23a. "Peeper" pore water reactive silicate in uM/1 vs. depth in
Hudson sediments off Alpine, New Jersey.
58
-------
0
15
£ 20
CJ
Q_
UJ
25
30
35
40
cr %0
234
Fe PPM
5 6
x
x
\
• X
,\
D 062775
x 072876
• Fe 072876
Figure 23b.
Chloride concentration in ppt and "colorimetric" iron values
in ppm vs. sediment depth for Hudson/Alpine peepers.
59
-------
phosphate and substantially greater quantities of sewage derived particulates
than at Foundry Cove. This difference may be the reason the peeper
phosphate profiles at Alpine (Figure 22) are 3 to 4 times higher than those
at Foundry Cove, reflecting the large sewage input rate to the lower Hudson
The SiO peeper profile from Alpine on 22 June 1975 (062775) (Figure 23a)
is very similar to those from Foundry Cove except that the maximum values are
slightly less. The SiO profile for 072876 presents a major problem in
interpretation. This profile and the corresponding PO profile (Figure 22,
072876) are both approximately 3 times lower than they were in previous years
and do not show the sharp gradients at the surface that the other profiles
do. Several explanations are possible for these discrepancies. The two
which seem the most likely are that the peeper was not inserted correctly
and substantial exchange with the overlying water occurred along
the back, sides, and face or that the peeper was inserted into a pocket of
recently resuspended or deposited sediment. The chlorinities for two of
the profiles are shown in Figure 23b and represent the average of many tidal
cycles. A single iron profile for the peeper (072876) is also shown on
Figure 23b.
SUMMARY OF PORE WATER RESULTS
The data from Foundry Cove indicates that for conservative species such as
chloride, the equilibration of the peeper cells with the pore waters is nearly
complete after a week and approximately linear curves were observed. Non-
conservative materials were controlled by dissolution and/or absorption
reactions in the sediment and either reflect a quasi equilibrium state as in
the case of silica or display considerably more variability as in the case of
PO and the metals Fe and Mn. The species which display variability in the
sediment with depth also appear to have greater horizontal variability than
Cl , silica, or the major ions. The difference in PO concentration between
Foundry Cove and the Hudson estuary at Alpine is consistent with the direct
correlation found by the URI group between PO levels in the pore water and
level of sewage contamination in the overlying water. The measurements of Cd
in Foundry Cove have shown that with suitable care to avoid contamination
trace metals with low solubilities can be measured in pore waters. Finally,
with the proper choice of models, reaction rates, and diffusion constants
some estimate of the exchange between the sediments and the water column could
be made from the pore water data for each species.
60
-------
SECTION 6
CESIUM-137 AS A TRACER FOR REACTIVE POLLUTANTS IN ESTUARINE SEDIMENTS
INTRODUCTION
A substantial number of the pollutants discharged to natural waters can
be classified as "reactive" in terms of their propensity to be associated with
particles, either in the original effluent or after becoming dispersed in the
receiving water. For example, metals from the electroplating industry and
some types of artificial radionuclides released from nuclear power plants are
transported and accumulated on particles in natural waters, as well as in
solution. The particles which are most important in reactive pollutant
transport are usually relatively small and often contain both organic and
inorganic components. We will not discuss the composition or sorption
characteristics of these fine particles, but instead will describe some of
their characteristics as vectors of pollutant dispersal and accumulation.
In estuaries, fine particles (< 63 microns) are quite mobile and often
undergo many episodes of deposition and resuspension by the variable currents
of tidal waters. In theory, it should be possible to describe and predict the
pathways of fine particle transport in estuaries, based on the physics of the
particle motions and numerical models of sufficient complexity or from
properly-scaled physical models. Actually, it is more practical to make
direct field measurements of particle transport or to use tracers to infer
the net motion of particles over extended periods of time. The approach
described here uses a "natural" tracer (cesium-137), which has become
associated with fine particles in estuaries, as a guide to the distribution
and transport of fine-grained sediments and several types of pollutants.
The pattern of accumulation of fine particles in estuarine sediments is
complex and essentially unique to each estuary. As a first approximation,
estuarine sediments can be grouped into three end members: (1) large mineral
particles, such as quartz sands, which are relatively unimportant in the
transport of reactive pollutants; (2) fine particles (generally < 63 microns)
which have not acquired significant quantities of pollutants, primarily
because they have had relatively little contact with soluble phase pollutants;
and (3) fine particles with readily-measurable quantities of pollutants, which
will be referred to here as "recent" fines. Obviously the degree of
contamination of recent fines can be extremely variable, but as will be shown
in the case of the Hudson River estuary (U.S.A.), there is often a relatively
uniform dispersal of reactive pollutants in recent fine particles over large
areas and a surprisingly coherent distribution of several types of pollutants.
61
-------
CESIUM-137 AS AN INDICATOR OF RECENT SEDIMENTS
Atmospheric testing of large nuclear weapons during the 1950's and early
1960's, predominantly by the U.S.A. and U.S.S.R., dispersed a great variety
of radionuclides over the entire earth. A number of these nuclides have
sufficiently long radioactive half-lives to be valuable as tracers of global
scale processes. The pattern and time scale of deposition of strontium-90
(t 1/2 ^ 29 years), especially in the northern hemisphere, have been followed
closely (Volchok, 1966; Volchok and Kleinman, 1971) because of its long half-
life, potentially serious biological impact, and the existence of relatively
direct pathways by which this nuclide can reach man. The depositional history
of Cs-137 (t 1/2 ^ 30 years) has not been documented as well as Sr-90 because
it does not appear to be of nearly as much biological concern to man as Sr-90.
Available data indicate that the pattern of delivery of atmospheric fallout
Cs-137 to the earth's surface can be assumed to be identical to Sr-90, with
an activity ratio of Cs-137 to Sr-90 of ^ 1.5 (Harley _et_ _al. , 1965; Hardy,
1974). The peak delivery of fallout Cs-137 to the earth's surface by rain
and snow occurred during the years 1962-1964, and the quantities deposited
since then have been relatively small. Most of the Sr-90 (and Cs-137) fallout
on land has been retained in the upper 10-20 cm of the soil profile and the
total activity present per unit area is proportional to the annual rainfall
(Hardy and Alexander, 1962) as well as being a function of the latitude
(Volchok, 1966).
In the open ocean, both Sr-90 and Cs-137 appear to have remained
predominantly in solution (Broecker et_ _aJL. , 1966; Folsom et_ jil_. , 1970; Bowen
and Roether, 1973), although there is some indication of preferential removal
of Cs-137 into the sediments (Noshkin and Bowen, 1973). The fraction of
total fallout Cs-137 delivered to the ocean which is now in the sediments is
quite small. In most fresh water lakes, Sr-90 stays in solution to the first
approximation, but Cs-137 is nearly completely removed onto particles
(Wahlgren and Marshall, 1975; Farmer et_ _al_., 1977). In rivers and estuaries,
the fraction of fallout Cs-137 associated with sediment particles (compared
with that which passed through these systems in solution) is not well-defined
(Riel, 1972), but readily measurable amounts are found in the sediments of
estuaries which we have studied.
We usually measure Cs-137 in estuarine sediments by gamma counting 50-100
gram samples of dried sediment which have not undergone chemical steps to
enrich the specific activity of the samples. Our counting equipment consists
of a high resolution lithium-drifted germanium detector and a multichannel
analyzer, which allows us to simultaneously measure the activity of many
other radionuclides (both natural and artificial) as well as the Cs-137 gamma
emission peak at 662 Kev. Because of our ability to measure Cs-137 at
"normal" environmental levels in sediments with non-destructive gamma counting
we are able to process a large number of samples with relatively little effort
in laboratory preparation, compared with the analytical techniques required
for most pollutant measurements. The detection limit for most of our samples
was 10-20 pCi/kg, which is a few percent of the activity typical of surface
soils in the northern hemisphere.
62
-------
CESIUM-137 AND OTHER ANTHROPOGENIC COMPONENTS IN HUDSON ESTUARY SEDIMENTS
The total delivery of fallout Cs-137 to the Hudson estuary, decay
corrected to 1975, has been about 120 mCi/km (USERDA, 1975). There is an
additional supply of Cs-137 from a nuclear electrical-generating facility
located near the upstream end of the salinity intrusion in the Hudson. The
total release of Cs-137 from this facility over more than a decade of operation
has been comparable to the amount supplied by rain to the surface of the Hudson
estuary from global fallout. Thus the direct supply of Cs-137 to the Hudson
estuary is roughly a factor of two greater than might be expected if fallout
were the only source.
The specific activity of Cs-137 in surface sediments in the Hudson ranges
over more than two orders of magnitude, with the lowest values in sandy
sediments (typical of areas scoured of fine particles by strong currents).
Fine-grained surface sediments (< 63 y) usually range between 0.2 and 2 pCi/g
of Cs-137, which is comparable to fallout Cs-137 activity in surface soils
throughout the northern hemisphere (Hardy, 1974; Ritchie et_ al., 1975). There
is large variation in the depth to which Cs-137 is found in sediment cores.
In most areas Cs-137 activity is confined to the upper 5 cm of the sediment
column whereas in some it extends to nearly 3 meters below the sediment
surface. Thus the integrated amount of Cs-137 per unit sediment area is not
uniform, and ranges over more than two orders of magnitude. As a result,
relatively limited geographical areas account for large portions of the total
sediment burden of Cs-137. In the Hudson estuary (Figure 24) the dominant
areas of Cs-137 accumulation are the harbor and shallow coves upstream of the
harbor. These areas are not in close proximity to the site of localized
discharge of Cs-137 to the Hudson, and primarily reflect the zones in which
fine particles are rapidly accumulating (Simpson et al., 1976).
We have found the distribution of other man-made reactive contaminants in
Hudson sediments to be very similar to that of Cs-137, despite significant
differences in chemistry, and mode of input to the system. The locations of
sediment sampling sites for data reported here are shown in Figure 24. These
sites extend from approximately the upstream limit of salinity intrusion
during summer months, to the harbor area which typically has salinities of
approximately two-thirds that of sea water.
In Figure 25, activities of Pu-239,240, determined by alpha spectrometry
following chemical separation procedures as described by Wong (1971), are
plotted against Cs-137 in the same samples. The covariance over two orders of
magnitude of these two parameters in Hudson sediments is clear. Thus, if the
present distribution of Pu-239,240 in Hudson sediments (mostly derived from
fallout) were to be measured, the most efficient procedure would be to use the
distribution of Cs-137, which is relatively easy to measure by gamma
spectrometry, to guide the selection of samples for Pu-239,240 analysis.
(The alpha particle energies" of Pu-239 and Pu-240 are nearly identical and
the sum of their activities is usually reported.)
In Figure 26, the concentration of polychlorinated biphenyls (PCB's) in
Hudson sediments is plotted .against Cs-137. Although our data is limited at
this time, the covariance of these constituents is also obvious. The levels
of PCB's are high in sediments over large areas of the Hudson because of
63
-------
HUDSON
Hudson Drainage Basin
ESTUARY
4l°30'«
Figure 24. Locations ( ® ) for which data in Figures 25, 26, 27 and 28
are reported. Locations of discharge of polychlorinated
biphenyls (PCB's), cadmium and nickel (Cd, Ni) and radioactive
cesium (Cs-137) are also indicated.
64
-------
Jjoo
CT>
-XL
\
o
o
CM
CM
£- 10
100
Cs-137 (pCi/kg)
1000
,10
00
o
Q- I
100
Cs-137 (pCi/kg)-
1000
137 239,240
Figure 25. Activities of Cs and ' Pu in Hudson Estuary sediment (dry weight) core samples
(Figure 24) including both surface samples and a number from well below the sediment-
water interface. Two suspended particulate samples (^ ) collected near the middle of
the sampling range are included.
Figure 26. Activities of Cs and concentrations of polychlorinated biphenyl's (PCB's) in samples
of surface sediment (Figure 24). All data is expressed in terms of dry weight.
-------
industrial releases during the 1950's and 1960's at two sites more than
200 km upstream of the locations of our sampling area. Considering the great
differences in chemistry between Cs-137 and PCB's, it is perhaps surprising
to find their sediment distributions to be as similar as they are, but their
covariance is a good indicator of the ability of fine particles to transport
and accumulate quite a variety of reactive pollutants.
In Figure 27 is shown the concentration of several trace metals relative
to Cs-137. Zinc, copper and lead concentrations in recent Hudson sediments
are several times the concentration levels in pre-industrial sediments. All
of the samples shown in Figure 27 are upstream of the harbor area, and thus
reflect diffuse sources of these metals to the Hudson over a number of decades.
Sediment samples from New York harbor have somewhat higher concentrations for
all three metals because of electroplating and other industrial discharges.
Vertical distributions of all three metals in harbor sediments also are similar
to that of Cs-137.
In Figure 28 the concentrations of cadmium and nickel in Foundry Cove at
mp 53 are plotted against Cs-137 activity. High level contamination of the
sediments of this small (^ 0.5 km ) shallow (mean depth ^ 1-2 meters) area by
effluent from a battery factory has resulted in Cd concentrations ranging from
a few percent to '\< 100 ppm (Bower, 1976). Some surface sediments in the cove
which are apparently in areas of active current scouring contain relatively low
concentrations of Cd, Ni and Cs-137. Thus Cs-137 is useful in mapping the
pattern of trace metal accumulation in sediments in relatively small, highly
contaminated areas, as well as for diffuse sources over large areas.
All of the "reactive" pollutants we measured in Hudson estuary sediments
are found preferentially on fine particles and in sediments rich in organic
matter, as would be expected. However, many sediment samples with very
similar particle size distributions and organic carbon contents did not have
appreciable reactive pollutant, concentrations (and they did not have measurable
Cs-137). Thus the activity of Cs-137 was a much more accurate indicator of
probable pollutant concentration than were classical sedimentological
techniques, especially in estimating the depth to which appreciable pollutant
concentrations would be found in estuary sediment cores.
CESIUM-137 AS A POLLUTANT TRACER IN OTHER AQUEOUS SYSTEMS
Fallout Cs-137 has been used as an indicator of recent sediments in the
Delaware Estuary. Sites with appreciable activity of Cs-137 in surface
sediments also have hydrocarbon constituents typical of recent pollution,
whereas surface sediments free of Cs-137 have hydrocarbons typical of
unpolluted marsh sources (Wehmiller, personal communication). In lake
sediments, the distribution of Cs-137 has been shown to be nonuniform in large
lakes and to be closely related to that of fallout Pu-239,240 (Edgington et al.
1976).
The concentration of a number of reactive pollutants in Hudson estuary
sediments, although extremely variable in both surface and depth distributions,
has been shown to have considerable coherence from one pollutant to another and
to have a strong correlation with Cs-137. Thus the task of mapping contamin-
ated sediment distributions in complicated sedimentary regimes can be simpli-
fied" through the use of a "natural" tracer, Cs-137.
66
-------
300 -
200
.0
CL.
O
100
1000 2000
Cs-137 (pCi/kg)
3000
100
en
E
TJ
O
10
100 1000
Cs-137 (pCi/kg) >-
137
Figure 27. Activities of Cs and concentrations of zinc ( X ), copper ( ^ ) and lead ( Q ) i-
sediments (Figure 24) from both surface samples and samples well below the sediment
water interface.
Figure 28. Activities of Cs and concentrations of cadmium ( • ) and nickel ( ® ) in sediments from
the least contaminated area of Foundry Cove (Figure 24).
-------
SECTION 7
RADIOCARBON GEOCHEMISTRY IN THE ESTUARY OF THE HUDSON RIVER
INTRODUCTION
Estuaries can be characterized as complex, three-dimensional zones of
mixing of river water and sea water in semi-enclosed areas bounded by land
surfaces. Many processes in estuaries are variable on several scales in both
time and space. Since a number of estuaries have become the sites of major
population centers, the spectrum of natural estuarine processes is now often
intricately interwoven with the effects of urban industrial activities. Thus,
estuarine processes are important not only to understanding natural chemical
cycles but also for management of activities as diverse as disposal of sewage
and industrial wastes and maintenance of adequate navigation depths in rapidly
shoaling harbors.
Although radiocarbon (C-14) has been extensively used in reconstructing
sedimentation history in estuaries, especially over the period since the end
of the last glacial maximum, it has been exploited in relatively few studies
of processes in urban estuaries. One type of tracer application which has
been made is in studies of organic material in aqueous systems significantly
perturbed by man (Rosen and Rubin, 1964; Kolle et^ al., 1972; Zafiriou, 1973;
Erlenkeuser _et_ jil_. , 1974; Spiker, 1976; Spiker and Rubin, 1975; Wakeham and
Carpenter, 1976). Another potentially valuable application is in establishing
the rate of gas exchange between the atmosphere and estuarine waters, one of
the critical processes in maintaining the balance of oxygen in estuaries
receiving large amounts of sewage.
There are a number of factors which make the distribution of C-14 in
estuarine sediments difficult to interpret. Sediment accumulation patterns
tend to be complex, with large variations in net deposition rates from area to
area within an estuary and considerable reworking and movement of materials
likely to occur after they reach the sediment surface. In addition, dredging
activities can greatly perturb natural sediment deposition patterns (Simmons
and Hermann, 1972). Thus the apparent C-14 age of carbon in sediments is
likely to be a complicated function of the deposition and reworking history
of the sediment. In many estuaries, most of the carbon in fine-grained
sediment is in organic matter, as opposed to calcium carbonate shells or
detrital carbonate from the drainage basin. This presents additional
difficulty due to the number of important sources of organic carbon in
estuarine sediments. Even carbonate shells, which are probably more
desirable as dating materials for estimating accumulation rates of estuarine
sediments, also present difficulty in interpretation. Fresh water dissolved
inorganic carbon in many lakes (Deevey et al., 1954) and rivers (Broecker and
68
-------
Walton, 1965) has long been known to be deficient in C-14 relative to the
atmosphere. Estuaries can also be expected to have C-14 deficient inorganic
carbon. The magnitude of the effect is a function of the proportion of
weathering of carbonate rocks relative to silicate rocks which occurs in
the drainage basin, and the degree of equilibrium with atmospheric C-14 by
gas exchange which is accomplished in the riverine (Broecker and Walton, 1959)
and estuarine portions of the system. The extent to which carbonate shells
reflect the carbon isotopic ratio of dissolved inorganic carbon is somewhat
in question (Keith and Anderson, 1963; Rubin and Taylor, 1963; Broecker, 1964;
Keith, Anderson and Eichler, 1964; Mook and Vogel, 1968; Fritz and Poplawski,
1974) but appears less of a source of uncertainty than other factors in
estuarine systems. Equilibration of the C-14/C-12 ratio of fossil estuarine
carbonate shells with dissolved inorganic carbon cannot be directly
verified by studies of currently-forming carbonate shells because of the
presence in the atmosphere and aqueous carbon systems of C-14 produced from
atmospheric testing of nuclear weapons. This perturbation does, however,
provide additional opportunities for using bomb C-14 as an estuarine tracer.
HUDSON ESTUARY MORPHOLOGY AND SEDIMENTATION HISTORY
Estuaries are transitory coastline features produced primarily by
the rapid rise in sea level as ice melted following the most recent maximum
of continental glaciation (> 18,000 B.P.). Along the eastern coast of the
United States, much of the detrital load of rivers settles out before reaching
the continental shelf (Emery, 1967), the estuaries of the isostatically-more
stable southeastern U.S. coastline becoming filled with sediments to a greater
extent than those farther north (Mead, 1969).
The Hudson River (Figure 29) is tidal for approximately 250 kilometers
between New York City and a dam located north of Albany, New York. Upstream
of the tidal waters, fresh water is derived from two main sources, the upper
Hudson River draining crystalline rocks of the Adirondack mountains and the
Mohawk River flowing through sedimentary terrain with large areas of carbonate
rocks as well as evaporites. About two thirds of the total fresh water runoff
of the Hudson is derived upstream of tidal waters, and much of the dissolved
inorganic carbon comes from the Mohawk. Mean annual discharge of the Hudson
to New York harbor is about 550 m3/sec, with monthly means usually ranging
between 150 nrVsec and 1500 m^/sec. Saline water intrudes one-third to one-
half of the total reach of tidal water during low fresh water flow but
occupies only the area downstream of the northern end of Manhattan Island
during seasonal high flow. Upstream of the southern tip of Manhattan (mile
point 0), the Hudson has a remarkably constant total cross-sectional area for
approximately half of the reach to the head of tidal influence, despite the
fact that the channel transects at least four quite distinct geomorphic
provinces (see Sanders, 1974). Downstream of mile point (mp) 0, to the
mouth of the estuary at approximately mp -15, the geometry is extremely
complex, and includes a number of tidal straits connecting several segments
of New York harbor with the New York Bight and Long Island Sound.
The entire Hudson channel is believed to have been excavated by fluvial
processes during the Tertiary, and over-deepened by ice scouring during the
glacial maxima (Lovegreen, 1974). Downstream from about mp 100, where
basement lies about 60 meters below current sea level, the bedrock channel
69
-------
LOCATION OF CORES
4I°30
4I°00
40°30
74° 30
74°00'
73°30'
14,
Figure 29. Location of cores for which C data are reported in shell layers
70
-------
deepens to about 200 meters between mp 50 and 20 (Worzel and Drake, 1959)
and then shallows again to about 60 meters at the Narrows (mp -6), indicating
that the Hudson estuary probably should be described as a fjord (Newman et al.,
1969) which is now largely filled with sediment. Seismic profiles (Worz"eT and
Drake, 1959) and boring information (Wiess, 1974; Newman et_ al., 1969) have
established the general nature of the material accumulated~iiT"the Hudson since
ice scouring ceased. The lower fill is glacial drift, consisting of gravels
and till, which is overlain in most borings by one or more sequences of
lacustrine varved sediments representing 2000-3000 years of accumulation. The
lake sediments were apparently formed when a terminal moraine crossing the
Narrows near the mouth of the present estuary served to dam up the glacial melt
waters. The uppermost layer of sediments is estuarine organic silt, repre-
senting in most sections a third or less of the total sediment thickness. The
oldest reported C-14 dates of fill materials are 12-13,000 radiocarbon years
(Newman et al. , 1969) for peats formed at the base of the estuarine silt
accumulation. Newman et al. (1969) believe that there was a substantial
errosional event between the end of lacustrine sediment deposition and the
onset of estuarine silt accumulation, possibly related to a catastrophic
decanting of the large proglacial lake(s) occupying the Hudson Valley. The
deposits of estuarine silt are generally continuous between the top of the
lacustrine sediments and the present sediment surface. The total thickness
of estuarine organic silt is generally between 10 and 60 meters, with the
greatest accumulation in areas with deepest bedrock. Pollen zonation and a
limited number of C-14 dates from samples of engineering borings have been
used to establish the history of silt accumulation and foraminifera assemblages
have been used to delineate the variations in mean salinity intrusion since
estuarine conditions were established (Newman _e_t _aJL. , 1969; Weiss, 1974). The
published data are not sufficient to establish a detailed history of
sedimentation of estuarine silts, but accumulation rates appear to have been
reasonably uniform with time. Assuming a mean total accumulation of 30 meters
of estuarine sediment in 10-12,000 years, the sedimentation rate in the Hudson
estuary averaged over the Holocene has been about 3 mm/yr.
There have been major perturbations of the sedimentary regime of the
Hudson over the past century. During the past 70 years, extensive dredging has
occurred in New York harbor mostly between the mouth of the estuary (mp -15)
and mp 12. Initially, dredging was primarily for deepening of docking
facilities and some navigation channels, but during the past twenty years or
so, sediment removal has been largely directed at maintaining the present
harbor depth in the face of rapid shoaling in a few zones, especially between
mp 0 and mp 12 (Panuzio, 1965). Dredging activities since World War II have
annually removed on the average about 4 million metric tons (dry weight)
(Panuzio, 1965; Gross, 1972). Dredging in the main channel of the Hudson
outside of New York harbor has been confined to a few areas. The most
extensive was in the tidal river reach of the Hudson north of mp 100 to allow
Albany to be developed as a port for deep water vessels.
During the past few years we have obtained a considerable amount of
information about recent (last two decades) sediment accumulation in the
Hudson by measuring the distribution of cesium-137 in sediments (Olsen et al.,
1978). This nuclide has been supplied to the Hudson since the middle 1950's
by global fallout from nuclear weapons testing, with peak delivery from
rainfall during the years 1963-1965, and since the middle 1960's by low-level
71
-------
releases from a commercial nuclear power plant located at mp 43, with peak
discharges in 1971-1972 (Simpson et_ al_. , 1976). The relative contributions
of fallout and reactor Cs-137 can be estimated from the activities in the
sediments of Cs-134 and Co-60, both of which were derived almost exclusively
from the reactor. Using Cs-137, which is found in the fine-grained estuarine
sediments of the Hudson in readily-measurable activities, as an indicator of
sediments which have been in contact with the water column during the past
two decades, we have mapped the accumulation pattern of recent sediments.
Deposition rates range over approximately two orders of magnitude, with New
York harbor between mp 10 and mp -2 being the dominant zone of sedimentation,
having accumulations of 10-20 cm/yr of fine-grained particles over large areas
of the harbor. Outside of the harbor, recent net deposition rates are
considerably less than a cm/yr over most of the total area, with marginal
coves between mp 56 and mp 43 and areas of the western side of the broad reach
of the estuary between mp 40 and mp 20 having accumulations of up to a few
cm/yr. Downstream migration of remobilized fine sediment particles from mp 40
to mp 20 in the harbor occurs on the time scale of a few years, as indicated
by the history of arrival in the harbor of Cs-137 released at the reactor
site at mp 43. Thus, present sediment accumulation patterns in the Hudson
estuary are extremely variable, ranging from areas of active scour and little
net deposition to those in which dredging is employed to maintain channel
depths in zones accumulating sediment at a rate two orders of magnitude
greater than the average accumulation rate over the past 10,000 years.
SAMPLE DESCRIPTION AND PREPARATION PROCEDURES FOR HUDSON RADIOCARBON
MEASUREMENTS
We have determined C-14/C-12 ratios in several types of samples from the
Hudson: (1) dissolved inorganic carbon from above the upstream end of tidal
water (mp 154) to the coastal waters adjacent to the mouth of the Hudson
estuary; (2) organic matter from sediments between mp 41 and mp 0, and in
recent sewage sludge; (3) shells from sediment cores in a restricted reach
of the estuary between mp 24 and mp 18, plus shells collected from the fresh
water reach of the tidal Hudson in 1887.
Data are reported in Table 13 for the C-14/C-12 ratio in dissolved
inorganic carbon in a number of large water samples (100-200 liters) from the
Hudson collected during the last several years. Suspended particles were
usually removed from the samples with a high flow continuous centrifuge and
carbon dioxide was chemically precipitated in concentrated NaOH with a closed
circulating gas system after acidifying the centrifuged water samples.
Counting data are given in terms of 6C-14 in which:
(C-14/C-12) - (C-14/C-1,., ^ , ,
/o/ x sample standard
(/oo) = (c-i4/c-i2)—— * 100°
standard
The standard for which our data are reported is modern wood (0.95 x NBS
Oxalic Acid Standard). In this notation, modern wood, free of bomb C-14,
would have 6C-14 = 0°/oo, a sample with no C-14 would have 6C-14 = -1000°/
and the atmosphere in 1975 had. a 6C-14 of approximately +400°/00 due to the
presence of bomb C-14. The standard error (la) we have quoted is based on
counting statistics and variations in the background characteristics of our
72
O O
-------
TABLE 13
14
C Measurements of Inorganic Carbon in Water Samples
from the Hudson River Drainage Basin
13
Lamont-Doherty
Sample No.
1439 G
1439 H
1439 C
1439 A
1439 B
1439 D
1439 Ea
1371 B
1371 A
Sample
Description
Mohawk River
Upper Hudson River
Hudson River
(mp 61; S=0°/00)
Hudson Estuary
(mp 18; S^5V00)
Hudson Estuary
(mp 18; S^5°/00)
Hudson Estuary
(mp -7; S
-------
counters. Stable carbon isotope ratios for all of our samples were measured
on splits of the purified CO,., used for C-14 countings and not from a separate
preparation. The 6C-13 data are quoted relative to the PDB-scale (Craig, 1957)
with standard errors of replicates approximately + 0.2°/00, not including
fractionation and effects during the CO^ separation and purification. All
of the water samples had 6C-14 values greater than modern (0°/00) and thus
contained appreciable bomb C-14. Most of the reported data were from a
relatively short period of time and are indicative of the water column
inorganic carbon C-14 for 1976 during flow conditions somewhat less than mean
annual flow.
Samples of organic matter in surface sediments from three sites in the
Hudson estuary were analyzed for C-14 (Table 14). All of the organic matter
samples were leached with dilute acid (2% HCl) to remove carbonates and dilute
base (2% NaOH) to remove humic acids before burning to collect carbon dioxide
from the residual organics. The humic acid fraction of two of the samples
were also analyzed for C-14. These fractions from mp 41 and mp 18 yielded
apparent C-14 ages of 1500-1900 radiocarbon years, while the residual organics
had apparent C-14 ages of 4100-4600 radiocarbon years. Sediment organic
matter from a rapidly shoaling area of New York harbor (mp 0) and sewage
sludge (preparation of this sample did not include treatment with acid and
base) from a New York City secondary treatment plant both had appreciable
bomb C-14, yielding future ages when expressed in terms of C-14 dates.
Carbonate shell materials are not common in Hudson sediments. We have
collected gravity cores (^0.5 meter length) at approximately a hundred sites
in the Hudson between mp 60 and mp -2 and have found only one zone with
significant shell material. This area, between mp 24 and mp 18, currently
appears to have little net accumulation of fine-grained recent sediment except
on the shallow subtidal bank along the western shore, based on the distribution
of Cs-137 in the sediment (Simpson et_ al., 1976). Considerable net downstream
transport of fine particles bearing Cs-137 occurs through this reach, but
the main channel is predominantly covered by a coarse sand lag deposit. The
main channel is not an area of active dredging, and is deeper (_>_ 15 meters)
than the project depths maintained by the Corps of Engineers in New York
harbor. Near the navigation channel of the Hudson between mp 24 and mp 18
we have observed a number of distinct subsurface shell layers, up to 10 cm
in thickness. These layers consist almost exclusively of many single valves
of Mulinia lateralis, a small marine mollusk, and coarse sand fragments,
predominantly quartz. At several sites containing shell layers, we collected
a number of duplicate ^0.5 meter gravity cores, a few small diameter 3 meter
piston cores and two 6 meter piston cores, the latter from Lamont-Doherty's
oceanographic vessel VEMA. In general, the shell layers are lenticular and
appear to be continuous only on a horizontal scale of less than 50 meters,
and cannot be visually correlated between cores except when duplicate cores
are collected or precisely located sites are reoccupied.
The locations of the sediment samples from which shell materials have
been analyzed for C-14 are shown in Figure 29. All but four of the reported
samples with C-14 dates between 1000 and 3000 radiocarbon years (Table 15)
were from subsurface layers of Mulinia lateralis. Four samples consisted
of detatched valves of Macoma balthica and large oyster fragments (Crassostrea
virginica) two of which were collected from the top of cores and a third from
74
-------
14
Lamont-Doherty
Sample No.
TABLE 14
Measurements of Organic Carbon in Surface Sediments from the Hudson Estuary
,14 ^13
Sample
Description
6CJ
(V )
\ / o O '
Apparent Age
(Radiocarbon Years)
1378 BH
1378 B
1378 AH
1378 C
1439 F
Humic acid fraction
of organic matter
(mp 41)
Residual organic fraction
after removal of humic
acids (mp 41)
Humic acid fraction of
organic matter (mp 18)
Residual organic fraction
after removal of humic
acids (mp 18)
Organic matter - whole
sample with no fractions
removed (mp 0)
Sewage sludge - N.Y. City
treatment plant (1975)
-167 + 35
-403 +_ 12
-209 +_ 51
-436 +_ 35
+381 +_ 46
+226 + 17
-36.1
-27.3
-27.4
1470 +_ 340
4140 +_ 160
1890 +_ 520
4600 +_ 500
Future Age
Future Age
a) This sample was from the Ward's Island Sewage Treatment Plant in New York City and was not a
sediment sample from the Hudson.
-------
Table 15
Dates of Carbonate Shells from Hudson Sediments
Lamont
Sample
No.
13631
1363D
1363C
1363B
1363G
1363E
1363J
1363F
1363K
1363L
1363M
1363N
1387Ab
1387D
1387C
1387E
1387B
1363H°
1363A
1377A6
Core
No.
A
B
B
B
C
C
C
C
D
E
E
E
F
F
F
F
F
G
H
Core
Location
(mp)
24
22
22
22
18.
18.
18.
18.
18.
18.
18.
18.
18.
18.
18.
18.
18.
18.
18
98
6
6
6
6
6
6
6
6
5
5
5
5
5
3
Depth
(cm)
260-270
35-45
130-140
200-210
0-5
21-23
27-31
45-47
75-80
110-116
200-212
222-228
0-10
109-111
127-130
199-201
266-271
0-5
137-142
Surface
sc14
("/„„)
-315+_13
-244+21
-289+17
-310+JL1
+455+14
-148+10
-149+14
-152+14
-208+14
-288+19
-138+12
-261+30
+58+19
-130+33
-170+13
-241+35
-228+26
+356+16
-232+11
-76+11
Age
(Radiocarbon
years)
3040+160
2240+230
2740+190
2980+130
-3010+80
1290+100
1290+130
1320+130
1880+140
2720+220
1190+110
2430+330
-450+150
1120+300
1490+130
2210+370
2080+270
-2450+100
2120+110
630+100
6C13
(V )
V / 0 Q '
-11
- 7
- 1
- 3
- 2
-
- 4
-
_
-
-
_
-3
-2
-
-
-4
-2
-11
.4
.2
.6
.2
.2
.1
.1
.6
.2
.9
.2
6C"
Initial
(°/0J
(-20)+10d
-68+5
-43+5
-10+5
(bomb
-13 + 5
(-201+10
-25+5
(-201+10
(-201+10
(-201+10
(-201+10
(bomb
-19+5
-16+5
(-201+10
(-201+10
(bomb
-18+5
-67+11
Corrected
6*C
(VoJ
-295+16
-176+22
-246+18
-300+12
14
cr i
-135+11
-129+17
-127+_15
-188+17
-268+21
-118+16
-241+_33
c14,
-111+33
-154+14
-221+36
-208+28
c14)
-214+12
-
Estimated
Minimum
Age (yrs)
2810+180
1550+210
2270+190
2860+_140
0-20
1160+100
1110+160
1090+140
1670+_170
2510+230
1010+150
2210+_350
0-20
950+300
1340+130
2010+370
1870^280
0-20
1930+120
-
All samples were from discrete subsurface layers of Mulinia lateralis unless otherwise noted.
Core top layer of oyster fragments (Crassostrea virginica) and clam shells (primarily Macoma balthica).
Oyster fragments (C. virginica) and clam valves (M. balthica) from a composite of about a dozen grab samples.
Initial <$C for samples without <5C data were estimated from the observed range of <5C values for similar samples.
Lampsilus radiatus Gmel. (fresh water mollusk) collected by W.S. Treator at Barrytown, New York (mp 98) in 1887.
-------
a surface grab sample. The three surface shell samples all contained bomb
C-14, and hence give future ages when expressed as C-14 dates (Table 15).
One of the subsurface samples (1387 C) consisted of a single oyster valve and
yielded a date of 1490 + 130 radiocarbon years, in agreement with adjacent
layers of Mulinia lateralis. The subsurface shell layers for which dates are
reported in Table 15 were obtained during seven separate coring operations.
Four of the cores (B, C, E, F) have C-14 dates reported for more than one depth
interval in the sediments. Two of the four cores (B, F) which have more than
one depth interval dated show increasing ages with depth. Dates from one
core (C) were obtained by preparing composite samples for several shell layers
from about half a dozen 0.5 meter gravity cores which were collected
sequentially while anchored at one site in the estuary. In this core, the
surface layer of oyster fragments contains bomb C-14, while the three
subsurface layers between 20 and 50 cm have the same age (^ 1100 years) within
the uncertainty of the measurements. One core (E) has an apparent age
reversal with depth. We believe the reversal in core E is real, but there is
a possibility that it resulted from an error in labeling following extrusion
of the core from its plastic liner or during later sampling. Sample G
represents a composite of oyster fragments and clam shells from a large
number of surface grabs collected while anchored at one site.
We have been able to obtain some fresh water mollusk shells which were
collected in 1887 from the Hudson River at mp 98 (tidal fresh water reach of
the Hudson). Measurement of C-14 in one sample consisting of about half a
dozen large single valves gave an apparent age of 630 +_ 100 years (Table 15).
The shells were quite fresh in appearance, having their organic coatings
intact, and were apparently collected alive.
DISCUSSION
Water column 6C-14 values in' the Hudson (Table 13) indicate the presence
of appreciable bomb C-14 in the inorganic carbon pool, but that complete
equilibration with atmospheric C-14 does not occur. Our water samples from
1976 indicate a downstream increase in 6C-14, with the highest value (+200°/00)
about mp 18, followed by decrease toward the coastal ocean as salinity
increased. This decrease is probably largely due to mixing of fresher water
with coastal sea water having a lower value of 6C-14 (+155°/00). One
explanation of the increase of 6C-14 downstream to the maximum at +200°/00
is exchange of atmospheric C02 with the dissolved C02 in the Hudson as water
moves downstream. We can estimate the change which might be expected to
occur as fresh water with 6C-14 of +140°/oo at mp 61 flowed downstream to
mp 18, exchanging carbon dioxide with the atmosphere, using an expression
derived from Broecker and Walton (1959):
(CA - C^) = (CA - CRi)e - Rt/kh
where C = 6C-14 of atmospheric carbon dioxide (estimated to be ^ +400°/00 in
A 1976)
C = 6C-14 of river inorganic carbon at the upstream end of the flow
Rl segment of interest (+140°/0o)
C = 6C-14 of river inorganic carbon at the downstream end of the flow
R2 segment of interest (to be calculated) 2
R = gas exchange rate (estimated to be £ 10 moles/m /year)
77
-------
t = time to transect segment of river (^0.1 year at flow conditions
of '^ 400 m /sec and volume of channel -^ 1.1 x 10 m ) ^
k = concentration of total inorganic carbon in Hudson (^ 1 mole/m )
h = mean depth (^ 10 meters)
If atmospheric 6C-14 was +400°/00, the dissolved inorganic carbon would
increase from +140°/00 to +164°/00 as the Hudson flowed between mp 61 and
mp 18, assuming only downstream advective flow. Our measured data indicates
higher than predicted 6C-14 for mp 18 (average of two samples = +195°/00).
There are several possibilities which could explain the data. We do not have
a well established value for the 6C-14 of the atmosphere in 1976 which was
used in our calculation (+400°/o0)» but it is not likely that the actual
value is significantly higher than the one we used, and thus cannot explain
the higher value at mp 18 than predicted from our gas exchange calculation.
We obtained our value of +400°/o0 by extrapolating the trend of Rafter and
O'Brien (1972) beyond 1972. Published tree-ring data (Cain and Suess, 1976)
from adjacent to Hudson (Bear Mountain, '^ mp 45) in 1970 and a value of +413
+ 12°/00 reported for atmospheric C-14 in Monaco in 1975 (Rapaire and Hugues,
1977) are in good agreement with the trend of Rafter and O'Brien (1972).
Another potential explanation is that the actual gas exchange could be
greater than the first order estimate we made of 10 moles/m /yr, but this
does not seem very likely in light of observed gas exchange rates in marine
and fresh water systems (Broecker and Peng, 1974; Emerson, 1975). The gas
exchange rate which we used is that derived by Broecker and Peng (1974) as an
average for the world ocean (equivalent to a stagnant boundary layer thickness
of 50 microns in a one parameter gas exchange model). Most other published
estimates of carbon dioxide exchange rates in the ocean based on the
distribution of natural C-14 and bomb C-14 (see Broecker and Peng, 1974 for
references), have ranged between a factor of 2-3 faster down to a value
approximating that used here. Thus the value we have used is a reasonable
upper limit for the Hudson, and the most probable value for summertime gas
exchange rates is approximately half of that used for the calculations
reported here (Hammond _et_ al. , 1977).
We can also calculate the expected effect of gas exchange on 6C-14 in
the Hudson as water moves downstream from the head of tidal water (mp 154)
to mp 61. From the 6C-14 data in the Mohawk and Upper Hudson for early
October, the 6C-14 of water in the Hudson was probably about +125°/00, since
the Mohawk supplies more than two thirds of the dissolved inorganic carbon
to the water entering the tidal Hudson at the upstream end. Using C =
+40°/00, CR = +125°/00, R = 1Q moles/m /year, t = 0.09 year (flow conditions
in October 1 were also ^ 400 m /sec and volume of ^ 1.0 x 10 m between
mp 154 and mp 61), k = 1 mole/m , h = 7.5 meters, the calculated 6C-14 value
at C (mp 61) is +157°/oo5 which is greater than the measured value of „
2 +140°/oo. If we were to use a model gas exchange rate of 5 moles/m /yr
measured and calculated 6C-14 would agree. If the total change in 6C-14
between mp 154 (+125) and mp 18 due to gas echange were estimated from the
same parameters as used for the two segments of the total length (R = 10 moles/
m /yr), the calculated 6C-14 value at mp 18 would be +176°/0o, compared with a
measured value of +195. Thus the change in 6C-14 measured between two points
near the extreme ends of the tidal Hudson is the right order of magnitude to
be produced primarily by gas exchange with the atmosphere, but the rate of
change indicated near the downstream end appears substantially higher,
suggesting the possibility of a second important mechanism for increasing 6C-14
78
-------
in the dissolved inorganic carbon in the water column near New York City.
Oxidation of sewage particles, which have a higher C-14/C-12 ratio than
inorganic carbon in the Hudson, seems likely to be important in the C-14 budget
of the dissolved inorganic carbon the estuarine reach of the Hudson near New
York harbor. Two other measurements of dissolved inorganic carbon 6C-14 are
reported for 1974 (Table 13). These two samples were collected at different
t^mes and with differing flow conditions (May ^ 950 m /sec, September ^ 400
m /sec) and thus are not nearly as useful for gas exchange calculations as
1976 data. Both of the 1974 samples had higher 6C-14 than samples collected
at similar locations in 1976.
Spiker and Rubin (1975) have found the proportion of C-14 in dissolved
organic carbon in a number of streams to be dependent upon the amount of
domestic sewage wastes and industrial fossil fuel wastes dicharged. They
calculated the proportion of fossil carbon in the dissolved organic carbon
assuming the contribution of dissolved organic carbon from other sources,
including municipal wastes, had a C-14/C-12 ratio equivalent to the present
atmosphere. Our data indicate that the discharge of municipal wastes also
affects inorganic carbon 6C-14 and tends to bring the water column values
toward equilibrium with the atmosphere. Kolle et_ al. (1972) attributed down-
stream increases in C-14/C-12 in the inorganic carbon of the Rhine River to
the oxidation of municipal wastes dominated by sewage. The downstream trends
of C-14/C-12 in the Rhine are complicated and difficult to interpret in detail
because of the contribution of large quantities of industrial contamination
of C-14 free carbon. In the Hudson, the primary locus of both sewage and
industrial discharge is adjacent to New York City and the level of pollutant
carbon loading to the tidal river section of the Hudson north of this is
relatively low. A sediment sample from New York harbor (mp 0, Table 14) had a
6C-14 value of +381°/00, indicating a significantly higher component of bomb
C-14 than we observed in our water column inorganic carbon measurements.
Sewage sludge from one of the dozen or so large treatment plants in the New
York City area also had appreciable bomb C-14 (Table 14, +226°/00)3 Very
large discharges of sewage are made to the harbor complex (^ 100 m /sec) from
a number of point sources (Simpson et^ al., 1975). Fine-grained sediments from
the harbor lose about 8% of their weight upon heating from 100°C to 500°C over-
night, while sediments farther upstream lose only about 4% of their original
weight. The sediments near New York City have approximately twice as much
organic carbon as average Hudson sediments, and a much higher fraction of
recent sewage organic particles, based on the C-14 data reported in Table 14.
The levels of algal growth in the Hudson estuary are surprisingly low (Malone,
1976; Simpson
-------
Washington and Narragansett Bay, the contamination was assumed to be primarily
petroleum hydrocarbons. In the Hudson, based on our knowledge of general
sediment accumulation and pollutant discharge patterns we would expect the
extent of fossil fuel contamination to be greater at mp 0 than at mp 41.
Recent unpublished gas chromatography data on solvent extracts of Hudson
sediments suggest that a higher proportion of the normal hydrocarbons upstream
of the harbor area are natural, whereas the harbor sediments are dominated by
a complex mixture of unresolved peaks (J. Wehmiller and E. Keenan, personal
communication). The weight fraction of a total pentane extract (TPE) and
normal hydrocarbons (NH) in the harbor sediment sample analyzed for C-14 (mp 0)
are greater (TPE = 0.8%, NH £ 0.15%) than the weight fractions in the upstream
sample (mp 41) analyzed for C-14 (TPE = 0.32%, NH <_ 0.057%). Thus our chemical
data suggest a greater extent of recent petroleum pollution at mp 0 than in the
sample at mp 41, yet the apparent C-14 age is much older in the upstream
sample, the opposite of that which would be predicted if recent petroleum
pollution were dominant in the C-14 data. Both of the samples we analyzed have
higher weight fractions of hydrocarbons than the surface sediments of Lake
Washington, and thus are definitely contaminated with fossil fuel carbon.
However, in the harbor sediments, despite the highest level of hydrocarbon
contamination in our samples, the effect on bulk organic carbon C-14 dates is
overwhelmed by recent sewage organic particles.
The C-14 data on fossil carbonate shells reported in Table 15 present a
substantial geochemical problem in interpretation. The apparent ages, except
for the three surface samples of oyster fragments containing bomb C-14, are
all between '^ 1000 and '^ 3000 radiocarbon years. Apparent ages of this
magnitude have been found for modern carbonate shells in fresh waters (Broecker
and Walton, 1959; Keith and Anderson, 1963; Rubin and Taylor, 1963). We
believe the shell layers in Hudson sediments that we analyzed for C-14 to have
been deposited prior to the last 50-100 years, and were probably formed up to
several thousand years ago. We will present several qualitative arguments in
support of the above statement, and one semi-quantitative approach to
estimating the true time of growth of the estuarine carbonate shells we
analyzed.
During a recent study of trace metal concentrations (Zn, Cu and Pb) in
Hudson sediments, we have been able to establish good estimates of pre-
industrial heavy metals in this estuary. Samples from the longest core
reported here (F) have "base-line" heavy metal concentrations below the upper
20 cm. All of the subsurface C-14 data on this core are found between
laminated layers of fine-grained sediment with no recent metal contamination.
These data are the strongest evidence we have that if reworking of the
carbonate layers has occurred, it predated the period of heavy metal contam-
ination which presumably has occurred over at least the last 50 years.
Modern fresh water carbonate shells with apparent C-14 ages of several
thousand years (Broecker and Walton, 1959) generally come from carbonate rock
drainage basins, the water of which usually have bicarbonate concentrations on
the order of 2 mM or more. The Hudson drainage basin is not, however, predom-
inately carbonates or other sedimentary rocks. The Upper Hudson (bicarbonate
£0.3 mM) drains crystalline metamorphic rocks of the Adirondack mountains,
and the mean bicarbonate of the Hudson below the confluence with the Mohawk is
about 1 mM. Thus, qualitatively one would expect considerably less depletion
80
-------
of C-14 in Hudson inorganic carbon than that equivalent to an apparent age of a
few thousand years. We can make a somewhat more quantitative attempt to esti-
mate the original C-14/C-12 ratio of shells formed in the Hudson Estuary based
on our one measurement of carbonate shells formed in the Hudson River at a well-
known date prior to the arrival of bomb C-14. If we assume the shells collected
in 1887 from mp 98 (Table 15) have a C-14/C-12 ratio (6C-14 = -76°/oo,
apparent age 630 years) which is representative of the fresh water carbonates
formed in the pre-nuclear era Hudson, then after correcting for the known age,
an upper limit of 540 years (6C-14 = -67°/oo) can be placed on the extent of
age correction which should be applied to carbonate shell material formed in
the saline reach of the Hudson.
We can also attempt to use stable carbon isotope data for the estuarine
carbonates to estimate the original 5C-14 of these samples. Fresh water
carbonate shells from the Hudson prior to significant sewage pollution
probably had a 6C-13 value of -11°/00 (Table 15 - assuming the one sample we
analyzed from 1887 is typical), whereas carbonate shells formed from surface
sea water have 6C-14 values of ^ 0°/00. Pre-nuclear era 6C-14 in marine
carbonates was ^ 0°/00 (Broecker and Olson, 1959; Broecker and Olson, 1961).
If 6C-14 of shells formed in the saline reach of the pre-nuclear Hudson are
assumed to lie on a mixing line between a marine carbonate value of 0°/00
and a fresh water carbonate endmember of -67°/00 (age corrected value of 1887
Hudson shells), and 6C-13 of carbonate shells also lie on a mixing line
between marine carbonates (0°/00) and fresh water shells from the Hudson
(-ll°/0o)j then SC-13 of a carbonate shell layer can be used to estimate the
initial 5C-14 of that shell layer (see Figure 30). We have applied age
corrections by this procedure to all of the dates reported in Table 15. The
maximum correction was -690 years, with most of the changes being on the
order of -190 years. In the absence of a measured 6C-13 value, we estimate
the original 6C-14 of a sample to be equivalent to about -190 years based on
the average 6C-13 of similar samples. The uncertainty of the correction for
these samples is obviously greater than for the samples for which measured
SC-13 values were obtained. Dates corrected for initial 6C-14 are plotted in
Figures 31 and 32, with the uncorrected dates in radiocarbon years also
included to indicate the magnitude of the correction terms. No major changes
in estimates of sedimentation rate result from using corrected dates rather
than the original dates, except for core B (Figure 31) where the effect is
to reduce the estimated sedimentation rate from about 2 mm/yr to about 1 mm/yr.
The procedure we have suggested for estimating original 6C-14 of
estuarine carbonates is obviously oversimplified in that it ignores a number of
processes which could be significant, but it does provide a relatively simple
conceptual framework. One weakness to our approach is the assumption that the
one sample of pre-bomb fresh water carbonates in the Hudson we analyzed is
representative. We cannot support that assumption much beyond the statement
that one sample is a whole lot better than none. The magnitude of suggested
corrections in original 6C-14 would not be greatly altered by choosing another
tie point on the 6C-13 - 6C-14 graph, provided 5C-13 and 6C-14 were varied
roughly parallel to the line drawn. Gas exchange would be the most likely
process to cause variation in 6C-14 of the fresh water endmember, and as can
be seen from the dotted line on Figure 30, the computed effect of gas exchange
on 6C-13 and 5C-14 is similar but not identical to mixing of fresh water with
marine bicarbonate. The magnitude of suggested corrections in original 6C-14
81
-------
Apparent Initial Age (SC14)
from 8C13
40-
20-
t
o -OH
-^~
E
o
-20-
-40-
-60-
-8Q-*-j
Mixing line for
complete exchange
of atmospheric C14
(Pre-bomb) with.
estuarine bicarbonate/
Two end-
member mixing
+ gas exchange
(Pre-bomb)
Marine
Carbonates
(Pre-bomb)
River Shells
(Pre-bomb)
T I I I I | T I
-12 -10 -8 -6 -4 -2 0 +2 +4
Figure 30. Measured 6C in Hudson Estuary shells can be used to estimate
the original 6C1 value of the estuarine shells, assuming
mixing of only two end members. Gas exchange does not cause a
major deviation (dashed line) from this mixing line for the
Hudson prior to bomb testing.
82
-------
00
OJ
0
CL
(1)
Q
I
200-
300
Subsurface Shell Dates
mp 22
.3mm/yr/7 . \
\ 2.2mm/yr
\
\
\ \
\ \
1000 2000
Age (years)—*
3000
o
100-
CL
Q
200-
300-
Subsurface Shell Dales
mp 18.5
-oH
I X—cH
1000 2000
Age (years) -—
3000
Figure 31. Data points indicated by circles are ages derived from measured radiocarbon values.
Data points with error bars were corrected for estimated initial 6C14 values using 6C
(Figure 30). The most probable ages for the shell layers lie between the two plotted
ages at each depth.
Figure 32. Symbols have the same meaning as in Figure 31. The core extends another 3 meters below
the deepest shell layer.
-------
of the samples in Table 15 is not large compared to the measured 5C-14. Thus,
changes in the details of Figure 30 would not greatly alter the magnitude of
the "corrected" ages.
Another weakness in the approach we have taken in estimating initial 6C-14
on the basis of 6C-13, is the lack of detailed knowledge of the variation of
6C-13 in the Hudson. We have no way to obtain such data for the "pre-
pollution" era of interest. Three transects of 6C-13 and salinity have been
reported for the Hudson in 1963-1964 (Sackett and Moore, 1966). All three of
these sets of data have different 6C-13 - salinity slopes, but appear to
indicate conservative mixing between two endmembers within each data set
(Mook and Vogel, 1968). Only one of the salinity transects approached fresh
water, and that data set, collected in April 1964, extrapolates to 6C-13 =
-9°/00 in fresh water, which is certainly compatible with the value of 6C-13 =
-ll°/oo which we used for a location considerably farther up the estuary in
the construction of Figure 30.
Another possible perturbation of 6C-13 in the Hudson is production of
methane and carbon dioxide of greatly different 6C-13 than the values we have
used for Figure 30. A recent study of methane geochemistry in the Hudson
(Hammond and Simpson, 1977) indicates that methane supplied to the water column
is derived primarily from methane bubbles produced in the sediments which
partially dissolve (^ 5-10%) on the way up through the water column. Methane
oxidation does not occur in the water column to any appreciable extent, and
dissolved methane is lost primarily by gas exchange. The total production rate
of methane which forms bubbles and passes through the water column is ^ 1.5
mole/m /yr. This is an order of magnitude less than our estimate of the gas
exchange rate of carbon dioxide in the Hudson. CH, is not very likely to
significantly perturb inorganic carbon 6C-13 since about 90% of the bubble
flux passes through the water column without dissolving. Assuming a 6C-13 =
-60°/00 for CH, in the water column and that 1% of the total bubble flux is
oxidized (we believe this is a fairly strong upper limit) the effect on the
inorganic 6C-13 value would be ^ -0.5°/0o.
CONCLUSIONS
The dissolved inorganic carbon in the present day (1976) Hudson River and
estuary contains appreciable bomb C-14, but is not in equilibrium with
atmospheric C-14/C-12 ratios. The downstream increase in 6C-14 in the water
is compatible with reasonable values of gas exchange rate and physical trans-
port of the water, and thus could conceivably be used to compute gas exchange
rates for the Hudson. There appears to be at least one additional mechanism,
oxidation of sewage organics, which can significantly increase the 6C-14 of
the inorganic carbon in the Hudson estuary near New York City.
Surface sediments in New York harbor, despite measurable contamination
with petroleum hydrocarbons, have a large component of organic carbon with
appreciable bomb C-14, presumably mostly recent sewage particles since primary
production of organic carbon in the lower Hudson is an insignificant fraction
of the discharge rate of sewage organics. This recent sewage contamination
dominates the C-14/C-12 ratio in bulk organic matter in New York harbor
sediments, resulting in apparent future C-14 ages of that material as opposed
to apparent C-14 ages of a few thousand years in surface sediments in the
84
-------
Hudson upstream of New York harbor. The mean accumulation rate of sediments
in the Hudson since the onset of estuarine condition ^ 12,000 B.P. has been
about 3 + 2 mm/yr, based on a total accumulation of 10-60 meters. Such a rate
of sediment accumulation is approximately the rate of subsidence of the New
York area during the last century (Fairbridge and Newman, 1968) and is also
compatible with the early Holocene eustatic submergence (0.5 mm/yr). Cores
collected in or adjacent to the present main navigation channel of the Hudson
between mp 18 and mp 24 penetrate up to several thousand years of sedimeint
within a few meters, and sedimentation rates calculated from C-14 dating of
subsurface marine shell layers are in the range of 2 + 1 mm/yr. The age of
the tops of these cores, extrapolated from uncorrected C-14 dates of subsurface
shell layers, are in the range of 1000 to 2000 radiocarbon years, although
oyster and clam shell fragments at the tops of cores contain bomb C-14.
Our best estimate of the pre-bomb C-14 for inorganic carbon in the Hudson
estuary indicates a correction of ^ -200 years should be applied to C-14 dates
of estuarine shell materials from the sediments with ages of 1000-3000 years,
if no additional information on the initial 6C-14 of the sample is available.
A simple correction using 6C-13 based on two component mixing between surface
sea water and fresh Hudson River water can be made. The magnitude of this
correction in our sample was as high as ^ -700 years, but most commonly was on
the order of -200 years as suggested above, and thus represents a relatively
small change for shell materials with ages of 1000-3000 radiocarbon years.
Conceptually, a correction for the effect of gas exchange on SC-14 downstream
of the fresh water endmember composition should be included, but equilibration
with the pre-bomb atmosphere would have increased the values for both 6C-13
and SC-14 of inorganic carbon in the Hudson Estuary in a similar manner as an
increased contribution from the sea water endmember.
None of the corrected shell dates we obtained extrapolate to the present
at the sediment surface. We collected our cores from an area (mp 18 to mp 24)
for which there was no evidence available to us indicating that dredging had
occurred, but it is possible that some of the most recent sediment from the
core sites had been removed, either by dredging or current scouring, before
the recent oyster fragments containing bomb C-14 were accumulated. Another
possibility is that sediment accumulation at the core sites had not been
appreciable over the past 1000 years or so until quite recently. Sediment in
the harbor region (mp 0) is presently accumulating quite rapidly in a number
of places. Assuming the dredge spoils of 4 x 10 tons/yr (dry weight) were
spread uniformly over the bottom between mp 12 and mp 0, the rate of
accumulation would be ^ 25 cm/yr. We know from dredging records that one
large shoaling area has a sediment accumulation rate of at least this value
(Panuzio, 1965). We also know from the distribution of other anthropogenic
radionuclides (e.g., cesium-137) in harbor sediments that shoaling rates of
comparable magnitude (10-20 cm/yr) are found over considerable areas
(Simpson et al. , 1976). Thus sediment currently accumulating over substantial
areas of NewTork harbor is being deposited at a rate of 10-100 times faster
than the mean for the past 12,000 years, and could be expected to bury
significant amounts of sewage organic carbon before oxidation could occur
(Figure 33).
85
-------
\ r
SEDIMENTATION RATES IN THE HUDSON
Coves
I0-30mm/yr
Navigation
Channel
< 1-2 mm/yr
Shoal
100-300 mm/yr.
Long Island
Sound
Harbor
50-lOOmm/yr
4I°00
Subtidal
Channel Bank
1-5 mm/yr
40°30
Figure 33. Sedimentation rates in the Hudson from the measured distribution
of 137cs in the sediments.
86
-------
SECTION 8
A NEW ENZYMATIC METHOD FOR ANALYSIS OF CELLULOSE IN SEDIMENTS
Sediments in natural systems such as estuaries, lakes and the ocean
consist of a complicated mixture of materials derived from biological processes,
as well as from chemical and physical processes which could be classified as
"geological" in origin. The organic fraction (non-calcium carbonate) of
sediments ranges from less than a tenth of a percent in slowly accumulating
sediments of large areas of the deep ocean up to more than 30% in some lakes.
The history of production, preservation and destruction of sedimentary organic
matter is strongly dependent on the type of organic compounds involved, as
well as on environmental factors. For example, within one group of compounds
(the carbohydrates) glucose, a simple sugar, and starch, a polymer consisting
of a a-linked glucose monomers, are rapidly cycled by bacterial communities,
while cellulose, a polymer consisting of repeating 3-linked glucose units,
is much less reactive in sediments.
Carbohydrates have been measured in environmental samples at many
localities (Lewis and Rakestraw, 1955; Rogers, 1965; Biggs and Wetzel, 1968;
Artem'yev, 1970), using relatively simple chemical procedures which have
evolved over the past several decades (Dreywood, 1946; Morris, 1948; Dubois
et_ al_. , 1956; Zein-Eldin and May, 1958; Handa, 1966; Strickland and Parsons,
1968; and Gerchakov and Hatcher, 1972). These are based on forming colored
solutions by the reaction of simple sugars or polysaccharides with anthrone
or phenol or similar reagents in sulfuric acid. The anthrone and phenol and
other similar procedures provide some measure of the total carbohydrate
content of a sample, including a substantial variety of both monomeric and
polymerized carbohydrates. Pre-treatment with acid and base leaches can be
used to remove water-soluble monomeric carbohydrates (Strickland and Parsons,
1968). Thus the same analytical procedure can be used to measure two
operationally-defined quantities, "total carbohydrates" and "crude fiber",
the latter including what is left after strong acid and base leaches.
The amount and type of compounds present in sedimentary organic matter can
be expected to vary substantially, depending upon the original source of the
organic matter and upon processes occurring after deposition in the sediment.
Our present knowledge of the details of source composition and post-
depositional changes of sedimentary organic matter is very primitive. This
is perhaps to be expected, considering the structural complexity and range of
compounds encompassed by environmental organic materials included under terms
such as "humic acids". One additional factor which tends to perpetuate the
present level of understanding is the collection of laboratory data with
analytical techniques which aggregate a large range of compounds into one
measurement. Thus a measurement of "total carbohydrates" provides a convenient
87
-------
indicator of the amount of one large class of organic molecules which could
make up anywhere from a few percent to more than one half the carbon content
of a sample of sediment, but it incorporates a tremendous range of compounds,
some of which are monomers with molecular weights on the order of a few
hundred, while others are highly polymerized, having molecular weight in the
millions.
We have developed a new analytical technique, based on the use of a
sequence of enzymes, to measure cellulose in sediments. This technique
introduces a new type of analytical approach to the study of sedimentary
organic matter, and provides a procedure for the measurement of one specific
type of highly polymerized carbohydrates which is -an important., relatively
stable component of most recent sediments. Use of enzymes in analytical
procedures is common in biochemical research and in routine biomedical
laboratory tests, but has not, to our knowledge, been previously exploited
in organic geochemistry.
One possible application of the enzymatic analytical method for cellulose
we have developed is for establishing the amount of sewage sludge in sediments.
Sewage sludge contains large amounts of cellulose (Hunter and Heukelekian,
1965), which potentially could be used as a very sensitive indicator of the
presence of such pollutants. Total carbohydrate measurements have already been
made in the area of the coastal ocean near New York City (Hatcher and Keister,
1976) in an attempt to map the present distribution of sewage sludge which has
been discharged for a number of decades into a dumping zone about 15 kilometers
from the mouth of the Hudson River estuary.
METHODS
Chemical methods for carbohydrate measurements in environmental samples
use strong acid solutions to break up polymers and produce colored complexes
of basically unknown structure, using reagents such as anthrone or phenol.
These methods are nonspecific and have varying sensitivities, depending upon
the type of sugar monomer. Interference from non-carbohydrate materials
occurs, and empirical corrections for these effects must be made.
Cellulose may be hydrolyzed enzymatically under mild conditions of
temperature and pH. Cellulytic enzymes cleave only the 3-D-(l->4) glucosidic
bonds linking the glucose units, and act only on cellulose and certain of its
derivatives. Enzymatic hydrolysis, therefore, lends itself to the selective
analysis of one type of compound, leaving other components of the system
unaltered.
Materials
- Extracting acid: 20 ml of concentrated H^SO, added to distilled water
and brought to a total volume of 1 liter.
- Extracting base: 25 g of NaOH added to distilled water and brought to
a total, volume of 1 liter.
- Acetate buffer: 0.05 M, pH 5.0.
-------
- Glucose stock solution: 1 g B-D-glucose (Fisher Chemical Company) was
dissolved in acetate buffer and diluted to make 1 liter of stock solution
(1 mg/ml).
- Glucostat Reagent Set (Worthington Biochemical Company).
- High activity cellulase, lyophilized preparation from Trichoderma
viride (obtained from Dr. E.T. Reese, U.S. Army Labs, Natick, Massachusetts,
Lot No. 9414). Prepared by dissolving 400 mg of enzyme in acetate buffer and
diluting with buffer to 200 ml.
- Redistilled chloroform.
- Avicel microcrystalline cellulose, TG 101 (FMC Corporation).
Pretreatment of Samples
Sediments were dried and ground with mortar and pestle. Approximately
800 mg of sediment was weighed out for each sample and transferred to 100 x
18 mm polypropylene centrifuge tubes. Extracting acid (10 ml) was added and
the tubes were thoroughly mixed by shaking and stirring, and placed in a
boiling water bath for 30 minutes, after which the tubes were centrifuged for
15 minutes at 15,000 RPM. Supernatant was removed by vacuum pipet, care being
taken not to dislodge any solids from the pellet. Extracting base (10 ml) was
then added, followed by mixing, heating, centrifuging and removal of super-
natant. Then distilled water (10 ml) was added, mixed, centrifuged, and the
supernatant removed. The rinsing procedure was repeated to remove soluble
residues from the sediment.
The residual sediment from the acid-base treatment was transferred to
an Erlenmeyer flask as follows: 5.0 ml of acetate buffer was drawn into a
Einnpipet; about half was delivered to the centrifuge tube, and the tube was
mixed on a tube buzzer or stirred, until all of the pellet had broken loose;
the contents were then rapidly poured into an Erlenmeyer flask. The remainder
of the 5.0 ml buffer was then used to wash out the residual sediment in the
same manner. This two-step procedure was found to effectively transfer all of
the sediment in a precise volume of buffer.
To each flask 80 yl of chloroform was added as a bacteriostat, and the
suspensions were ultra-sonically homogenized, sealed with parafilm and stored
overnight in a refrigerator.
Enzyme Procedures
For each sample of sediment to be analyzed, three 800 mg aliquots
of sediment were treated with acid and base. Sample A was used as a zero
reaction-time blank and samples B and C were incubated with cellulase for 24
hours. All reactions were carried out in 50 ml Erlenmeyer flasks, with a
total liquid reaction volume of 10 ml, consisting of 5 ml of sediment
suspension in acetate buffer and 5 ml of cellulase solution.
The enzyme reactions were begun on the day following initial acid-base
leaching by repipeting 5.0 ml of cellulase solution into flasks B and C of each
89
-------
sample. The flasks were then sealed with parafilm, a pinhole punched in each,
and placed in a gyrotory shaker bath set at 50°C.
When 24 hours had elapsed the contents of the three flasks were poured
into centrifuge tubes. To flask A of each sample, 5 ml of the cellulase
preparation used for flasks B and C was added, and the contents immediately
transferred to polypropylene tubes; all three samples were centrifuged 15
minutes at 15,000 RPM. It was essential to keep flask A cold until addition
of the enzyme, and to centrifuge immediately, before any enzymatic hydrolysis
could take place. From the supernatant of each tube, a 2 ml aliquot was
withdrawn, with a Finnpipet, and placed in 20 x 150 mm glass test tubes, for
glucose determination.
Glucose, the product of the hydrolosis of cellulose was then analyzed
enzymatically, using the glucostat macro method. One vial each of chromogen
and glucostat reagent were injected with distilled water, the contents
transferred to a graduated cylinder and diluted to 80 ml. Multiples of these
quantities were prepared depending on the number of samples to be analyzed.
The dissolved reagent was dispensed with a repipet: to each 2.0 ml sample
aliquot, 8.0 ml of glucostat reagent was added, and the tubes were agitated
with a tube buzzer. After exactly 10 minutes, 2 drops of 4N HC1 were added
to terminate the reaction and the tubes were mixed again on the tube buzzer.
It was found to be convenient to deliver the glucostat and later the HC1 to
one tube every 15 seconds.
After an interval of at least 5 minutes to allow full color development,
the absorbance of the solutions was measured in 1 cm cuvettes against
distilled water at a wavelength of 420 nm and a slit width of 1 mm. The
absorbance of sample A was subtracted from the mean of samples B and C to give
the corrected absorbance due only to glucose produced by the enzymatic
hydrolosis of cellulose.
Glucose Standards
The most suitable standardization procedure was found to be addition of
glucose to a reaction mixture exactly like the actual samples, containing
sediment, buffer, cellulase and chloroform. For each batch of samples to be
analyzed, an extra aliquot of a representative sample was weighed out and
acid-base treated exactly like the other samples. The residue was transferred
to an Erlenmeyer flask with 5 ml of glucose standard instead of acetate buffer.
This standard was prepared by diluting 20 ml of stock solution (1 mg/ml) to
100 ml with acetate buffer (0.2 mg/ml). After addition of 5 ml of cellulase,
the total reaction volume of 10 ml yielded a standard addition of 0.1 mg/ml
of glucose. The flask was incubated for 24 hours with the other samples and
then analyzed for glucose content. The absorbance of the same sample without
the glucose spike was subtracted from that of the standard; this difference
was set as the absorbance due to 0.1 mg/ml of glucose.
To determine the percent of cellulose in a sample the following formula
was used:
90
-------
inn abs. B + abs. C 0.1 me/ml ,A .,
100 x 7. abs. A x 7—- &H- ._,^. . x 10 ml
i (abs. std. - abs. (B+C) )
2
weight of sample (mg)
This number represents the weight percent of enzyme-hydrolyzed cellulose in
the sediment expressed in terms of glucose (see Discussion of Method).
In addition, a fresh glucose standard of 0.1 mg/ml was made up and
measured at the time of the glucostat analysis, and an acetate buffer blank
was run as well. The difference of these two absorbances was calculated, and
the two values were compared as a check for bacterial activity, contamination
or inhibition of glucostat enzymes.
Discussion of Method
A ce'llulase preparation of high specific activity was necessary to
achieve good results. Several commercial cellulase preparations were found
to have too low an activity or were found to cause significant background
absorbance at 420 nm. Worthington's cellulase II typically yielded only
30 yg/hr per mg of enzyme when assayed as recommended with Avicel micro-
crystalline cellulose as substrate. The only enzyme preparation which
proved suitable for quantitative analysis of sediment samples was a lyophilized
T. viride cellulase provided by Dr. E.T. Reese (U.S. Army Labs, Natick,
Massachusetts). When assayed under the same conditions as for commercial
cellulases the enzyme supplied by Dr. Reese (50°C, 1 hour), produced
1200 yg glucose/hr per mg of enzyme, 40 times the rate we found for commercial
cellulases. This preparation caused negligible interference at 420 nm,
thereby increasing the sensitivity at low levels and eliminating another source
of error. With the procedure described here, 30 mg of lyophilized enzyme was
required for each sample analysis including a blank and two duplicates.
The acid-base treatment derived from Strickland and Parsons (1968), was
found to be necessary for several reasons. Untreated Hudson River sediment
was found to inhibit the Natick enzyme. Inhibition of the cellulase was tested
by saturating with Avicel cellulose: to one group of tubes no sediment was
added; to a second and third group 400 mg of acid-base treated sediment and
400 mg of untreated sediment were added, respectively. The first and second
groups of tubes showed similar activity while the third was about 40% lower,
indicating inhibition of untreated sediment.
Treatment of the sediment by boiling acid and base, ultra-sonic dispersion
and overnight soaking prior to addition of enzyme undoubtedly increased the
rate of enzymatic hydrolysis by swelling the cellulose fibers. Reese and
Mandel (1963) report that the reactivity of cellulose fibers is enhanced by
increasing the surface available to the enzyme. The acid-base treatment
also served to remove soluble substances which might interfere with or inhibit
the glucostat measurement.
The glucostat reagent set (Worthington Biochemical Company) provided
a highly specific and accurate colorimetric method for measuring glucose
produced by cellulolytic activity. The set utilizes a coupled enzyme system
91
-------
1.4-
t
1.2-
1.0-
E
C
O
cvj 0.8 H
O)
o
c
o
.a
0.6-
04-
0.2-
0.1
0.2
0.3
mg/ml of Glucose
1
o
<\j
0)
o
c
D
\_
O
.a
0
Figure 35.
Standard curve of absorbance at 420 nm as a function of the
amount of glucose added to a sample.
8 12 16
Cellulose Reaction Time (hours)
x->
1
24
Time response of glucose production by the enzymatic hydrolysis
of cellulose in a sample of Hudson Estuary sediment (mile point 43)
92
-------
consisting of glucose-oxidase and peroxidase enzymes. Figure 34 shows a
standard curve obtained by diluting glucose stock with acetate buffer.
Chloroform was found not to inhibit the response of the glucostat system. It
was found, however, that turbidity or discoloration of the solution to be
analyzed caused a depression of values. Thus high speed centrifugation of the
reaction mixture resulting in a clear supernatant proved to be essential.
Sediment samples that were not acid-base treated often produced yellowish
supernatants which caused interference at 420 nm.
With each batch of samples separate glucose standards and standard
additions to sediment samples were run; the glucostat response of the standard
additions to samples were consistently only several percent lower, indicating
that no significant inhibition of the glucostat system was occurring in the
cellulase-sediment suspensions, and that no bacterial degradation of glucose
was taking place during the course of the reaction.
Figure 35 shows a time curve for the cellulytic hydrolysis of Hudson River
sediment relatively high in cellulose. Most of the hydrolysis occurs within
the first few hours, and then the rate of hydrolysis falls off rapidly. Assay
of reaction mixtures after 24 hours by saturation with 200 mg of Avicel
substrate showed that considerable cellulase activity remained, .so that
cessation of cellulose hydrolysis in the samples was not due to enzyme
denaturation. Twenty-four hours was chosen as a standard reaction time for
analysis of sediments because it is far enough into the flat part of the curve
to minimize the effect of differences in the initial rate of hydrolysis.
Susceptibility of cellulose to enzymatic attack varies with great many factors
and the relatively long reaction time allows less susceptible forms of
cellulose to be fully hydrolyzed.
Cellulose Standards
Addition of known amounts of pure cellulose (Avicel) to sediment samples
prior to any treatment other than drying yielded only about 70% of the glucose
theoretically available. However, the amount of glucose produced was directly
proportional to the amount of cellulose added. A standard curve was prepared
in the following manner: 5, 10, 20, 30, 40, 1000 and 2000 mg of Avicel were
weighed out and transferred to 150 ml beakers. After 100 ml of distilled
water was added to each beaker, and while the suspensions were ultrasonically
homogenized, 1 ml aliquots were withdrawn using a Finnpipet, and added to
identical 400 mg weights of Hudson River sediment in centrifuge tubes. These
samples were then acid-base treated and incubated with cellulase enzyme exactly
as described above. Figures 36 and 37 show a standard curve prepared by
plotting absorbance produced against weight of cellulose added. On the basis
of molecular weights 162 grams of cellulose should produce approximately 180
grams of glucose when hydrolyzed completely due to the addition of a molecule
of water at each glycosidic linkage. We found the quantity of glucose produced
to be consistently about 70% of the theoretical value, whatever weight of
Avicel was added. The pretreatment of Avicel with acid and base, as was also
done for samples, results in some loss of weight by the Avicel. We subjected
4 different aliquots of Avicel, ranging in weight from * 10 mg to -v 80 mg, to
pretreatment with acid and base and then reweighed the Avicel after drying.
The average weight loss for the 4 samples was 36%. Correcting for this weight
loss from acid and base pretreatment, the yield of glucose from Avicel was very
93
-------
E
c
O
OJ
O
C
O
O
(f>
.0
0
100
200
400
Figure 36.
300
fj.g of Cellulose Added - >-
Standard curve of absorbance at 420 nm as a function of the
amount of cellulose added to a sample of Hudson Estuary sediment
(mile point 43 - 400 ing) .
0.7.
0.6-
E
c
O
20.5
0.4-
CJ
c
D
.0
k-
o
.0
0.3-
0.2-
O.M
0
2000
Figure 37.
500 1000 1500
/LLQ of Cellulose Added >-
Normal working range for samples reported here (expansion of
low absorbance portion of Figure 36).
94
-------
close to that predicted from complete reaction with cellulase. We have no
independent way to estimate the weight loss of cellulose in sediments resulting
from pretreatment with acid and base, which obviously introduces some degree
of uncertainty. Because of this uncertainty, we chose to report the weight of
original sediment cellulose as equal to that of the glucose, omitting any
correction factors based on molecular weights or the loss of cellulose by
pretreatment which could vary in proportion between the Avicel standard and
the cellulose in environmental samples. The maximum sediment weight that could
be effectively analyzed with our procedures was found to be 800 mg. The amount
of cellulose detected was found to be proportional to the weight of sediment
analyzed, up to a sample weight of 800 mg. Larger weights of sediment were
difficult to transfer quantitatively, and after centrifuging yielded super-
natants with some discoloration and turbidity.
The analysis for cellulose proved extremely sensitive and reproducible
in low ranges. In any given batch of samples the reaction blanks gave a
glucostat absorbance of + 0.002. Assuming a minimum sensitivity of 0.003
absorbance units, the minimum detectable amount of cellulose in an 800 mg
sample was 0.001%.
RESULTS AND DISCUSSION
We have analyzed 14 samples of sediment from the Hudson River estuary and
New York Bight and one sample of sewage sludge from Ward's Island sewage treat-
ment plant in New York City for cellulose (Table 16). The locations of these
samples range from near the upstream limit of saline water intrusion in the
Hudson Estuary to beyond the edge of the continental shelf at depths of ^ 2000
meters in the Hudson Submarine Canyon (Figure 38). Except for the sewage
sludge sample with a cellulose content of ^ 7%, all of the sediment samples
had less than 0.06% cellulose by weight.
The variations in grain size of the Hudson estuary samples was
relatively small, but in the New York Bight the samples ranged from
predominantly clay and silt (< 63 y) to predominantly sand (> 63 y) . We
normalized the cellulose data by two procedures, one on the basis of the
fraction of fine particles, and the other on the basis of the weight loss of
the sediment when heated between 100°C and 500°C. This weight loss has
been found to be approximately twice the organic carbon content of Hudson
sediments (Gross, 1972), and thus provides an indicator of the fraction of
organic matter which is cellulose. Assuming the organic carbon fraction is
approximately half of the weight loss on ignition, all of the samples of
sediment organic matter consist of less than 1% cellulose by weight, while
sewage sludge organic carbon approaches 20% cellulose.
We analyzed several samples which are not recent sediments from cores
in the Hudson estuary. One sample (mile point 19, 550 cm) is probably greater
than 1000 years old, based on radiocarbon dating of shell layers higher up in
the core. This sediment sample contained no measurable cellulose (< .001%
by weight), although the total organic content was still appreciable (^ 2%).
Another old sample from a core in New York harbor (mile point 0, 65-70 cm) had
a very low cellulose value (^- 0.001%). This sample has a trace metal composi-
tion indicative of pre-industrial levels (unpublished data), but we do not have
good age control on this sample other than that it is probably at least 50-100
years old.
95
-------
Table 16
Cellulose Content of Hudson R Lvur Ks tuary and New York Bighc Sediments
Depth in
Location Sediment
(Mile Point) (cm)
54
54
43
19
0
0
0
Sewage
Sludge
-26
-38
-67
-117
-136
-147
-157
20-25
50-55
0-10
550
0-10
5-10
65-70
0-10
0-10
0-10
0-10
0-10
0-10
0-10
Weight 7.
93
97
95
90
90
95
95
100f
100s
33
17
5
37
98
93
Weightb%
lost
on ignition
(LIC)
6
6
9.
4,
a.
9.
5.
70
12.
5.
2.
T
9.
10.
10.
.4
.2
.5
,1
.0
2
.3
7
0
.8
.3
7
.&
.9
Weight 7.
Weight % total
Weight %C Crude Fiber Carbohydrates
cellulose by phenol by phenol
.008
.008
.035 .08 1.5
.000
.028
.013
.001
7.3 9 20
.056
.006 .05 .86
.001 .02 .27
.001 .007 .09
.000 .04 .83
.004 .04 .72
.000 .03 .52
Cellulose
'/, < 63u
x 105
8
8
37
0
31
19
1
7300
56
18
•v. Oh
•>. oh
0
4
0
Cellulose
LIG
x 104
13
13
37
0
35
20
2
1050
44
12
* o»
•*. oh
0
4
0
Locations are given relative Co the Hudson River mileage reference system, with the origin located at the southern
tip of Manhattan. Positive numbers are upstream, and negative numbers downstream of the origin.
Weight loss on ignition (LIG) represents the weight loss of samples heated from 110 °C to 500°C. In Hudson sediments,
this weight loss is usually about two times the organic carbon content.
Data given in terms of the weight of glucose produced by enzymatic hydrolosis.
"Chemical analysis by method of Gerchakov and Hatcher (1972) jfter leaching with strong acid and base.
Chemical analysis by method of Gerchakov and Hatcher (1972).
No size fraction analysis was made because of the fibrous nature of dried sewage sludge. The weight < 63p was assumed
Co be 100% for the purpose of the calculation in the next to last column,
SThe sediment from this site was more than 90% sand (> 63u). To increase sensitivity for cellulose determination, we
removed all of the sand fraction, which had a weight loss on ignition of ^ 1%, and thus had a relatively insignificant
proportion of organic matter.
The cellulose content of these samples was very hear the sensivitity limit of our analytical technique, so we have not
reported ratios involving those concentrations.
96
-------
Hudson River
mP43 i Estuary
'30
74°30
39°00
30'
Figure 38. Locations of samples reported in Table 16. Samples range from
near the upstream limit of saline intrusion (mp 54) to beyond
the edge of the continental shelf (mp -157).
97
-------
Recent sediments in the Hudson, both in the harbor (mile point 0,
0-10 cm, 5-10 cm) and upstream (mile point 43, 0-10 cm), have relatively high
cellulose values, as does the fine fraction of sediment from near the sewage
sludge dumping area in the New York Bight (mile point -26, 0-10 cm). Thus,
recent sediments in the harbor and in the dumping area, both of which
obviously contain appreciable sewage contamination, have high cellulose
values (.02 - .06%). However, recent Hudson sediments (mile point 43) well
upstream of the major sewage loading also are relatively rich in cellulose
(^ .04%). We have analyzed organic matter from both the harbor and
upstream for radiocarbon and found the harbor sediments to contain bomb
carbon-14 in the organic matter (from sewage) while upstream the organic matter
in surface sediments has a much lower carbon-14 content and thus appears to
have a substantially lower component of recent sewage organics. We, therefore,
conclude that high cellulose levels in Hudson sediments are not unequivocal
indicators of sewage contamination. Presumably the same is true for the New
York Bight, as is discussed in more detail below.
The magnitude of the cellulose values we observed indicates that total
carbohydrate measurements [ranging from 10% to 60% of the organic carbon
in New York Bight sediments (Hatcher and Keister, 1976)] provide very
little information about the distribution of cellulose in sediments, and
should not be interpreted as indicative of sewage cellulose.
It is also evident that there is little correspondence between cellulose
and total carbohydrates or crude fiber. Cellulose concentrations are more
than an order of magnitude smaller than crude fiber values in Bight sediments,
while they are comparable to crude fiber and total carbohydrate numbers
in sewage sludge and to crude fiber in one sample of Hudson sediment.
The lack of cellulose in old Hudson sediments, compared with the high
values of recent sediments, indicates that cellulose is probably not
stable in sediments on the time scale of hundreds to thousands of years,
in contrast to condensed organic matter, such as humic acids which apparently
can remain in soils and sediments for very long periods. The gradual break-
down of cellulose in sediments over long periods is reasonable considering that
some soil fungi and certain wood eating organisms contain cellulases.
The time scale of breakdown of cellulose in coastal sediments
would be very useful to know. We cannot place adequate constraints on the
rate at which this process occurs from the data reported here, but a rough
calculation indicates the half time may be at least 5 years.
If sewage sludge (SS) and dredge spoils (DS) were assumed to be present
in the same weight proportion as that originally dumped (weight ratio of solids
SS/DS ^ 0.06; Gross, 1972) and contained the original cellulose values reported
here (SS = 7% and DS = 0.02%), then composite dumped material should have
^ 0.44% cellulose. Assuming a mean age of ^ 15 years for dredge spoils and
sewage sludge discharged to the dump site (dumping of these materials in the
New York Bight has occurred over a number of decades) and also assuming that
our one measured value (0.056%) is typical of composite waste discharges, the
half time for destruction of cellulose would be ^ 5 years. Our sample was
from the fringe of the dumping area and was mostly coarse sand (> 90% by
weight). We removed the sand fraction prior to analysis for cellulose, and
98
-------
considered only the composition of the fine particles as relevant to this
discussion.
Actually it is likely that substantially more of the sewage sludge
particles than the dredge spoils would be carried well away from the
discharge zone. Assuming cellulose to be stable, and that virtually all of
the cellulose in our sample came from sewage sludge (ratio of cellulose in SS
to DS = 350) the value of ^ 0.06% in our sample is a little more than 10% of
that expected from the composite of sewage sludge and dredge spoils dumped
O 0.44% cellulose). Thus the two calculations indicate that if the sample we
report is representative, (1) cellulose has a half time for breakdown
of ^ 5 years, or (2) assuming cellulose is stable, only 10% of the sewage
sludge discharged is in the sediments. The most likely situation is
between these extremes, i.e., cellulose has a half time for destruction
of >_ 5 years and only some .fraction of sewage sludge (>_ 10%) is found in the
dumping' area sediments.
The range of cellulose values we observed in the Hudson and New York
Bight indicates that detailed studies of this component of the organic
matter in sediments could provide valuable insights into the accumulation
and diagenetic history of organic matter in recent sediments. Another
possible application of enzymatic cellulose measurements is in the study of
blooms of coastal planktonic organisms, such as Ceratium tripos, which may
possibly play an important role in episodes such as the recent anoxic condition
in the nearshore waters of the New York Bight (T. Malone, personal communica-
tion) . These organisms have cell walls containing cellulose and appear to be
capable of producing intense blooms under certain conditions. The cellulose
from blooms would probably be relatively stable compared with most non-
cellulose organic matter produced by coastal planktonic organisms, and could
thus be traced for significant time periods in both the water column and the
sediments.
99
-------
SECTION 9
RADON-222 AS AN INDICATOR OF TRANSPORT RATES FROM THE SEDIMENTS
TO THE WATER COLUMN IN THE HUDSON
INTRODUCTION
Striking differences in chemistry often exist across the sediment-water
interface. Lake, river, and ocean waters are usually oxygenated and relatively
depleted in nutrients. Interstitial waters of estuarine and nearshore
sediments are usually anoxic very near the sediment surface and are often
enriched in nutrients and in some transition metals. Thus, sediments consume
oxygen from the water column and return nutrients and perhaps some transition
metals. The movement of dissolved substances across the sediment-water
interface can play an important role in establishing the composition of a
water body and in regulating the rates of reactions which occur in sediments.
Three approaches have been employed to estimate the rate of exchange of
dissolved substances across the sediment-water interface.
1. Direct measurement of flux into overlying water in either cores
returned to the laboratory or in a confining device placed over
the sediment in situ. Chemical flux is determined by a time
series of analyses (Fanning and Pilson, 1974; Pamatmat and Banse,
1969).
2. Measurement of gradients in interstial waters and calculation of
diffusive transport (Berner, 1971 and others).
3. Construction of a mass balance for a substance in the water column
which includes transport into or out of sediments.
The first approach can be criticized as it alters the turbulent regime in
the vicinity of the interface. The second approach is complicated for
biologically active substances because the effects of reaction and variable
diffusion on concentration vs. depth profiles are difficult to separate and
because many of the critical processes may occur within a few millimeters of
the sediment surface. Turbation by organisms and bottom currents can further
complicate this approach. The third of these approaches avoids these
difficulties, but averages flux temporally, spatially and is not always
possible to employ.
Radon (Rn-222) is a noble gas with a four day half-life. Its parent,
radium-226 (1600 year half-life) may be dissolved in the water column, but
is primarily bound to or within solid phases. Thus high concentrations of
100
-------
radon are found in sediment pore waters or in groundwaters, and much lower
concentrations are found in the water column. In the Hudson, pore water has
^ 600 dpm/1 and overlying water ^ 1.4 dpm/1. If Hudson water were in
equilibrium with the atmosphere, the concentration would be ^ 0.03 dpm/1 at
25°C.
Since radon was originally proposed as a tracer for processes in aquatic
systems by Broecker (1965), it has been used to study mixing rates and gas
exchange rates in lakes and in the ocean. Figure 39 is a schematic radon
profile in the open ocean. Throughout most of the water column, radon is in
secular equilibrium with its parent. In the surface ocean, a deficiency of
radon exists due to a net evasion to the atmosphere. In the deep ocean, an
excess is found due to input from the sediments. The shapes of these
deficiencies and excesses depends on the characteristics of the mixing
processes in the ocean.
In the Hudson estuary the surface and bottom zones overlap. Figure 40
is a radon profile collected in 1971 at mp 25. Radon activity is clearly in
excess of Ra-226 activity indicating that the rate of input from sediments is
greater than the rate of evasion to the atmosphere.
Although first proposed more than ten years ago for studies of exchange
across the sediment-water interface (Broecker, 1965), little has been done
since then to exploit radon as a tracer for such processes. As a noble gas,
radon is free of biological complications so its distribution will be
controlled by a balance among production from its parent, radioactive decay
and physical transport processes. Some fraction of the radon which is
produced by decay of radium in sedimentary phases emanates into interstitial
waters. It may decay there, or migrate into overlying waters before decaying.
Migration can be effected by molecular diffusion or by turbation of sediments
by currents and organisms. We will present results of radon measurements
made in the Hudson estuary (Figure 40A) between 1971 and 1975 in an effort to
quantify the importance of processes other than molecular diffusion as
mechanisms for transport across the sediment-water interface in this estuary.
Preliminary results of this research have been discussed by Hammond et al.,
(1975).
ANALYTICAL METHODS
The technique for extraction and counting of Rn-222 from water samples
has previously been described (Broecker, 1965). After radon -analysis some
samples were stored to permit ingrowth of radon and a second extraction for
radium-226 analysis. Water samples (usually about 20 liters) were generally
collected at least 200 m from shore to ensure that samples were representative
of the major portion of the estuary. A few samples collected near shore in
very shallow portions (< 1 m) of the estuary showed concentrations which
sometimes approached twice those in the center of the channel.
Sediment samples were collected by gravity cores in plastic liners of
either 3.5 or 6 cm diameter. To process each sample, the sediment was
extrudedj sections were cut and quickly dropped into a glass kettle which
contained 100-200 cm of distilled water. The kettle cover was sealed to
101
-------
Deep Ocean Radon
Meters
Below
Surf 50
100
100
Meters
Abov€50
Bottom
Figure 39.
Surface Rn
Depletion
Gas Exchange Rates
Bottom Rn
Excess
> 3000 meter separation
—;> Near Bottom
Mixing Modes
Rn222/Ra"6 •
Deep ocean schematic radon profile: Through most of the water
column radon-222 is in equilibrium with its dissolved parent
radium-226. Near the surface a deficiency exists due to evasion
to the atmosphere and near the bottom an excess exists due to
diffusion from the sediment pore waters.
Hudson River Ra 226 -Rn222
Ideal Profile (Sept. 27, 1971) Piermont
depth 4
(meters)
8
12
i i -i "i 1 1- i i i i i
ox' 3 5 7 9 o
\ Salinity (%„) A
V x -— n
"h ~~~~~--. rV
226 222-
Ra Rn
i i i i i i < i i i i
10 30 50 70 90 110
iSurf ' DeeP. i — Bse£_. Hnm /inO litprc.
Figure 40.
Range of Ocean Ra
Hudson Estuary radon profile: Activity of radon-222 always
greatly exceeds that of dissolved radium-226.
102
-------
MAP OF
LOWER HUDSON
ESTUARY
HUDSON
DRAINAGE
BASIN
HUDSON
HIGHLANDS
(mean depth = !2.8m)
TAPPAN ZEE
.)(mean depth=5.3m)
LONG
ISLAND
SOUND H
East R.
MANHATTAN
UPPER BAY
(N.Y. Harbor)
Raritan R.
N.Y. BIGHT
LOWER BAY
LO 20 40
KILOMETERS
111111111 i
f.
1 "-"• "" •>•'•'
STATUTE MILES
- 4I°30
4I°00'
- 40°30'
74° 30'
74°00'
73°30'
Figure 40a. A diagram of the lower Hudson Estuary indicating mean depths of
the Tappan Zee and Highlands regions.
103
-------
the base by an 0-ring and contained 2 teflon stopcocks with 0-ring seats.
One of these stopcocks was connected to a bubbling tube. The kettle was
shaken to create a slurry which could then be extracted and counted in the
same manner as a water sample.
Radon extraction from sediments was .accomplished within 2 days of
collection and the kettle was resealed to allow regrowth of radon and measure-
ment of the effective parent activity (Ceq). Results of sediment analyses
are listed in Table 17. The rapid transfer of the kettles was apparently
fairly efficient in preventing gas loss as the activity ratio of Rn-222 to
Ra-226 was generally close to 1, with the exception of a sample from mp 53
which was apparently saturated in situ with CH, and formed large gas pockets
while still in the core liner. One puzzling observation is that all sediments
collected in the large diameter liner showed '^ 20% radon deficiency while
those from the small diameter liner showed none. Since gas escape should
occur primarily from exposed sediment surfaces, one would have expected a
larger diameter core to have less radon loss during sampling and transfer
operations, not more. After the Ra-226 measurement was completed, the volume
of the slurry was measured to compute the original volume of sediment sampled.
The slurry was then dried and weighed.
RESULTS
The Hudson estuary is classified as a partially-mixed estuary whose
circulation has been discussed by a number of authors, most recently by Abood
(1974) and Simpson and Hammond (1977). The estuary is tidal as far upstream
as a dam (at Green Island) located 154 miles from the southern tip of Manhattan
(mp 154) with stage changes of 1-2 m and maximum current velocities of 50-100
cm/sec . The estuary morphology is controlled significantly by the regional
geology, recently summarized by Sanders (1974). The channel is narrow and
quite deep where it passes through the Precambrian crystalline rocks of the
Hudson Highlands, and then broad and shallow immediately south of the Highlands
in the Tappan Zee region, an area formerly occupied by a large glacial lake.
South of the Tappan Zee region, the Hudson is bounded on the west by the
Palisades diabase, a thick sill formed in the Triassic (^ 195 million years
ago).
Initially, we expected to see a significant gradient in radon across the
sharp depth discontinuity separating the deep waters of the Highlands from
the shallow waters of the Tappan Zee. This was based on the assumption that
a uniform radon flux from sediments into water of different mean depths would
result in smaller concentrations in the deeper water. However, no discontin-
uity was ever observed, due to a number of processes discussed below.
Several surveys of radon concentration in the water column were made
along the axis of the estuary. The first of these, made in late July 1972
(Figure 41), showed no consistent gradients, either with depth or with location
along the channel of the estuary. Note that maxima may occur at the surface,
the bottom, or at intermediate depth. To test the time variability, several
profiles were taken at two locations throughout the summer of 1972 (Figure 42).
Another series (primarily surface samples) was collected at more locations and
shorter time intervals during August 1974 (Figure 43), and compared with
104
-------
TABLE 17
Activity of Mobile Radon in Sediment
226
mp
76
53
49
41
41
37W
34
27
25
25W
18
18
18W
13W
9
7
1
-5
Water
Depth
(m)
16
45
28
24
15
9
10
5
5
13
1
8
15
15
15
14
Coll.
Date
Aug 06 1974
Mar 02 1974
Aug 06 1974
Apr 06 1974
May 18 1974
Aug 08 1974
Apr 06 1974
Nov 14 1973
Mar 02 1974
May 18 1974
Nov 05 1973
May 19 1974
Aug 27 1974
Jan 18 1974
Jan 18 1974
May 18 1974
Mar 02 1974
Mar 02 1974
Interval
(cm)
2-10
10-18
2-10
10-18
2-6
2-10
10-18
2-6
2-6
2-10
10-18
0-4
9-19
19-29
39-49
49-59
2-12
12-22
2-6
0-10
10-20
20-30
30-40
40-46
2-6
1-6
13-15
22-24
2-7
7-12
12-17
2-12
12-24
2-6
6-18
2-10
Rn222
Ra226
1.02
1.01
.68
.68
.75
1.01
1.21
—
.79
1.05
1.24
.80
.69
.71
.71
.78
1.04
—
—
.81
1.01
1.17
.78
.68
.38
.50
.58
.46
1.11
1.03
.94
.98
1.17
.93
.45
Ra
(dpm/gm
dry)
.61
.58
.88
.69
.59
.43
.39
.55
.52
.45
.40
.05*
.41
.35
.34
.41
.37
.43
.63
.55
.34
.40
.49
.50
.35*
.48
.48
.34
.41
.39
.24
.19
.23**
.24
.06*
Ce 3
(dpm/cm
sediment)
.45
.54
.64
.48
.33
.25
.36
.39
.27
.32
.30
.06
.29
.29
.29
.26
.39
.39
.56
.29
.26
.40
.43
.52
.10
.22
. 31
.47
.21
.25
.27
.16
.11
.23**
.28
.08
*Sandy sediment with shell debris
226 _ _ 222
**Assuming Ra - Kn
105
-------
RADON(dpm/l) AND SALINITY (%o) VS. DEPTH(m)
mp
63 22 JULY mp 60 22 JULY mp 59 23 JULY
\ e i p • i **. ,**.•• . ^- A i— iv* *
mp
47 23 JULY
5
10
15
20
0.5 I.5.JLO 0.5 1.5 0.5 1.5 2.5 0.5 1.5 2.5
" T 1 "Jf
? 5
15 5 10 15 5 10 15 20 25
%o %o %c
Figure 41. Radon sampling transect along the axis of flow - July 1972.
-------
Radon Profiles - Tappan Zee
1971 -72
222
dpm Rn /IOO/—
meters) .
100 20O
-\ \
.i 1 9/27/71 4
- • 6 m.p. 26
- ~-,.x \ 8
12
10
Surface - 9/27/71
100 20O
A
4
A
/Y\ 12
100 200 IOO 200 100 200
, . « • •; i i i p • i • l
1 L 10/10/71 4 _ ', ^^ 4 \ j 8/22/72
- ll m.p. 26 - \ / 8/8/72 - '\x m.p.22
" \\ 8 " / m.p.22 8 - \
- \\ - / j
- \ b 12 - 6 12 - ! /
! 3
10 10 IO
Surface -9/4/72 m.p. ,9_ n/7 mpz7-../7
100 200 Surface - 11/7/72 I00 -^ I00
1 ' ' 1 ' ' ' 1 ' ' • • L 1 • • • i. h | • •
A 80 A A - T -T
A - ^ j
6°- f 8 - 8 '. 1
A 40- A 1 - ' - i
A 12 . i J I2 .
L 20- '
^rlnm I0 I0
mile point — •
Figure 42. Time series of radon within a limited reach of the Hudson over
a period of four months.
-------
o
00
(O
o
z
o
o
CO
UJ
UJ
N
0.
0,
1 — 1 — i — 1 —
1.05
* *
1.53
1.40
~ * * 1 4.7
1.34 ll47
1.54
1.42
2.01
107 1.82
. ,.L i. I 1 i J
i i i i i • i i n i T — i — i — r
RADON (dpm/l) TIME
O.95 | 0\ ^ \ A i
**""**' t,\J\J 1 «T -9
1.35 1.39 1.53
1.30 1.44 1.42
U6 1.20 1.47
1.27 1.85 I'll
1.29
0.97 1.18 1.33 1.45
I i l i i 1 1 1 I I i i i
(iii;
SERIES
-
l.33_
# ~
1,29
1.15
I 1 1 1 1
o
z
3:
8-5
Figure 43.
56
48
24
8-10 8-15 8-20
AUGUST 1974 SAMPLING DATE
^
8-25
O
10
5
0
5 «
E
Time series of radon in the Hudson over one month: The number indicates activity
(dpm/l) in a surface sample in mid-channel, * indicates a surface and deep sample
were averaged, ** indicates a surface and two deeper samples were averaged. No
correlation with rainfall, wind or tidal amplitude was apparent.
-------
several potentially significant environmental variables. No consistent
variation in radon levels could be correlated with tide heights (which varied
by a factor of two) although the bottom currents and hence the rate of
stirring of surface sediments by flow of water near the bottom (current
turbation) might be expected to vary with tidal amplitude, with rainstorms,
or with wind velocity. The median radon concentration for all of the summer
(June-September) and winter (December-March) samples was determined for various
segments of the estuary (Table 18). Since no consistent axial or vertical
trends were observed, all samples from north of mp 16 were lumped to form the
histogram in Figure 44, establishing a median for the estuary and displaying
the extent of variability. The median is 1.43 + 0.25 dpm/1 in summer and
1.30 + 0.35 dpm/1 in winter.
A few measurements of Ra-226 were made on both filtered and unfiltered
water samples. About half the Ra-226 is dissolved (passes -45 n filter) and
half is on suspended material. The total is 0.085 + .04 dpm/1 (median of
36 analyses). The analytical precision was about 7% for both Ra-226 and Rn-222
analyses.
Several small streams entering the estuary have radon concentrations 5-10
times that of the estuary (Table 19), due to their shallowness and the
contribution of groundwater flow. All of these cascade over dams or waterfalls
shortly before entering the estuary. The turbulence introduced by this process
is very effective in degassing radon from streams as shown by the measurements
in Sparkill Creek on August 6, 1974.
McCrone (1967) has observed that the grain size, cation exchange capacity
and percent organic carbon of Hudson estuary sediments are rather uniform with
depth and location along the estuary between mp 20-76. This observation is
supported by the rather constant value of radon released per gram of dry
sediment (Ra-226 in Table 17), although samples north of mp 40 average ^ 30%
greater than those to the south. As one approaches the harbor, the release
rate drops sharply and the variability increases, probably due to an increasing
sand fraction and dredging activities. An average value in the top 20 cm was
computed for each location and each location was weighted equally to estimate
mobile radon in each of two large areas (listed in Table 21 as C£q per wet
sediment volume) with the uncertainty expressed equal to one standard
deviation. Sandy samples from the navigation channel were not included in the
average since they cover less than 5% of the total area.
DISCUSSION
The transport of radon across the sediment-water interface can be
calculated by constructing a mass balance for radon in the water column (Table
20). The potential inputs are migration from sediments, production in the
water column, stream inflow, and groundwater. The sinks are decay and evasion
to the atmosphere. Gas bubbles which escape from sediment and rise through the
water column are of negligible importance in transporting radon (Hammond et al. ,
1975) When both sinks are considered, the mean residence time of a radon
atom in the estuary is T - 3 days. With an effective horizontal eddy
diffusivity of K = 700 m /sec computed on the basis of a one-dimensional
advection-dif fusion model (Simpson and Hammond, 1977), a radon atom will move
109
-------
RADON HISTOGRAM
SUMMER (JUNE- SEPTEMBER)
MEDIAN =l.43± 0.25
n =106
8
WINTER (DECEMBER-MARCH)
MEDIAN" l.30±0.35
n = 17
ih
.80 1.00 1.20 1.40 1.60 1.80 2.00
RADON 222 (dpm/l)
Figure 44. Histogram of radon data for the Hudson: All samples were
collected at least 200 m from shore between mp 16 and mp 76.
110
-------
TABLE 18
Estuary Geometry and Median Radon Concentrations
mp 16-20 20-40 40-57 > 57
Mean depth (m)
Area (km )
10
9
5.3
116
12.8
29
Median summer radon
(dpm/1) 1.45-K 31 1.43+^26 1.40^.27 1.50+.35
Number of samples 28 35 33 10
Median winter radon
(dpm/1) 1.22+.15 1.42+.57 1.10^.30
Number of samples 4 76
111
-------
TABLE 19
2
Radon (dpm/1) in Streams (mp 0-91) with Drainage Area > 100 km
18
25
34
45
47
58
59
60
67
91
mp
Saw Mill
Date
(1974)
River
Sparkill Creek
Croton River
Annsville
Popolopen
Creek
Creek
Moodna Creek
Quassaick
Fishkill
Wappinger
Wallkill
Creek
Creek
Creek
River
( above
(above
( above
(above
(below
dam, no ice)
dam, under ice)
dam, no ice)
dam)
dam)
Jan
Feb
Feb
Aug
Aug
29
13
20
6
6
Rn
dpm/1
39
62
37
23
9
.6
.5
.6
.8
.8
(summer flow)
(below
rapids)
(in rapids)
(below
(below
falls)
dam)
Aug
Aug
Aug
Aug
7
7
7
7
9
0
11
2
.3
.9
.3
.1
D.F.**
.002
.0006
13 km
Anomaly
.02*
.001
.002
.010
.004
.010
.006
.013
.011
.087
.02*
.10*
.04
.09
*
.06*
.01
.12
.18
Rondout Creek
* Assuming Rn = 10 dpm/1
** D.F. = .01 yrs x 0.5 m x Drainage ARea/(River Cross-Section) x 13 km
-------
TABLE 20
Radon Budget for Hudson Water Column
(atoms m sec"1)
Region
Season
LOSSES
Decay
Evasion
INPUTS
Radium-226
Streams
Groundwater
From Sediments
Tappan
Summer
126+23
163+82
8+4
3+5
0
278+85
Zee
Winter
125+50
126+63
8+4
3+5
o'
240+81
Hudson
Summer
298+58
160+80
18+8
36+10
80+40
324+107
Highlands
Winter
234+64
96+48
18+8
36+10
0+40
276+80
TOTAL LOSS = TOTAL INPUT
289+85
251+80
458+99
330+80
113
-------
TABLE 21
Comparison of Diffusive, Observed and Turbated Fluxes
C (dpm cm )
eq
Number of locations
Season
2-1 9
D (m sec x 10 )
e
Calculated diffusive
— 2 T
flux (atoms m sec )
Observed flux
_ n
(atoms m sec)
Incremental flux
(atoms m~2sec~-'-)
C (dps cm~3) x 10~4
eq
Minimum depth (cm)
Tappan Zee
0.33 + 0.10
8
Summer Winter
1.4 0.7
1. 2+_. 1 1.2+_.1
118+48 83+_34
278+85 240+81
160+98 157+88
55+17 55+17
3+2 3+2
Hudson Highlands
0.42 +_ 0.11
5
Summer Winter
1.4 0.7
1.2+.1 1.2+_.l
151+^55 107+_39
324+107 276+_90
173+^120 169+98
70+JL8 70+18
2.5+2 2.5+2
114
-------
1/2
an average of (KT) = 15 km from its source. Thus, radon should be well
mixed across the estuary, although large regional source variations could
result in gradients along the flow axis. We have divided the estuary into
segments and have constructed a mass balance for radon in two of these, the
Tappan Zee region (mp 20-40) and the Hudson Highlands (mp 40-57). Since no
consistent axial or vertical gradients exist, we treat each segment as a well-
mixed box. In the absence of axial gradients, we can ignore horizontal
advection and diffusion into and out of each box. The principles used to
construct the mass balances are outlined below, and the construction is done
for a unit surface area in each box.
Production in the Water Column
Radon is produced at a measurable rate by the radioactive decay of Ra-226
dissolved or suspended in the water. This rate can be calculated by inte-
grating ,Ra-226 activity over the mean depth. No consistent seasonal or
regional variation in this activity was observed, and the average concentration
from all of the data of 0.085 jf 0.04 dpm/1 was used. The variation is
probably largely due to variations in the amount of suspended material in each
sample. Suspended plus dissolved Ra-226 accounts for less than 5% of the total
radon input.
Stream Inflow
The contribution of stream runoff was calculated by taking the average
stream flow over the year, assuming an average concentration of 10 dpm/1, and
dividing by the estuary area in the segment of interest. In the Tappan Zee,
runoff is 0.13 km yr and the input is 3 +_ 5 atoms m sec . Injthe Hudson
Highlands, runoff is 0.42 km yr and the input is 36 + 10 atoms m sec
This input is about 1% of the total in the Tappan Zee, and about 10% of the
total in the Highlands.
The anomaly which a stream might produce in the estuary is listed in
Table 19 and was calculated as follows. The tidal excursion of the estuary
is about 13 km. Thus, a stream input will be mixed throughout at least this
range. For each large stream a dilution factor (D.F.) was calculated,
assuming rainfall of 100 cm/yr, runoff of 50%, estimating the stream drainage
area, and assuming that in a mean radon life (0.01 years) this volume of water
is added to 13 km. The anomalies (dpm/1) which could be expected from each
stream is listed in Table 19, and are generally less than 10% of the median
concentrations. An observed anomaly might be several times larger, since
rainfall is episodic.
Groundwater
Supply of radon to the estuary by groundwater is the most Difficult input
parameter to characterize. Groundwater typically contains ^ 10 dpm/1 and
thus a small input could add significant amounts of radon. To introduce 50%
of the total radon inputs would require groundwater flow amounting to 2.5% of
the precipitation falling in the drainage basin surrounding the Hudson
Highlands region and 22% in the drainage basin surrounding the Tappan Zee.
Thus, the Tappan Zee should be free of major groundwater inputs derived
locally. The bedrock geology does not indicate any outcrops of potential
115
-------
aquifers and the highly impermeable glacial clays which underlie the Tappan
Zee make groundwater flow even more unlikely.
We have observed a type of groundwater flow in the Hudson Highlands,
however, along coarse fill used in embankments for railroad tracks. Pumping
of water by the rise and fall of the tides occurs through these fill areas.
As the tide rises, the banks accumulate water. As it falls, the water
percolates back into the river. Water from one of these percolating outfalls
was found to have a radon activity of '^ 100 dpm/1. Assuming this is typical,
that the banks occupy 2% of the estuary surface area, and that 0.25% + .12 m
of water exchanges each tidal cycle, the summer groundwater input to the
water column in the Highlands is equivalent to 80 +_ 40 atoms m sec . In
winter, the embankments often should be frozen and tidal pumping greatly
reduced. Railway embankments are not nearly as common along the Tappan Zee,
and the great width of this region should make a tidal pumping input much
less important than for the Highlands. Thus, we assume that groundwater is
not a significant input to the Tappan Zee.
Radon Decay
The primary sink for radon in the water column is decay. The rate can be
calculated by integrating the average concentration over the mean depth.
Mean depths were computed from the Hudson River Navigation Charts and include
marsh areas.
Although mixing in the estuary is rapid it is not instantaneous. If the
combined effects of decay, evasion, and input from sediments (Table 20) were
allowed to act on a column of water 1 m deep which was isolated from the rest
of«the_estuary, this water would accumulate radon at a rate of about 90 atoms
m sec . In half a tidal cycle (2 x 10 sees), this would amount to 0.2 dpm/1
in concentration. Thus, incomplete mixing of marsh waters with the remainder
of the estuary may create some variation in radon concentrations.
Evasion to the Atmosphere
During October 1974, two weeks of daily measurements were made to
determine the concentration of radon in equilibrium with the atmosphere, by
flushing air through 19 liters of distilled water for ^ 2 hours at ^ 3 1/min.
Results ranged from 0.02-0.05 dpm/1 at 15-20°C. Thus, estuary radon is
considerably in excess of atmospheric equilibrium, and a net evasive flux
occurs. The rate of evasion has been calculated by use of the Lewis and
Whitman (1924) stagnant film model which envisions gas transport across the
air-water interface to be limited by molecular diffusion through a thin film
of water. The surface of this film is assumed to be in equilibrium with the
atmosphere and the base of the film has a concentration equal to the measured
surface water concentration. Film thickness has been related to wind speed
(Broecker and Peng, 1974; Emerson, 1975) and we have estimated a thickness of
95 y for the Hudson in summer (average wind speed - 4.4 m sec ) and 60 y in
winter (average wind speed - 5.9 m sec ). As discussed by Hammond and
Simpson (1977), this estimate is based on more representative wind speed data
than the film thickness used by Hammond e_t al_. (1975). Using surface water
concentrations listed in Table 18 and assuming atmospheric equilibrium to be
0.03 dpm/1 in summer and 0.06 dpm/1 in winter, the values in Table 20 were
116
-------
calculated. Evasion and decay are of comparable Importance.
Variations in evasion could cause changes in radon concentration. If the
exc.hang| rate was doubled in the Tappan Zee, radon would be lost at about 160
m sec , and in one day this would amount to a decrease of 0.35 dpm/1. If
the exchange rate was halved, the anomaly would be +0.18 dpm/1. Variations in
wind speed can produce changes in the water column concentrations similar to
the observed variations.
Diffusional Input from Sediments
We have now established estimates of the loss terms for radon from two
segments of the Hudson and all of the important input terms except one, the
supply from sediments. This input must balance the radon budget, and its
magnitude can be compared with the amount that molecular diffusion in the
upper layers of the sediment should supply.
The problem of molecular diffusion in sediments has been discussed by
Berner (1971) and equations for radon diffusion has been derived by Broecker
(1965). Writing a balance for radon production, decay, and diffusion in one
dimension
^ = AC - AC + f- (D |£) (1)
dt eq dx s dx
where C = concentration of radon per wet sediment volume at x
C = concentration of radon supported by emanation at x
A = decay constant for radon
D = effective diffusivity of radon in sediments
x = depth in sediment
Porosity is assumed to be constant with depth and does not appear in the
equation because of the definition of C. Assuming C and D to be independent
of depth, eq. 1 can be solved for the flux across the sediment-water interface
per unit area,
J = (D A)1/2 (C - C ) (2)
o s eq o
where C = concentration at the sediment-water interface, taken to be equal
° to that in the overlying water column.
Under molecular diffusion Dg = -j where Dm = molecular diffusivity and
9 = the effect of tortuosity on diffusive path length (0 > 1), Dm at 18°C
has been measured by Rona (1971) and the effect of temperature has been
calculated by Peng (1973). Li and Gregory (1974) have pointed out that
diffusivity should increase by no more than 8% when salinity increases from
zero to 35°/00. The effect of tortuosity is less well constrained. Li and
Gregory (1974) found 9 = 1.35 for Pacific red clay with a porosity of 50%.
Hudson sediments typically have a porosity of 60-80%, so an estimate of 9 = 1.2
should be accurate within 10%. The calculated diffusive flux is listed in
Table 21. The uncertainty is primarily due to variability of measurements of
C . The calculated flux accounts for only 40% of the total input required by
tni'budget (from Table 20). To supply the required flux, Dg must be increased
by a factor of 6. It is unlikely that the measurement of Dm is this much in
117
-------
error. Instead, we suggest that stirring of the surface layer of sediment
(turbation) is significant in transporting radon across the sediment-water
interface. Since the turbation effect appears to be similar in both summer
and winter, and the benthic macrofauna are expected to be much less active at
low temperatures, currents are probably the primary cause. An additional
factor which appears to limit benthic organism activity in the area of the
Hudson of interest is that of extremely variable salinity. Both the Hudson
Highlands and Tappan Zee contain completely fresh water for appreciable periods
of the year, although average salinities range from a few parts per
thousand up to about 10 parts per thousand.
It is conceivable that roughness of the sediment surface might increase
the total area of sediment sufficiently to enhance the radon flux above that
computed using a planar surface. We can evaluate the significance of surface
roughness with some simple calculations. Radon gradients should have a half-
distance of about 2 cm, so topography with a wavelength greater than this will
increase the surface area contributing to the diffusional flux beyond the
surface area of a plane. Assuming a sinusoidal morphology with amplitude A
and wavelength 2R, the path length dl covered by horizontal movement dx is:
JT i , /Air TTx,2 , ,ON
dl = 1 + (-— cos —) dx (3)
K K.
Large scale features of the channel examined with a PGR indicate that A/R
< 0.06 for features with amplitudes of a few feet or more. The absence of
microtopography is more difficult to demonstrate but an upper limit might be
A/R = I/TT. This morphology would increase the diffusional area by 22%.
However, it is unlikely that such topography exists in the muddy sediments of
the Hudson and we take the surface area of the sediment-water interface to be
equal to that of the map area.
Turbation Input from Sediments
To quantify transport due to turbation is a difficult problem. The
actions of organisms have been modeled as a diffusive mechanism (Goldhaber
et_ _§!_. , 1975; Hammond ej^ _al_. , 1975; Guinasso and Schink, 1975 and others).
Transport in this type of model is characterized by an eddy diffusivity which
exceeds molecular diffusivity within a turbated zone. Assuming that turbation
is uniform spatially and temporally, we can calculate the minimum thickness of
the turbated zone required to supply to observed flux (Table 21).
Alternatively, by assuming a depth to which turbation is effective, the
eddy diffusivity required in the turbated zone to produce the observed flux
across the sediment-water interface can be calculated, treating the sediments
as a two-layer system. In the upper turbated layer (thickness = d), a uniform
eddy diffusivity D prevails. Below this, transport is by molecular diffusion
and D = D . A solution to the two-layer problem for radon in the surface
ocean has Been presented by Peng et_ _al_. (1974). Following this approach, the
general solution to eq. 1 is:
(C - C) = Me~aX + Ne+ax (4)
eq 1/2
where a, M and N are constants. In the upper layer a = (A/D ) and in the
lower layer a^ = (A/D,p . The unknown quantities are M N M N and a .
118
-------
0.2 0.4 0.6 0.8
1.0
0.2 0.4 0.6 0.8 1.0
h-
UJ
2
O
LU
CO
8
&°
O
12
UNIFORM
STIRRING
MODEL
o
H
<
-------
Defining the lower layer to begin at y = 0 (so y = x-d), the boundary
conditions are that radon concentration at the sediment-water interface must
equal the concentration in the overlying water column
C = 1.43 dpm/1
x=o
the flux across the interface must equal the observed flux
ip —9—1
J = D -:— = 260 atoms m sec
o s dx
x=o
the concentration at the base of the upper layer must equal that at the top
of the lower layer
-ax ax -ay ay
M + N = M2 + N2 at x = d, y = 0
e e e e
the flux leaving the lower layer must equal the flux entering the upper layer
D! f = D, f
1 dx , 2 dy .
x=d J y=0
and deep in the sediment the concentration must equal the equilibrium value
C = C at y = oo
eq
With five boundary conditions and five unknowns, the concentration vs. depth
profile can be calculated. If the thickness of the turbated zone is large,
the calculated profile of radon deficiency (Figure 45) is a simple exponential
and D = 8 x 10 cm sec . If the thickness is chosen to be 4 cm in the
Tappan Zee, the calculated summer profile_(Figure 45) clearly reflects the
two-layer character and D = 1.2 x 10 cm sec . The predictive power of this
model is limited, however, unless either D or d can be evaluated independently
since D is quite sensitive to d. Surficial sediments in this estuary have
high porosity and low viscosity (Olsen et al., 1976). The thickness of this
soupy layer observed in cores averages about 2 cm in many areas. Below this
depth, the viscosity increase is sufficient for sediment to retain its shape
when extruded from core liners. On the basis of textural criteria, we chose
d = 2 cm. This value is smaller than the minimum thickness required to produce
the average observed flux, (Table 21), although it is within the estimated
uncertainty.
It seems likely, however, that uniform stirring to 2 cm is not the only
mechanism which can augment the diffusive flux in transporting substances
across the interface. Ten of the 106 summer measurements had radon concentra-
tions more than 3 a from the median. This is more than twice the variability
expected statistically. Three of these samples were collected on September 15,
1972 at mp 18 and are plotted vs. depth in Figure 45A. Their relationship
to salinity indicates that the high values observed are not analytical contam-
ination, but represent a localized and transient injection of radon into the
water column. Assuming a mean water depth of 10 m and an average concentration
of 6 dpm/1, to generate such an anomaly would require injection of 5 x 10
atoms m , or all of the atoms stored in 5 cm of sediment at equilibrium.
This estimate is a minimum since it neglects evasion, horizontal transport,
and decay after injection. Portions of the estuary are dredged periodically,
but not such activities were noted when samples were collected. Estuaries
are regions of active erosion and redeposition, and it seems reasonable that
120
-------
RADON222 (dpm/l)
2
4
_ 6
E
— *
f 8
r^
O.
LU
Qio
12
1 bv
14
24 6 8 10 12
^1 ^Ql 1 I 1 I
NORMAL \ \
-SUMMER i \ jo
VALUES ' \
\ \ I8
^ \ S
\ \ ^4
\ \ £2
\ \
\< .
/
/
A
/
j/
x
i i i i
—
__
\ \ 4 8 12 16
v \
\ Q SALINITY (%o)
S\ \
9-15-72 \
\
mp 18 \ \
\
i \
SALINITY 0 i \
' D 222 <*"\
. Rn -5 v
- ' U
1 1 1 1 1
5 10 15 20
SALINITY %0
Figure 45a. Radon-222 concentration (dpm/l) vs. water depth (m) in the Hudson
15 September 1972. Salinity values are also shown.
121
-------
transient anomalies can be produced by stochastic events such as slumps.
The magnitude of the anomaly which a slump might produce depends on the
thickness of sediment involved and the time which elapses between slumping
and sampling.
Assuming that current turbation occurs uniformly and rapidly enough to
remove all radon from the upper 2 cm of sediment, that transport occurs by
molecular diffusion alone below 2 cm, and that values in Tab^e 2I^are accurate,
stochastic reworking must supply a flux of about 50 atoms m sec . If this
flux is supplied from sediments containing equilibrium concentrations of radon
reworking can be characterized by a mass transport parameter, k, where
k = reworking flux = (depth of reworking) (frequency of reworking)
L/
eq
On the basis of our data± an average radon flux due to stochastic
reworking yields k = 0.6 m yr , but the frequency and depth of reworking
cannot be independently constrained. We can estimate the maximum frequency
on the basis of time required for ingrowth of radon in sediment (^ 10 sees),
requiring a reworking depth of 2 cm (below the turbated layer). A maximum
depth of 2 m might be estimated, requiring a reworking frequency of once every
3 years. If the relaxation time from a reworking event is one day (before
radon concentrations in the water column return to "normal") and we observe
an anomaly on one of twenty sampling trips, the frequency is about 5 x 10
sec and the average reworking depth should be 4 cm.
CONCLUSIONS
Radon-222 is distributed rather uniformly throughout the water column of
the Hudson estuary (Table 22). Vertical and axial variations are generally
less than 20%. Slightly lower concentrations are observed in winter than in
summer. By estimating the rate of gas exchange in the Hudson a mass balance
for radon can be constructed. The sinks for radon are decay (40-65% of total)
and evation (60-35% of total) with relative rates depending on mean depth.
Inputs from radium-226 decay in the water column, stream runoff, and ground
water are small in comparison to inputs from sediment.
The flux of radon supplied by molecular diffusion from sediments accounts
for only 40% of the total input from sediments. This indicates at least one
other transport mechanism involving the sediments must also be significant in
the supply of radon. If we assume that the upper two centimeters of sediment,
which has very low viscosity compared to underlying material, is stirred
rapidly and continuously by tidal currents (relative to diffusion), a supply
of radon comparable to that of molecular diffusion can be accounted for. A
second mechanism, attributed to stochastic reworking of sediment by erosion
and slumping, contributed a radon flux of approximately half of that supplied
by molecular diffusion.
Quantitative application of these models to the transport of other
dissolved species is not yet possible, because we cannot distinguish between
the depth to which such processes operate and their rate of occurrence.
Because nutrients, transition metals, and oxygen will have concentration vs.
depth profiles which differ from that of radon, a one-to-one correlation
between their transport and that of radon cannot be made. Detailed profiles
122
-------
TABLE 22
Summary of Rn Measurements
Coll.
Date
1971
9/23
9/27
11/10
1972
6/26
7/10
7/18
7/22
7/22
7/23
7/23
7/22
7/25
M.P.
25
25
26
25
18
18
18
63
60
54
47
37
-3
Depth
(m)
0
1
6
10
1
0
0
0
4
8
12
0
4
8
0
3
15
0
5
15
0
7.5
14
0
5
10
0
7
17
0
10
20
30
0
4
10
0
4.5
8
Sal.
1.0
0.6
1.9
8.1
0.5
1.2
0.9
3.7
4.0
4.7
7.3
0.0
0.0
0.2
0.4
2.0
0.0
2.0
12.3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
5.4
19.1
22.8
25.2
Temp.
20.4
20.6
21.7
VL3
19.3
21.6
21.4
20.9
24.4
23.7
21.6
24.6
24.5
29.2
24.7
23.0
22.9
21.2
20.2
Excess
„ 226 _ 226
Rn Rn
dpm/1 dpm/1 Notes
1.89 from boat mooring
0.96 0.052 station no. 1
0.99 0.052
1.12 0.085
1.23 0.032 station no. 2
2.33 0.079 from west shallows,
station no. 3
1.31 0.038 at pier,
station no. 4
0.42
0.49
0.45
0.73
1.22 0.099
1.20 < 0
1.21 0.107
1.25
^1.2 (could not find
^1.0 analyses recorded)
1.62
1.55
1.44
10.96
1.38
1.51
1.23
1.46
1.28
0.90
1.28
2.75
1.30
1.16
2.02
1.14
1.46
1.52
2.04
1.13
1.15
1'08 ("continued
123
-------
Coll.
Date M.P.
7/26 16
8/1 18
8/8 21
8/14 18
8/22 21
9/4 25
9/15 18
9/23 18
10/18
11/7
1973
8/23 31
Depth
(m)
0
9
15
0
5
7.4
14
0
5
12
0
0
4
10
14
0
0
0
0
0
8
12
5
10
14
0
5
9
12.5
0
0
12
0
0
0
0
10
0
6
>10
TABLE 22 (continued)
Excess 226
Sal. Temp. Rn Rn
(°/00) (°C) dpm/1 dpm/1
7.3 25.1 1.38
11.7 23.7 1.49
12.3 23.6 1.78
5.1 25.0 1.25
6.9 23.9 1.25
7.4 23.8 1.19
10.3 23.4 1.47
5.2 2.37
5.9 1.24
7.1 1.04
4.2 0.95
5.3 1.27
6.9 1.11
14.1 2.09
14.4 2.01
1.16
1.32
1.03
5.20
3.58
6.30
10.62
1.44
1.57
1.27
0.56
0.79
0.86
0.69
0.57
0.80
1.20
0.61 0.089
0.33 0.102
0.73
0.78
1.22
1.27 0.13
1.45 0.15
2.10 0.30
Notes
station no. 1 east
8 1
station no. 3,
8 1
station no. 4,
8 1
bucket sample, 8 1
8 1
8 1
8 1
8 1
8 1
8 1
8 1
8 1
8 1
8 1
station no. 1
M.6.5 1
station no. 1, 8 1
station no. 1, 8 1
station no. 2, 19 1
station no. 3, 19 1
station no. 4, 8 1
station no. 4, 19 1
station no. 4, 8 1
hit bottom, mud
on Niskin
(continued)
124
-------
TABLE 22 (continued)
Excess
Coll.
Date
1974
8/5
8/6
8/7
8/9
8/13
8/16
8/18
8/19
8/22
8/23
M.P.
25
18
91
76
62
57
53
49
46
51
47
42
37
29
18
52
47
37
31
25
18
18
52
47
37
31
25
18
52
47
37
31
26
25
24
18
Depth
(m)
0
0
0
0
0
0
10
20
0
20
40
0
24
0
12
24
0
0
0
7
14
0
8
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Sal.
0.0
0.0
0.0
0.1
0.1
0.1
0.5
0.5
0.4
0.7
1.3
1.4
1.6
1.7
0.2
0.5
1.3
1.4
1.5
2.3
2.9
5.0
8.4
Temp.
25.8
25.7
25.7
25.5
25.6
25.6
25.7
25.5
25.8
25.8
25.7
ND
ND
26.8
25.8
25.9
27.0
26.3
Rn "u Rn "u
dpm/1 dpm/1 Notes
1.73
0.92
0.36
1.31
0.90
0.77 0.049
0.78
0.90 0.083
1.03 0.089
1.04
0.82
1.30
1.20
1.14
1.29
1.13
1.38 0.154
1.32
1.27
1.30 0.104
1.59
1.25
1.28
1.86
1.67
0.80
1.20
1.15
1.01
1.12
0.82
1.03
1.35
1.24
1.29
1.05
1.70
1.18
1.28
1.38
1.27
1.32
1.07
1.08
1.14
( f-< /-\TI 1 — i -n 1 1 ^ /-I 1
125
-------
Coll.
Date
8/27
8/28
9/10
12/13
1975
1/4
2/10
3/7
M.P.
26
18
37
1
-12
76
47
25
18
57
53
47
41
35
25
18
25
18
25
18
Depth
(m)
0
6
0
5
0
12
0
0
1
1
1
1
20
1
1
15
1
9
1
9
1
1
1
1
1
TABLE 22 (continued)
Excess
Sal. Temp. Rn
(°/oo) (°C) dpm/1
0.97
1.31
1.00
3.6 27.1 1.18
18.8 25.2 1.01
29.0 22.0 0.76
0.0 ND 1.70
ND ND 1.56
1.18
1.11
0.68
0.66
1.33
1.89
0.86
1.04
1.93
1.13
2.63
1.05
0.92
1.27
0.88
1.84
1.27
r. 226
Rn
dpm/1
0.22
0.22
0.078
0.075
0.079
0.085
0.140
0.062
0.068
0.208
0.122
0.095
0.102
0.109
0.054
0.076
0.057
0.073
Notes
from dock
from dock
from pier
from dock
from pier
from dock
from pier
from dock
126
-------
of radon distribution in sediments could help provide answers to these
questions. It is clear that uniform current turbation of surficial sediments
will accelerate consumption of oxygen, decay of organic matter, and recycling
of nutrients to the overlying water column. Stochastic reworking of sediments
from depth may be effective at introducing trace metals to the overlying water
column as this allows materials to bypass the zone of oxidized sediment near
the sediment-water interface by mixing anoxic interstitial waters directly
with the overlying water column.
127
-------
<
CO
ATMOSPHERE
STAGNANT WATER FILM 100-250 JJL
OXIDIZED WATER
COLUMN
I-30M
o
CO
ID
IL.
u_
Q
>
Q
Q
UJ
OXIDIZED
SEDIMENT
SULFATE "
REDUCTION ^''
^ rf»
METHANE PRODUCTION
0-20CM
or
51
o to
d t
^ Q
NO METHANE
OXIDATION?
„METHANE
„ OXIDATION^.
"NO"ME"THANE
OXIDATION
INTERFACE
PARTIAL
SOLUTION
INTERFACE
BUBBLES
FORM
Figure 46. Biogeochemistry of methane in estuaries.
-------
SECTION 10
METHANE AS AN INDICATOR OF TRANSPORT PROCESSES
BETWEEN THE SEDIMENTS AND WATER COLUMN IN THE HUDSON
INTRODUCTION
Methane concentrations in natural waters are controlled by a balance
among production, oxidation and mass transport. Methane production and
oxidation have been studied in a wide variety of aqueous environments by both
microbiologists and geochemists, although neither discipline has yet produced
a fundamental understanding of the role of micro-organisms in either of these
processes. Some of the controlling factors have recently been reviewed by
Reeburgh and Heggie (in preparation). Methane production occurs by at least
two mechanisms (Doelle, 1969; Mechalas, 1974), including reduction of C0?
(Claypool and Kaplan, 1974) and acetate fermentation (Koyama, 1963). The
microorganisms responsible for these processes are strict anaerobes but may
be inhibited by the presence of sulfide (Cappenberg, 1975). This usually
results in a lack of methane production in the presence of sulfate reduction
(Atkinson and Richards, 1967; Tsou et. _al., 1973; Martens and Berner, 1974).
Methane oxidation can be carried out by a large number of aerobic
microorganisms commonly found in lakes (Cappenberg, 1972; Rudd et_ _al_. , 1974),
rivers, river mud, soils (Whittenbury et_ aJL. , 1970) and marine mud (Hutton
and Zobell, 1949). Isotopic evidence and mass balance considerations suggest
that methane oxidation can also be accomplished by sulfate reducers (Feely and
Kulp, 1957; Nissenbaum _e_t al. , 1972; Reeburgh, 1976).
Methane transport through sediments may occur by diffusion, turbation and
bubble migration (Figure 46).. Diffusive processes have been discussed by a
number of authors and a concise treatment can be found in Berner (1971). The
importance of turbation by organisms and currents in transporting dissolved
substances has recently begun to be quantitatively assessed (Goldhaber et al.,
1975; Hammond &t_ _al., 1975). The importance of bubble transport across the
sediment-water interface (hereafter called sediment flatulence) has been
recognized by a number of authors (Koyama, 1963; Reeburgh, 1969; Emery, 1969;
Martens, 1976 and others). To date, quantitative estimates of bubble transport
have been difficult to make. Three approaches have been used: (1) examination
of the decrease in organic carbon with increasing depth in the sediment, (2)
counting the rate at which bubbles burst at the air-water interface, and (3)
leaving a bubble trap above the sediment-water interface for a period of time.
Each of these approaches presents complications. The first approach will be
difficult to use in the presence of turbating organisms. Their activities may
homogenize sediments and made the decreases in organic carbon with depth
difficult to observe. The second approach is only useful for systems with
129
-------
rapid flatulence and quiescent'surfaces. Systems with slow flatulence would
require observation of large areas for long periods of time to determine fluxes
and while the number of bubbles can be counted, their mean size must be
estimated to calculate fluxes. Estuaries rarely have surfaces sufficiently
calm to permit use of this technique. The third method creates some logistical
problems and is only useful for monitoring a rather small area. We present an
application of an alternative approach for estimating flatulent releases,
based on constructing a mass balance for methane in the water column.
MEASUREMENT TECHNIQUES
Water samples were collected in Niskin samplers and transferred through a
tube to glass bottles (^ 115 cc) which were overflowed by one volume to
minimize gas loss. To prevent oxidation during storage, 0.5 cc of HgCl^
(saturated) solution was added and the bottle was capped with a serum cap
pierced by a hypodermic needle, preventing bubble entrapment. In practice,
a small bubble always forms. To minimize diffusive loss from this bubble
through the cap, bottles were kept cool' and stored upside down. An aliquot of
this sample was injected into a stripping system patterned after those of
Swinnerton j2t_ al. (1962) and'Weiss and Craig (1973). Its dissolved gases were
extracted by continuously purging with helium carrier. Methane was separated
from 0- and N~ on a molecular sieve column (#5A) and measured on a flame
ionization detector. Peak resolution is essential for samples containing less
than 1 yM CH, since at hight sensitivity the detector we used gives a negative
response to 0 and a positive response to N.. Since high sensitivity also
increases baseline noise, peak height can be determined with greater
reliability than peak area, although the latter must be known to compare the
signal from samples to those' from a gas standard which was injected
periodically from a gas sampling loop for calibration. The relationship
between peak height and area is a function of the condition of the column and,
to a lesser extent, of both stripping temperature and sample salinity. The
area/height ratio for water samples .was determined periodically by use of a
freshwater liquid standard with CH, ^ 5 yM and is typically 1.1 to 1.3 times
that of the gas standard. The sensitivity with the above procedures is 0.01 yM
and the absolute accuracy is ^ 6%, with the major uncertainty introduced by the
peak area/peak height ratio. Precision for poisoned samples run on different
days is about + 3%.
A number of storage tests.were performed at various temperatures, on
waters of varying salinities, as a check "on the procedure and to determine if
CH, is oxidized by bacteria. The storage procedure for poisoned samples
appears to be good for at least two weeks and is independent of temperature
or light conditions. Storage temperature for unpoisoned samples was chosen
to be within a few degrees of in situ temperature. These tests showed that the
rate constant for methane oxidation is equivalent to a mean lifetime for
methane of more than 20 days in Hudson waters with a salinity of less than a
few parts per thousand and a lifetime much longer than this in higher salinity
water.
Measurements of CH, in sediment were made by extruding sediment from a
gravity core within a few minutes of collection, slicing a section of sediment
and quickly sealing it in a glass kettle (volume -\/ 0.3 liters) along with
130
-------
3
100 cm distilled water. The resulting slurry was equilibrated with the
gas phase and the concentration of CH in the gas was determined by gas
chromatography on a column of molecular sieve #5A with a flame ionization
detector. Knowing the PCR within the kettle, the air volume, the sediment
volume sampled and the 4 temperature, [CH,] per sediment volume can be
calculated to + 10%. This is the parameter required for calculating the
diffusive flux of CH,. Crude porosity measurements were made and these permit
calculation of [CH.] to +_ 20% in interstitial waters, the parameter necessary
to determine the degree of in situ saturation. The kettle technique seems
quite reliable for preventing gas loss as shown by measurement of Rn-222/Ra-226
activities on the same samples (Hammond, 1975). One core (P. . ^5 atm)
was apparently saturated with CH. jin situ and formed gas — Sltu
pockets when retrieved, causing ^ 30% loss of Rn-222. The CH measurement
here was corrected for this. Oxidation of methane during kettle storage
(usually < 24 hours) was, apparently not a problem since a number of in situ
values were observed.
RESULTS
10 3
A total of approximately 2.0 x 10 m of fresh water reaches the New York
Bight each year by passing through lower New York Bay. About 85% of this
enters the system via the Hudson River (mean annual flow ^ 550 m /sec) and
^ 15% is the combined inputs of the Raritan, Passaic and Hackensack Rivers.
The average monthly freshwater flow varies by an order of magnitude between
spring (high flow) and late summer (low flow), causing the distance of salt
penetration (> 0.1°/00) to range between 'v mp 25 and 'v mp 65. Large volumes
(^ 85 m /sec) of treated and untreated sewage from New York City and northern
New Jersey (DeFalco, 1965) are introduced to the harbor complex. These
discharges contribute more than 20% of the fresh water input during low flow
conditions. The large sewage loading reduces summertime dissolved 0- levels
to 25-50% of saturation in the Upper Bay and has delivered large amounts of
reduced carbon to the sediments. It is important to note that the Hudson
Highlands and Tappan Zee reaches of the estuary have quite distinct morphol-
ogies. The mean depths are 12.8 m and 5.3 m respectively.
A number of synoptic sets of CH, samples were collected along the axis of
the Hudson Estuary at different times of the year. Two typical transects
(with surface and deep values averaged) are plotted in Figure 47. Their
essential features are: CH, is greater in winter than summer; CH is greater
in reaches of the estuary with deeper water than those with shallow water;
and CH, rises dramatically in the zone of major sewage loading.
Figure 48 illustrates the variation of CH, with depth at individual
sampling stations. Fresher water in the upper half of the water column, low in
CH., enters the northern end of the harbor complex, while more saline water,
high in CH,, flows upstream in the lower half of the water column. At the
southern end of the harbor, lower salinity surface water leaving the estuary
is high in CH,, while saline water low in CH, is entering near the bottom. In
a more pristine region of the estuary, the Hudson Highlands, variations with
depth are less consistent. Methane sometimes increases with depth (mp 57,
mp 49) but may also have mid-depth minima (mp 53) or maxima (mp 46). Samples
collected in winter months show much less vertical variation, but the best
131
-------
HUDSON ESTUARY METHANE DISTRIBUTION
1.2
1.0
0.8
1.0.6
o 0.4
0.2
SEWAGE INPUT
Mill
HUDSON HIGHLANDS TAPPAN ZEE
mean depth=l3m mean depth-Sm
x'\
A A A-
MARCH 2-3,1974
T «Q°C
AUGUST 16-18,1973
T~25°C
i i i i i i
/
/ '
\ ~
I i I
I
I
I
72 64 56 48 40 32 24 16 8
MILES NORTH OF THE BATTERY
0
-8 -16
Figure 47. Methane and salinity versus depth in the Hudson.
-------
METHANE AND SALINTY VS. DEPTH
a)Harbor Complex (March 2J974)
IA mp-2 mp-5
mp '^ M mP ' M (UPPER BAY) (NARROWS)
MM /"-M MM uM
0.5 1.0 05 1.0 0.5 To 0.5 To
1 5
X
o! io
UJ
Q
15
It 1 1 1 i
'. °
I \
',
~'v ON
v* x
\ N
\ N^
"~ * C O
i O
rm
—
,,,,,. ,, , , , ,
> » i i
\ i i '
*• i i i
- -. \~z~) *
\ ' \ S '
\K\ ! \ 0
V / x '
5 / \ i
, _ > '
/
0 °
~ tin " 1 1 1 1
1 ' '0 ' '
/
\ /
"*sO
/A* js~i
f 1
0 ',
- 7777
0 20 0 20 0 20 0 20
%o c/eo %o '<••
b)Hudson Highlands (August 7, 1974)
mp 57 ., mp 53 .. mp 49 .. mp 46 M
u.M itM uM uM
^^
E
^•^
X
1- 20
a.
UJ
Q
30
O.I 0.2 „ 0. _ 0.2 JDJ 0.2 0.3
Y i i i v I '
i i
1 9
f 9
\^/ l^
• s s
^ S°
i
/////
-
0 1 2
%0
" T^WI I TA^— I 1 I |
! \ o
i \ i
O - i'
i xp
f \ ^
i \ Vj
- x o - \*t **
i . Xs N""O
\ I
- wfc-j O
S >
f
\ /
~ ' O
I r
i |
X O
I \
™ . « \
^s b
O ^*
— 1 /
i f
\ /
- ¥
j
/ II!
f 1 It
- x o 6 i 2 12
%0 %0
ii in
1 1 I
0 1 2
Figure 48. Methane versus mile point in the Hudson:
samples were averaged.
Surface and deep
133
-------
generalization that can be made is that CH, is uniform in the Highlands
within + 30% (one standard deviation).
Hudson estuary sediments have been discussed by McCrone (1967), Gross
et_ ad• (1971) and Olsen at ad. (1978). The predominant sediment type is a
silty clay with ^ 3-8% organic matter. Sandy sediments filled with shell
fragments cover portions of the navigation channel, some sections of which
are periodically dredged, and the sand fraction generally increases from the
Tappan Zee to the very sandy sediments of the Lower Bay and New York Bight,
with substantial areas of fine-grained sediments, accumulating rapidly in
certain zones of the Upper Bay and between Manhattan and New Jersey.
Methane concentrations within surficial sediments show large variations,
as abundances observed range over 4 orders of magnitude (Table 23). The upper
limit seems to be controlled by saturation of CH, in interstitial waters. The
ratio of observed concentration to in situ saturation (based on solubility data
from Atkinson and Richards, 1967) is listed as S.F. in Table 23.
Clearly, the distribution of methane-saturated sediments is more
heterogeneous than general sediment lithology variations. Surface (< 10 cm)
sediments may be saturated with CH,, but more commonly they are not. Cores
from mp 49 and mp 37 evolved gas bubbles from ^ 30 cm, although their surface
sediments contained less than 10% of saturation. There is a trend toward
greater incidence of surface saturation in the more northerly cores, but
insufficient data is available to say this with much confidence. It is
reasonable that this may be true, however, because less sea salt reaches these
regions and thus sulfide inhibition of methane-producing bacteria in surficial
sediments is less likely. The saturated sediments at mp 9 are located within
^ 1 km of a major input of raw sewage and also in a zone of very rapid sediment
accumulation. Methane distributions in Hudson sediments can be summarized by
saying that interstitial waters of upper sediments at some locations are
saturated in situ, but these are probably isolated pockets. All locations
are probably underlain with CH,-saturated material at some depth. The depth
of this saturation is probably a complicated function of the amount and type
of organic material present and the average annual salinity.
Three additional observations indicate the widespread nature of CH -
saturated sediments in the Hudson estuary. Low seismic velocities reported
by Worzel and Drake (1959) at mp 25 were attributed to gas-saturated sediments.
Gas pockets have been observed to form at depth in several cores other than
those described here (C. Olsen, personal communication). The authors have
noted bubbles bursting at the river surface at various locations in the Tappan
Zee and Hudson Highlands.
DISCUSSION
The distribution of methane in Hudson Estuary waters must be controlled
by a balance between inputs and outputs. By treating parts of the lower
estuary as well-mixed reservoirs, the importance of methane as a component of
carbon cycling in Hudson sediments can be assessed. To develop these concepts,
a budget for methane in the water column of the Hudson Estuary has been
derived, which includes contributions of advective transport of methane along
134
-------
TABLE 23
Methane Distribution in Hudson Estuary Sediments
mp
76
53
49
41
41
37
34
25W
25
18W
18
13W
9
7
1
-2
-5
Depth
(m)
16
45
28
24
15.
9
10
5
5
1
13
8
15
15
15
19
14
(Interstitial
Coll.
Date
4/6/74
3/2/74
8/6/74
5/18/74
4/6/74
8/8/74
4/6/74
5/18/74
3/2/74
8/27/74
5/18/74
1/8/74
1/18/74
5/18/74
3/2/74
5/18/74
3/2/74
water [CH
Interval
(cm)
2-10
10-18
2-10?'
h
10-18
2-6
2-6
7-9
2-10
10-18
2-6
2-10
10-18
2-6
2-12
12-22
1-6
13-15
22-24
2-6C
14-16
2-7
7-12
2-12
12-24
2-6
7-7.5
6-18'
2-6
2-10°
])/(in situ
[CH4]
(in sed.)
3480
3500
7830
6550
178
20.8
1350
2260
123
19.6
28.6
0.4
4.7
6.0
21
504
543
3.3
3..1
6.0
, 2390
1500'
6.1
74.
12.7
1.8
saturation)
ymol/1
(in IW)
5520
6360
12400
11900
274
32
2140
4520
190
31
52
0.6
7.4
11
32
840
987
5.1
4.8
9.9
3800
2720
9.4
123
19.6
2.9
(SF)a
.99
1.14
.92
.88
.05
<.01
.40
.85
.07
<.01
.01
<.01
<.01
<.01
.02
.55
.64
<.01
<.01
<.01
.93
.67
<.01
.02
<.01
<.01
, o o o
Corrected value. Rn measurement
CSandy sediment with shell debris
on this sample indicated 32% gas loss
135
-------
the estuary axis, evasion to the atmosphere, diffusion from the sediments,
and the partial dissolution of bubbles created by sediment flatulence.
Oxidation of methane in the water column is neglected since the storage tests
indicated this process is much slower than evasion to the atmosphere. The
reservoirs discussed are the Hudson Highlands (mp 56 to mp 40), the Tappan Zee
(mp 40 to mp 20) and the Harbor complex (mp 20 to mp -5).
Advective Transport
It has been shown (Hammond, 1975) that for an estuary such as the Hudson,
whose circulation can be approximated in terms of a two-layer advective model,
the flux of a property C past any location is equal to the freshwater flux (Q )
past that location, multiplied by the zero salinity intercept of a tangent on
a plot of C vs. salinity at the location of interest. This approach is also
valid for estuaries in which one-dimensional advection-diffusion models may
be used (Boyle _et_ _al_. , 1974). If this is done at two different locations, the
change in intercept, AC, can be used to calculate the net addition or loss of
C between the two locations, equal to Q x AC. A plot of methane vs. salinity
for March 1974 is shown in Figure 48A. Plots like this were used to calculate
advective fluxes.
Evasion to the Atmosphere
Use of the stagnant-film model (Lewis and Whitman, 1924) for estimating
the rate of exchange of gases across the air-water interface has met with
considerable success in natural water systems (Broecker and Peng, 1974). This
model envisions the rate of mass transport to be limited by molecular diffusion
through a thin layer of water at the interface. The thickness of this film can
be considered as a convenient parameter for characterizing the rate of gas
exchange of an open water surface as a function of environmental variables.
In marine systems, the most important forcing function on the rate of gas
exchange is apparently wind velocity. Emerson (1975) has proposed an empirical
calibration curve for wind speed and film thickness based primarily on exchange
rates of Rn-222 in the ocean and in fresh water lakes. He notes that to use
such a curve, observations of wind speed must be made at a fixed height above
the air-water interface. Field data on which his curve is based are reproduced
in Figure 48B with wind velocities taken to be those 10 m above the interface.
The mean wind speeds (monthly average of observations at LaGuardia, Kennedy
and Newark Airports) during 1973 and 1974 ranged from 4.1 to 6.3 ms . This is
equivalent to a film thickness of 110-50 y. A previous estimate of 180 + 60 y
based on wind speed at the Central Park Observatory (Hammond et al., 1975)
must be too low. This film thickness would not permit evasion of methane from
the Tappan Zee to keep pace with advective influx (primarily from the Highlands)
during winter months, and consequently it must be an over-estimate. Wind and
measurements at the Central Park Observatory are typically 2 ms lower than
the regional average from other sites, possibly due to a sheltering effect
from surrounding buildings, or to the general decrease in surface wind
velocities usually observed in large cities.
The mean residence time of a methane molecule in a well-mixed column of
water (depth h) before evasion will be hz where z is film thickness and D is
D
136
-------
METHANE VS. SALINITY IN LOWER HUDSON
1.4
1.2
=L 1.0
LJ
h-
UJ
March 2-3, 1974
mp I
-2
mp-8
8 12 16 20 24 28 32
SALINITY(%o)
Figure 48a. Methane concentrations in yM/1 vs. salinity in Lower Hudson Estuary, March 2-3, 1974,
line indicate depth distribution at each sample station.
-------
FILM THICKNESS vs. WIND SPEED
mph
4 6 8 10 20 30 40
1000
100
to
i r
RANGE FOR
HUDSON
ESTUARY
O L.227 EMERSON 8
BROECKER (1973)
® N. ATLANTIC BROECKER 8 PENG (1971)
® N. PACIFIC PENG ET. AL. (1974)
j j 1
I
I
I
10
15 _20
5
m sec"1
figure 48b. Surface film thickness vs. wind speed as determined by gas
exchange experiments on oceans and lakes
0.6r
0.4
ec
LJ_
,-> 0.2
o>
o
0.0
r ' I ' i ' T
® FRESH WATER.TSIVOGLOU, 1972
O FRESH WATER (Z=100LL)]
• " " (2=250LL)\ PENG, 1973
£ SEA WATER (Z=IOOJLL) X
(2=300^7
O
BEST FIT
SLOPE = 0.74
0.0
0.6
0.2 0.4
iog(Di/DRN)
Figure 48c. The mass transfer (flux, i/flux, radon) vs. diffusion
(diffusion, i/diffusion, radon) coefficients for various
gases in several systems
138
-------
molecular diffusivity of the molecule. In Tappan Zee waters this time is
about three days during both summer and winter, since the temperature
dependence of D is nearly compensated by higher winds in the winter. A film
thickness for each sampling period was derived by averaging wind speed for a
period beginning three days prior to sampling. The evasive flux of gas is
then a function of the difference in concentrations between the bulk solution
(C ) and film surface (C , assumed to be in equilibrium with the atmosphere) :
Flux = D
-5 2-1
The average flux, as^umjng Cy = 0.2 uM, D = 2 x 10 cm s , and z = 100 y,
is 4 x 10 y mol m s
An alternative conceptual model for gas exchange in rivers and estuaries
has been exploited by O'Connor and Dobbins (1958) based on the film replacement
model of Higbie (1935). In this formulation, the exchange rate is limited by
the replacement time (r) of a surface film through which molecular diffusion
occurs and
1 /?
Flux = (D/r)X// (Cw - Ceq) (2)
Each model is characterized by a mass-transfer coefficient, D/z or
(D/r)l/2, and a laboratory comparison of the fluxes of gases with differing
diffusivity should test the relative superiority of the two models. Figure
48C is a plot of log (Flux. /Flux ) vs. D for experiments performed by Peng
(1973) and Tsivoglou (1972J during which a tank of water was agitated and
fluxes of different gases were monitored. The stagnant-film model predicts a
slope of 1.0 and the surface renewal model predicts a slope of 0.5. The best
fit line to the two sets of data has an intermediate slope of 0.74. The
O'Connor-Dobbins model has previously been applied to the Hudson by O'Connor
(1970) and Quirk et_ ai_. (1970), assuming that all turbulence is flow- induced.
If the mass transfer coefficients calculated by these authors are equated
to the stagnant-film model, a film thickness of z = 320 y or 230 y respectively,
is predicted. This yields an exchange rate which is clearly too low, as
pointed out earlier. This suggests that the empirical relationship between
film thickness and wind speed is more accurate here.
Diffusion from Sediments
A first-order material balance calculation provides some important
information about the effectiveness of diffusion in supplying methane to the
overlying water column. Assuming that methane is unreactive in the upper 10 cm
of sediment, a linear gradient should exist there, and if the data in Table 23
are representative of the estuary sediments as a whole, the diffusive flux is
calculated to be 1 x 10~ mol m s . This represents a minimum flux (if CH,
is unreactive) because the diffusional gradient in methane-saturated sediments
is probably greater than we assumed, and stirring of surficial sediments by
currents and organisms has been neglected. This flux is about three times
larger than the flux leaving the estuary by evasion alone, between mp 20 and
mp 60^ and consequently must be an over estimate. With so few sample locations
it is possible that a representative suite of samples has not been collected,
but in light of what is known about methane biogeochemistry, it seems much more
likely that the major fraction of methane which diffuses into the upper layer
of sediments (^ 0 to 10 cm thick) is oxidized there by microorganisms before
139
-------
it can diffuse into overlying waters. It is postulated that this layer forms
an effective barrier to methane diffusion into overlying waters, certainly in
saturated sediments and perhaps in all Hudson sediments. Since the seasonal
and regional distribution patterns of methane can be accounted for in the
absence of diffusion, we assume diffusion inputs to be small.
Bubble Transport
Diffusion is probably not the major mechanism supplying CH. to the
estuary waters. The seasonal and depth characteristics of concentration in
the water column can be most simply explained by another process. Bubbles can
form in sediments if the total pressure of dissolved gases reaches in situ
pressure. If a sufficient volume of gas accumulates, confinement is no longer
possible and bubbles will migrate vertically through the sediment and escape.
Martens (1976) has suggested that bubbles may follow tubes created by burrowing
organisms. The escape of gas will result in a salvo of bubbles and the process
may be termed sediment flatulence. The impact of this process in estuarine
sediments has been described by Reeburgh (1969) and Martens and Berner (1974),
but a quantitative estimate of its importance is not, to the author's knowledge,
available. As the bubble of gas rises through the water column a portion of it
will dissolve, resulting in a transfer of methane from the site of its produc-
tion to the overlying waters.
The problem of diffusion into a water column from an ascending bubble has
been dealt with by a number of authors. The treatment presented here,
summarized in Table 24, is a modification of discussions presented by Levich
(1962), LeBlond (1969) and Guinasso and Schink (1973). The mass flux entering
the water from a gas bubble is given in equation 4 (Table 24). Bubbles of
radius r > 0.3 cm are significantly non-spherical (Levich, 1962), but for
simplicity, spherical geometry can be assumed. Neglecting surface tension
(for r > 0.01 cm) the pressure within the bubble is given in equation 5. The
rise velocity (v = — < 0) depends on water viscosity, bubble size and shape.
Levich (1962) states that if r < 0.01 cm, v will be given by Stokes Law and if
r > 1.5 cm, bubbles tend to become unstable, breaking into smaller bubbles.
By analogy to Cape Lookout (Martens, 1976), bubbles are primarily methane and
are likely to be within 0.05 cm < r < 1.5 cm. According to data from Turner
(1963) cited by Guinasso and Schink (1973), this size class of bubble has v =
-27 cm/sec + 30% at 20°C.
The input from one bubble rising through a water column of depth h is
given by equation 6. If the gas is pure, from spherical geometry and the ideal
gas law, equation 7 follows directly. Assuming that only a small fraction of
the bubble dissolves (m = constant) and that C » C (recalling that C
aP), equation 8 is obtained. Integrating equation 8 yields the input of ^
methane from one bubble (equation 9, Table 24.
Equation 9 has two interesting functional characteristics: k/v depends
only slightly on temperature since the viscosity effects on D and v should
nearly cancel. Thus the temperature dependence of I should be close to that
of solubility. If flatulence is uniform seasonally (at N bubbles sec ), and
the concentration is controlled by a balance between bubbles and evasion (as
is nearly true north of mp 20)
140
-------
TABLE 24
Equations for Dissolution of a Rising Bubble
r = kA (c - c )
dt eq w
k = mass transfer coefficient (--)
= bubble boundary layer thickness
A = bubble surface area
P = P + aZ (5)
P = pressure at top of water column (P = 1 atm. when Z = 0)
Z = vertical distance, positive down
a = — — = 10 atm cm
4 Pr3
m -
R = ideal gas constant
T = absolute temperature
r = initial bubble radius
z
2/3
2/3
4 k .SmRT. . , oo/o
where K = — (— ) = 4 k r 2p 2/3
V Z Z
v = rise velocity (< 0)
z
a
141
-------
TABLE 25
Environmental Conditions and Methane Averages
(viM)
Sample Date
Aug. 16-18, 1973
Aug. 24, 1973
March 2-3, 1974
April 6-7, 1974
Aug. 7, 1974
Aug. 9, 1974
Aug. 27-28, 1974
Oct. 2-4, 1974
Temp.
24-27
24-27
0-3
4-7
24-27
24-27
24-27
14-17
Wind
Speed
(ms )
3.6
4.4
4.9
6.0
4.4
3.6
4.0
5.3
. 3 -1.
(m s )
280
280
950
1100
280
280
200
520
Hudson
Highlands
(mp 40-60)
0.20 (12)
0.59 (6)
0.28 (6)
0.15 (22)
0.32 (7)
0.10 (5)
Tappan
Zee Harbor
(mp 20-40) (mp -2-3)
0.11
0.13
0.27
0.17
0.22
0.21
0.13
(12) 0.61 (2)
(7)
(4) 0.93 (6)
(6) 1.28 (6)
(4)
(8) 1.07 (6)
(4) 0.73 (3)
TZ/HH
.55
.46
.61
.69
1.3
Predicted Ratio (h = 12.8 m, h = 5.3 m)
HH ±/i
* ( ) is the number of samples collected
-------
- Cw = ™
z W
and since 1} is also nearly constant,
z
INz
W D
variations in concentration with temperature should be similar to those in
solubility. This is approximately true (Table 25). Although Emery (1969) and
Martens (1976) have observed seasonal variations in flatulent emissions (low
at low temperature), we see no strong evidence for this. The reason may be
that our bubbles originate at considerable depth in the sediment and do not
"see" large temperature variations.
If flatulence is uniform areally, the model predicts high concentrations
in reaches with deep water_._ If the mass flux is uniform and r is constant, N
will be proportional to P . Thus the ratio of methane concentration in the
Tappan Zee to that in the Highlands should be
-
Sz (IM)TZ (pz WTZ „ ,,
S.H - ««HH = (PZ-1» VVffi] = '
This is slightly lower than the observed ratio, but we have ignored the
advective contribution from the Highlands to the Tappan Zee, which amounts to
about 30% of the total budget there during high flow.
It is clear that the bubble model comes close to accurately predicing the
observed seasonal and regional variations. It should be noted that alternative
models could be constructed. The seasonal variations could be explained by
assuming there is a reservoir of methane in the sediments which continuously
supplies an upward flux of methane and that the oxidizing bacteria in the
sediments are much less active at low temperatures. The observant reader will
have noted that all saturated surficial sediments were collected in cool months
rather than warm months. Thus, flatulence could be the controlling mechanism
in warm months and diffusion in cold months. Regional variations could be
explained by arguing that the Highlands contain a greater proportion of marsh
areas than does the Tappan Zee, that marshes are areas of high sedimentation
rates, and are consequently major sources of methane. We have measured
dissolved methane in one marsh, and while it indicates the marsh area to be
a somewhat more effective methane source than the adjacent river, marshes
cannot supply the major fraction of methane inputs. We feel that although
alternatives could be proposed, the assumptions required would be less general
and the quantitative description of methane flux would require introduction
of several adjustable parameters.
Budgets
The procedures described above were used to calculate methane input to
the water column. These calculations include advective, evasive and flatulent
components. Oxidation in the water column is assumed to be negligible and
diffusive transport across the interface is also assumed to be negligible.
Turbation is also neglected because of the apparent effectiveness of bacteria
143
-------
in oxidizing methane as it diffuses upward into surficial sediments. Methane
dissolved in sewage effluent accounts for 10% of the total input to the harbor
and this correction was included in the calculations. By assuming values for
6 and r , the flux of methane escaping from sediments can be calculated. These
resultsZare listed in table 26A. Several experimental schemes have been
employed to find 6, but the problem is complicated because 6 seems to depend
on the "age" of the bubble, at least for the first few seconds (Deindoerfer
and Humphrey, 1961). Since the rise time through an 8 m water column should
be about 25 sec, the cleverly-devised experiments of Wyman et al. (1952) should
yield the best value. They showed 6 to be -^ 20 y and independent of bubble
radius or temperature.
The escape of flatulence probably requires a certain buoyancy. Thus, it
is likely that the spectrum of bubbles leaving sediments is fairly uniform in
size. From visual observations we estimate r =0.5 cm. This will certainly
be correct within a factor of 3. Martens (1976) has estimated 0.3 < r < 1.0,
depending on mean depth and temperature. Between 4 and 16% of a bubble with
r = 0.5 cm will dissolve as it rises, justifying the earlier assumption about
a negligible loss of mass. Systematic errors introduced in Table 26 will be
proportional to the systematic error in assuming an effective r . Note that
the mass flux error which would be introduced by counting bubbles bursting at
the surface is three times the percentage error in r. Clearly, we need a
method to accurately determine bubble size spectrums.
The rate of flatulence can be compared to the rate of preservation of
organic carbon (Table 26B). The latter was estimated on the basis of regional
subsidence rates north of mp 20 and on the basis of dredging rates in the
harbor complex. Flatulence and sedimentation rates are probably best con-
strained in the harbor complex, indicating that approximately 10% of the
organic carbon which is buried escapes as flatulence. The fraction calculated
for regions to the north is somewhat greater (^ 30%), but is reasonably
consistent when the various uncertainties areconsidered. The diffusive flux
calculated earlier is equivalent to 0.3 mol m yr, suggesting that the major
fraction of methane which is produced escapes as flatulence.
If the rate of CO,., production is as great as the rate of CH, production,
we expect to find a decrease of about 20% in organic carbon with depth in the
sediment. Evidence for decreases of 30% or more have been observed in a few
locations, but are not consistent (C. Olsen, personal communication).
Decreases may be masked by turbation or may occur at depths below core
penetration. Major questions which remain to be determined are identifying
the depth at which methane is produced and the depth at which bubbles
originate.
_3
The rate of flatulence required by this model is approximately 10
bubbles m s . Emery (1969) has observed a rate about 15 times greater in a
small coastal pond. Our rate is equivalent to 1 bubble min in a 20 m area.
Since bubbles probably come in salvos of 10 or more, a visual observation to
determine the rate would require survey of a large area for a long time,
making such an observation in the Hudson impractical.
144
-------
A)
TABLE 26
Comparison of Flatulant Production Rate and Burial
Rate of Organic Carbon
a —2 —T 8
Flatulent Production (mol m sec x 10 )
Highlands Tappan Zee
Aug. 16-18, 1973 3.6 3.7
Aug. 24, 1973 6.5
March 2-3, 1974 4.9 2.2
April 6-7, 1974 3.5 3.4
Aug. 7, 1974 3.6
Aug. 9, 1974 5.3 7.3
Aug. 27-28, 1974 9.0
Oct. 2-4, 1974 1.6 4.9
— 2 — 1 R
Ave. (mol m s x 10 ) 3.8+1.3 5.3+2.4
(mol m~ yr"1) 1.2+0.4 1.6+0.7
B) Burial Rate of Organic Carbon
O V-i
Sed. rate (gm cm yr) 0.2+0.1
Dry wt. loss on ignition (%) 6
Carbon compounds (g mol ) 30
Preservation of organic carbon
(mol~ yr ) 4+2
Harbor
22
43
41
24
33+11
10+3
3.5+1.0°
8
30
90+30
Calculated as outlined in text from (evasion rate - advective rate)/
(area x f)
Assuming long-term sedimentation rate is equal to the subsidence rate
(Fairbridge and Newman, 1968)
CBased on the rate of dredging in the Harbor Complex (averaged over the
area from mp -8 to mp 20). The lower figure is derived from Panuzio's
(1963) statistics for 1926-1960 and the upper figure is Gross (1972)
estimate for 1964-1968
145
-------
CONCLUSIONS
The factors which primarily control the distribution of methane in the
waters of the Hudson Estuary are input by flatulence, advection and evasion
to the atmosphere. Oxidation by bacteria in surficial sediments probably
prevents significant amounts of methane from diffusing or being turbated into
overlying waters.
Using wind speed data, evasion rates are estimated to occur at a rate
equivalent to a film thickness of 100 y and a budget for methane in the water
column can be calculated. The budget requires an input from the sediments.
A model accounting for this input by partial dissolution of sediment flat-
ulence indicates that a bubble leaving the bottom, with an initial radius
of 0.5 cm, will lose 4-16% of its methane to the water column as it rises,
depending on temperature and depth. This model is used to calculate the
production of gas bubbles in sediments. The rate of methane escape is
equivalent to about 10% of the burial rate of organic carbon. We believe
that this calculation is accurate within a factor of three, indicating that
methane bacteria ultimately recycle a significant fraction of the organic
carbon buried in these sediments.
146
-------
SECTION 11
NUTRIENTS AND TRANSPORT MODELS IN THE HUDSON
NUTRIENTS IN URBAN ESTUARIES
In recent years substantial increases have occurred in algal populations
in lakes including large systems such as Lake Erie. These increases are
believed to result primarily from the loading of additional plant nutrient
elements, phosphorus, and nitrogen. When excess algal growth can be directly
related to man's activities, the changes are often ascribed to a loosely
defined process called "cultural eutrophication". In most cases, the most
practical management tool for decreasing algal standing crops in lakes is to
reduce the input of phosphate. As a result, the general policy direction for
nutrient management in several European countries and North American is to
decrease the phosphate levels in detergents, to construct tertiary treatment
facilities for phosphate removal from sewage whenever possible, and/or to
divert sewage discharge from lakes if an acceptable alternative is available.
Excess algal growth problems have also developed in some large estuarine
systems including the Potomac downstream of Washington, D.C., and removal of
both phosphorus and nitrogen from sewage effluent has been proposed in several
cases. Because of the costs of tertiary sewage treatment facilities and
problems of sludge disposal from such facilities, it is important to examine
critically the benefits of nutrient removal for estuaries on a case-by-case
basis.
Mathematical representation of algal populations, nutrients, and even
higher trophic levels such as zooplankton and fish, can be constructed in a
similar way to those generated to describe the balance of dissolved oxygen in
estuaries. The predictive capability of such models is sometimes not clear,
since the response of phytoplankton communities, as well as higher trophic
levels, to environmental parameters is more complex than for bacteria in
their role as oxygen consumers. In light of large uncertainties in critical
biological parameters, management alternatives can sometimes be reasonably
examined using extremely simplified conceptual models of water circulation
and nutrient-algae interactions. The approach described here (see also
Simpson et al., 1975) is to examine the distribution of phosphate in the
Hudson estuary and the rate of loading from sewage outfalls, in terms of a
very simplified description of water circulation and phosphate behavior in the
harbor region adjacent to New York City. Phosphate is the nutrient most
frequently considered for nutrient removal from sewage effluent and is
somewhat simpler to treat in estuarine budgets than nitrogen.
147
-------
HUDSON RIVER ESTUARY SEWAGE LOADING
HUDSON DRAINAGE BASIN
4I°30 •
40°30-
73°30
Figure 49. Location map for major sewage outfalls (> 10 million gallons per
day) to the Hudson Estuary. Total discharges shown are '^ 100 m
sec (^2.3 billion gallons per day). Areas of the circles are
proportional to the volume of sewage outfall at that discharge
site. Square in the New York Bight indicates zone of sewage
sludge and dredge spoil dumping.
148
-------
SEWAGE AND PHOSPHATE IN THE HUDSON ESTUARY
The Hudson estuary, portions of which are entirely surrounded by the New
York City metropolitan area, is an example of an estuarine system heavily
loaded with sewage and other discharges which substantially alter the ambient
water quality. Dissolved oxygen levels of less than 40% of saturation
with atmospheric oxygen concentrations are not uncommon during summer months.
Sewage is discharged to the Hudson Estuary near New York City from about a
dozen major treatment plant outfalls and numerous smaller plants with a
combined flow of approximately 80 m /sec O 1.8 billion gallons/day) plus a
number of ^aw sewage outfalls, largely from Manhattan Island, which total
about 20 m /sec O 0.5 billion gallons/day). Approximate locations of the
largest outfalls are shown in Figure 49. Of the major treatment plant outfalls
about 2/3 of the total volume is discharged from seconary treatment plants
operated by New York City, while the remainder is supplied by primary
treatment plants mostly from New Jersey.
The major secondary treatment plants in New York City, ranging in age
from a few years to about 40 years, are reasonably efficient in reducing
biological oxygen demand of sewage. Except for storm runoff periods when
treatment plants are essentially by-passed, about two-thirds of the oxygen
demand of the sewage is removed. Construction of major facilities is now
underway to treat most of the remaining raw discharges which currently
contribute on the order of 1/3 of the total dry weather biological oxygen
demand to the lower Hudson estuary. Primary treatment outfalls, mostly from
New Jersey, discharge about 1/3 of the total treated sewage flow and about 1/2
of the total oxygen demand. Thus in terms of the current oxygen budget, the
largest impact of sewage discharge is from partially treated New Jersey out-
falls. The oxygen demand from these sources will dominate to an even greater
extent when the current construction projects in New York City provide
secondary treatment for the raw discharge from Manhattan.
All of the major discharges of sewage to the lower Hudson estuary supply
large amounts of primary plant nutrients, especially ammonia and phosphate,
as well as organic materials which constitute the major load on the dissolved
oxygen resources of the water. The supply rate of nutrients is more directly
proportional to volume of sewage flow than is biological oxygen demand, since
present treatment operations '(raw, primary or secondary) do not have nearly
as great an effect on nutrients as they do on dissolved and suspended organic
carbon. It is important to establish to what extent and in what areas of the
estuarine-coastal water system these sewage-derived nutrients are converted to
organic matter by phytoplankton, and how this affects oxygen concentrations
and other water quality factors.
The concentrations of dissolved phosphate as well as other primary
nutrients in the harbor region of the Hudson estuary are approximately two
orders of magnitude higher than those usually considered as limiting to
phytoplankton growth. Phosphate concentrations do not vary greatly from one
location to another in the harbor, despite the presence of a number of large
discrete sources. The smoothness of observed phosphate concentrations
indicates the effectiveness of mixing by tidal currents and density-induced
non-tidal estuarine circulation. In general, the mean values of phosphate
in the Inner Harbor range between 2 um/1 during spring high fresh water runoff
149
-------
PHOSPHATE IN THE HUDSON RIVER ESTUARY
1
in
.+'**' 1
* ^_
J- + ^"*
* QJ
:'* t
to
°
+ + Q-
+ surface
+ August 16-18,1973 • bottom
- 6 -
- 5 -
- 4 -
- 3 -
- 2 -
.*.
'• ++ i
'
+
t
.
August 28, 1974
i i i i i i i i i i i i
10 20 30 10 20 30
t
1
, - **"+'' 1
, • «•
0
,+. a.
4-
} *
| March 2-3, 1973
- 2 -
- + '"
, +
~ 1 -t +
• •
April 6- 7, 1974
10 20
Salinity (%<>)•
30
10 20
Salinity (%o)-
30
Figure 50. Phosphate (molybdate reactive) concentrations versus salinity:
Fresh water1- fows were <_ 300 m3/sec for the August profiles,
450 m3/sec for October,- and ^ 1250 m3/sec for March and April.
150
-------
and 6 um/1 during summer low fresh water discharge. To illustrate the
general character of observed data several transects of phosphate and salinity
for the Hudson are shown in Figure 50.
Total phosphorus measurements, as well as particulate organic phosphorus
and dissolved organic phosphorus concentrations were made for most of the
transects shown in Figure 50. In general, about 1/2 to 2/3 of the total
phosphorus in the water was present as molybdate reactive phosphate, with
most of the remainder as particulate organic phosphorus. During high fresh
water flow (March and April) the fraction of particulate organic phosphorus
was somewhat greater. The data shown are for samples collected along the
axis of the Hudson, through the middle of the Inner Harbor, the Narrows, the
Lower Bay, and the apex of the New York Bight (Figure 49). Samples collected
in heavily loaded and more restricted zones of the Inner Harbor such as the
East River plot above the trends shown, sometimes by as much as a factor of
two in phosphate concentration, but the data shown are typical of most of the
harbor volume.
Upstream of New York harbor, phosphate and salinity decrease at
approximately the same rate until fresh water is reached, where the phosphate
levels remain nearly constant at 0.5 to 1.0 um/1, depending on the fresh water
flow rate. Composite plots of the data for several survey periods indicate
the general distribution of phosphate as a function of mile point (Figure 51)
and salinity (Figure 52) during both low and high fresh water flows. The
covariance of salinity and phosphate suggests that removal processes for
phosphate such as algal uptake do not dominate phosphate concentrations in
the salt intruded reach of the Hudson and that physical transport of the water
by estuarine circulation is most critical in establishing the phosphate
distribution.
The geometry of sewage loading is presented schematically in Figure 53.
Most of the sewage is discharged directly to the Inner Harbor, with less than
5% of the total added upstream of the Inner Harbor (segments A, B, C and D
of Figure 53). Saline water and sewage phosphate added to the Inner Harbor
are spread well upstream of the discharge sites by estuarine circulation, as
far upstream as segment D during low flow periods.
To illustrate the relative magnitudes of phosphate source terms to the
Inner Harbor, and provide some sense of the time scale of removal by estuarine
circulation, a one box model for phosphate in the Inner Harbor can be
constructed (Figure 54). The mean concentration of phosphate in the Inner
Harbor, multiplied by the volume of this portion of the estuary gives the
total amount of phosphate in solution. Inputs of phosphate are (1) direct
sewage discharge, (2) diffusion from the sediments, (3) downstream supply
from the fresh water reach of the Hudson, and (4) particulate oxidation in the
water column, with direct sewage discharge supply about 3/4 of the total.
Dividing the total standing crop of phosphate in the Inner Harbor by the
supply rate indicates a residence time for phosphate of 2 days during high
flow and 7 days during low flow.
A similar calculation can be done using a first order estimate of the
outflow rate of phosphate from the Inner Harbor. Removal of phosphate is
predominantly through the Narrows, into the Lower Bay and then to the adjacent
151
-------
1 4
«* ,
O 3
0.
PHOSPHATE DISTRIBUTION IN THE HUDSON ESTUARY
- A
. V
V
~ O
- ©
CD
®
_
.
-
— A
1 1
SURFACE, 8-16-73 LOW FLOW
ounrnuc. I T0 N
DEEP J 8-18-73 ) ^
SURFACE-i 00 ,„ \ ^ ^
DEEP I8'28-74 V X AA/A
SURFACE, A/A^A A
DEEP J- 3-2-74 /"A
SURFACE, /A ^
°EEP } V /$AA
A^ A
x^ HIGH FLOW $
/^ A 5~ SL-j5'c
Ax §-^" ""® @ (
AX/\ A A, _^ —'6? ™
.. gi —
A U$ ® _89 *^~ ™
0
i i . i i i A i i i , i i i i i i i
.
\ —
7 \
\ V7
\
\
\
\
& \
r\ o \
i ®v \
V \ '
ee \
0-
, i . i
72 64
56
16
8
0
-8
-16
48 40 32 24
MILE POINT
Figure 51. Phosphate (molybdate reactive) as a function of location for
several surveys in the Hudson Estuary during high flow (March
and April) and low flow (August).
'6.0
5.0
_J
\ 4.0
o
e
^1 ^ n
^*^. \j \j
O
PHOSPHATE vs.
A AUGUST1 73
- V AUGUST '74
0 MARCH '74
® APRIL '74 v
A
VA^V
\7 ^ ^A
A^/ A
r7 ^A/A
^
/\A A ~»
j^ 2.0A*" ® ® _J«
/A. ffiicy^ **** o
JfjAgr"" ~~~ """"
SALINITY IN THE HUDSON ESTUARY
V
yV ^/ \8 V
/ A VV
X^ A \
x^ A ^ ^ ~
AxA \
/ y LOW FLOW
.J —
\ ^^"^
-
HIGH FLOW vV
<". \
® \
O Or) ^^ V\ ~
Q\P \
O ° \ \
Q) \
6.0
5.0
4.0
3n
.u
2.0
1.0
I
J I
10 20
SALINITY %
30
Figure 52.
Phosphate (molybdate reactive) as a function of salinity for
several surveys in the Hudson Estuary. Data are from
Figure 51 plus samples from mid-depths.
152
-------
HUDSON RIVER ESTUARY
SCHEMATIC SEWAGE LOADING
< I mVsec I
10 mVsec
0 10 20
I I I I
Kilometers
ghl
0
10
Figure 53. Schematic loading of sewage to segments of ^ 30 km lengths
of the Hudson.
153
-------
PHOSPHATE BUDGET- INNER HARBOR
Input P04 (moles/sec)
Sewage
4.7
Sediments (0.4-0.8)
Particulate
Oxidation
(0.4-0.8)
Upstream P04 (moles/sec)
(0.9-0.4)
T(P04) = 2-7 days
P04 = 2-6 ^mole/I
6.4-6.7 moles/sec
Figure 54. Approximate fluxes and concentrations of phosphate in New York
harbor. The major cause of changes in residence time of
phosphate in the harbor is change in the rate of fresh water
discharge.
154
-------
coastal waters of the New York Bight. Some phosphate discharge does occur
through the Upper East River into western Long Island Sound. Based on
estimates of the magnitude of this transport, we have decreased the loading
rate of phosphate to the Inner Harbor used in Figure 54 by an equivalent
amount, about 10-15% of the discharge rate through the Narrows. Water flowing
out through the Narrows in the upper half of the water column has a mean
phosphate concentration similar to that in the Inner Harbor, while that
flowing upstream in the lower layer has a lower phosphate concentration and
a higher salinity. Using a two-layer advective transport calculation (Figure
55 - two box model of phosphate transport in the harbor area) based on field
salinity and phosphate concentrations during both high and low flow conditions,
the removal rate of phosphate is comparable to that estimated in Figure 54 for
the input rate. Thus to the first approximation, sewage phosphate is dis-
charged to the lower Hudson into waters with salinities averaging about 2/3
of sea water salinity, mixed reasonably well through the Inner Harbor over a
time period averaging between 4 and 14 tidal cycles, and discharged from the
system by estuarine circulation.
These greatly simplified representations of nutrient transport rates in
the Hudson can be expanded in the form of numerical models, and thus treated
in terms which are more conventially applied in water quality transport
calculations. The most common approach to one-dimensional descriptions of
estuary transport is to use observed salinities to establish model parameters
for the upstream transport of sea salt (and phosphate). Stommel (1953)
suggested the use of one-dimensional advection-diffusion segmented models
based on balancing downstream advection of fresh water with upstream diffusion
of salt. Thus the complex upstream transport processes of density-driven
circulation, tidal mixing and their interaction with channel morphology are
lumped into a single "diffusion" constant for each segment boundary.
We can describe the tidal Hudson in terms of equal length segments,
each of which are considered to be well mixed. The basis for the physical
dimensions of volume elements used here for the Hudson between mp 0 and the
head of the tide (about mp 154) is a report by Thatcher and Harleman (1972)
on a one-dimensional physical transport model for the Hudson. We took channel
dimensions reported in terms of a main "conveyance" channel and an adjacent
"storage" channel at approximately two mile intervals, and converted all
dimensions to metric units, combining the two-channel presentation into one
channel.
The geometry of the Hudson estuary downstream of mp 0 is quite complex
and not readily represented in one-dimensional terms, but we have chosen to do
so as a simple extension of the linear geometry of the Hudson upstream of mp 0
(Figure 56). A summary is given in Table 27 of the segment dimensions used to
approximate the Hudson as a chain of 6.4 km (4 mile) boxes extending from the
seaward end of the Narrows (mp -8) to the head of the tide north of Albany
(^ mp;154). The segments designiated A* and B* represent all of the harbor
upstream of the Narrows, including Newark Bay, Kill van Kull, the Lower East
River and Upper New York Bay, and thus are greatly simplified from the actual
geometry. All dimensions of segments 1-38 were derived from approximating
the main stem of the Hudson by rectangular cross-sections, based on dimensions
reported for linear distances of slightly greater than 3.2 km (2 miles)
(Thatcher and Harleman, 1972). We have rounded the segment lengths generated
155
-------
Hudson Estuary Phosphate Fluxes
(moles/second)
Sewage
1
Is
• r
1
1
1
E
Estuary
7*
^
_4
H
Harbor
6 ,
1 > \
1
_l
*•
-S
^ 6
Ocean
Figure 55. simple two box model of phosphate fluxes in the Hudson, with
discharge from the harbor expressed in terms of two layer
advective fluxes.
-------
Major Estuaries in the Northeastern U.S.
X' /
Narragansett
Bay
Hudson
350mVsec
Delaware
330 mVse
Ir
Lower
New York
Bay
Susquehanna
Q=950
Potomac
Q=340
Delaware
Bay
30
30
60
Scale in Kilometers
.Chesapeake
Bay
Figure 56. All of the major rivers flowing into the Atlantic between
Rhode Island and Virginia enter estuarine systems (mean annual
flows are shown for gauging stations above tidal influence).
157
-------
TABLE 27
Segment
No.
A*
B*
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
One-Dimensional Model Parameters for Tidal
Segment Lengths are 6.4 km (4 miles)
Mean Surface Mean
Width Area Depth
(km) (km ) (m)
2.00
3.50
1.33
1.25
1.35
1.47
1.48
2.20
3.64
3.78
4.13
3.06
1.27
0.73
0.73
1.30
1.99
1.54
1.14
0.87
0.78
0.74
0.93
1.30
1.50
1.31
12.8
22.4
8.6
8.1
8.7
9.5
9.5
14.2
23.4
24.3
26.6
19.7
8.2
4.7
4.7
8.4
12.8
9.9
7.3
5.6
5.0
4.8
6.0
8.4
9.7
8.4
16.8
11.6
11.7
11.1
9.8
8.9
9.2
6.8
5.3
5.5
5.1
6.3
12.7
18.9
18.7
10.8
8.0
9.9
10.5
12.3
14.1
12.8
11.7
9.0
7.9
7.1
Hudson
Segment
Volume n
(10* m3)
216
259
100
89
85
84
88
96
125
133
135
125
104
89
88
90
103
96
77
69
71
61
70
75
76
60
(continued)
158
-------
TABLE 27 Continued
Segment
No.
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Mean
Width
(km)
1.21
1.15
1.30
1.00
0.97
1.08
0.76
0.78
0.65
0.43
0.38
0.29
0.23
0.27
Surface
Area
(km )
7.8
7.4
8.4
6,4
6.3
7.0
4.9
5.0
4.2
2.8
2.5
1.9
1.5
1.7
Mean
Depth
(m)
6.4
5.6
4.3
4.7
4.6
4.3
4.7
4.2
5.7
6.3
6.9
6.4
4.5
3.6
Segment
Volume 0
(105 m3)
50
41
36
30
29
30
23
26
24
17
17
12
7
6
^Segments A and B are artificial in shape, and are designed to include
the following:
V (106 m3) A (km2) h (m)
Upper New York Bay 340 24 14
Lower East River 80 7 11
Newark Bay 34 11 3
Kill van Kull 21 37
The portion of the estuary seaward of the Narrows was excluded.
Approximate dimensions of Lower New York Bay, Raritan Bay and
Sandy Hook Bay are:
Estuary between
Narrows and Sandy
Hook-Rockaway
Transect
Q 3
V (10 m )
1.3
A (km )
250
h (m)
159
-------
in our compilation to distances of 6.4 (4 miles) and hence show an upstream
segment termination at mp 152, while the actual distance between the origin
of the location system at mp 0 and the head of tide is located about mp 154.
The mean depth of the entire estuary above the Narrows is slightly more than
8 meters.
The flushing characteristics of the lower estuary are difficult to
describe because of the complexity of physical geometry and current patterns.
We have chosen the Narrows as the logical seaward end of the estuary,- primarily
because the net transport past this location can be approximated reasonably
well by a two-layer advective model (Hammond, 1975). We have used phosphate
as an indicator of the rate of removal of pollutants from the harbor region
through the Narrows, because the rate of addition can be described reasonably
well (Hammond, 1975) and because its behavior to the first approximation
appears to be conservative within the lower estuary (Simpson ejt al^. , 1975).
Significant removal of sewage-derived nutrients from the harbor complex also
occurs through the Lower and Upper East River, into western Long Island
Sound, but this net transport appears to be substantially less than that
through the Narrows. Our best estimate at this time is that approximately
85% of the total phosphate discharge upstream of the Narrows, including the
discharge to the East River, leaves via the Narrows while the remaining 15%
leaves the system via the Long Island Sound.
Table 28 summarizes the field data for salinity and phosphate which we
have used to describe the flushing characteristics of the lower estuary.
The seaward boundary values of salinity and phosphate represent the
tidally-averaged characteristics of the upstream flowing water in the bottom
half of the water column in the Narrows.
We have made a series of calculations of the equilibrium phosphate
distribution in the Hudson with a one dimensional diffusion-advection numerical
model. The predicted phosphate concentrations of several of these calculations
are given in Figure 57. The calculations plotted are both low and high fresh
water discharge conditions (Figure 58) using a constant input rate of phosphate
from sewage of ^ 5 moles per second. In one case all of the sewage phosphate
is added to the second segment from the seaward end (B*) while in the other two
cases the phosphate is added to each of the segments in amounts approximating
those of the actual discharges (see Simpson et al., 1975 for a summary of
sewage outfall geometry in the Hudson).
The primary difference in the calculations based on a distributed sewage
phosphate source from those with a single aggregated source in segment B* is
a more uniform distribution of phosphate within the harbor (segments A*, B*,
1, 2 and 3) but the differences were not large (Figure 57). Several of the
calculations included additional inputs of phosphate from oxidation of sewage
particulates within the harbor and diffusion from harbor sediments. The sum
of these inputs was estimated to be 1/4 of the direct sewage discharge, which
is comparable to the uncertainty of the sewage discharge estimate. We have
not attempted to "tune" the calculations to provide a better fit to the
observed data, which are also subject to substantial uncertainty since no
really adequate time- and space-averaged data base exists. From the general
similarity of the calculation outputs and "observed" phosphate data (Figure
160
-------
TABLE 28
Segmented Model Observed Salinities
and Phosphate Concentrations
Segment
No.
Seaward
Boundary
A*
B*
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
High F
(1130 m
Salinity
V / o 0 '
22
19.7
16.4
13.8
11.2
8.9
6.5
4.5
2.8
1.4
0.6
0
low
/sec)
Phosphate
(Vim/1)
1.65
2.0
2.0
2.0
2.0
1.8
1.6
1.5
1.4
1.2
1.0
<1.0
Low F
(220 m
Salinity
* 00'
30
26.7
25.6
24.0
22.0
19.7
17.2
14.8
13.0
11.3
9.8
8.3
6.6
4.9
3.4
2.5
1.9
1.5
1.2
0.7
^0
low
/sec)
Phosphate
(ym/1)
3.0
5.5
5.5
5.5
5.5
5.5
5.2
4.6
4.2
4.0
3.6
3.4
3.2
2.9
2.6
2.2
2.1
2.0
1.9
1.7
<1.5
Fresh water flows for Segments A and B were slightly higher, due to the
addition of sewage from the New York City area and the inflow of the
Passaic and Hackensack Rivers. The totalsjj during high and low flow for
these segments were 1290 m /sec and 260 m /sec.
161
-------
4
« 3
D
_C
O.
CO
O
£ 2
A- Observed P04
n•
32 I B* A*
Seaward End
(mp -8)
Figure 57. Observed (A and A1) and computed (one dimensional diffusion-
advection model calculations using ^ 6 km, 4 mile, segments
assumed to be well-mixed) phosphate concentrations in the
Hudson Estuary. Curves B and B' assume all sewage added as
point source, C, C', D and D' assume a distributed source
approximately the actual pattern of discharge while C and C'
include additional phosphate input by diffusion from sediments
and particle oxidation in the water column. Primed letters
indicate high flow, unprimed indicate low fresh .water flow.
-------
HUDSON RIVER FRESH WATER FLOW
CO
500-
M/sec 300 - na
100
o
M/sec
1000
500
100
(Green Island gauging station)
25 year annual mean = 354
mean annual
flow
mean monthly
flow
61 62 63 64 65 66 67 68 69 70 71 72 73
Figure 58. Gauged flows at the head of tide in the Hudson (Green Island
mp +154) are approximately 2/3 of the fresh water flow at mp 0.
Mean monthly flows usually range over about a factor of ten.
-------
TABLE 29
Comparison of Phosphate Behavior in the
Lower Hudson Estuary and Lake Erie
Volume
Area
Mean Depth
Residence Time
Basin Population
E P Loading Rate
3-
Loading Rate
3-
PO
Predicted PO" (5)
Measured PO
Measured I P
Measured PO x 100%
Predicted PCL x 100%
Hudson
0.6 km3 (1)
2
60 km
10 m
2 days
10? (2)
6 mole/sec
2 ymole/£
2 ymole/£ (1)
2.7 yg- atom/£
100%
Lake Erie
500 km3
25,000 km2
20 m
1000 days
io7
26 mole/sec (3)
13 mole/sec (4)
2.5 ymole/£
.03-.06 ymole/£ (6)
0.6-1.2 yg-atom/£ (7)
1-25% (6)
(1) High flow conditions (see Figure 52)
(2) Only populations discharging above the Narrows and not to
the Upper East River is included (^ 65% of total in basin).
(3) See Vollenweider ejt aJU , 1974.
(4) Assuming PO = 50% of Z P loading rate.
(5) Assuming conservative behavior for PO
(6) Summer epilimnion near the lower value, winter epilimnion
near the upper value. See Dobson et al., 1974.
(7) See Dobson et al., 1974.
164
-------
57) it'is clear that very simple numerical models using observed salinities
and estimated phosphate inputs from sewage can approximate reasonably well
the gross features of the dissolved phosphate distribution within the salt
intruded reach of the Hudson (Table 30).
If the general features of nutrient behavior in the Hudson are
approximately correct as outlined here, then there are substantial implications
for phosphate management policies in the New York City area. The primary
purpose of decreasing phosphate discharge to receiving waters is to reduce
adverse effects resulting from excess algal growth in the receiving water.
The Hudson estuary upstream of the Narrows in the Inner Harbor has a large
surplus of nutrients for algal growth. The algal population within this
reach of the Hudson is surprisingly low. The maximum warm season standing
crops of algae are on the order of only 10% of those of Lake Erie, a system
that is generally accepted as having substantial areas that are clearly
eutrophic. During winter months, algal standing crops in the Hudson are
lower than summer months by more than an order of magnitude. Thus the Hudson
Estuary is not now characterized by nuisance growth of algae. In addition,
algal activity is not currently limited by available nutrients. Large
excesses of ammonia and nitrate as well as phosphate are present. Thus
reduction of the current discharge rate of phosphate by tertiary treatment
of sewage for phosphate removal probably would have little effect on the
activity of algae in this reach of the Hudson.
A first order budget for sewage-derived nitrogen in the Hudson has
been presented (Garside et al., 1976), which also indicates that dissolved
nitrogen and ammonia are removed from the Hudson estuary predominantly by
estuarine circulation, and not by phytoplankton uptake. There seems little
indication of nutrient limitation being significant for the Inner Harbor
area of the Hudson estuary.
The factors which currently limit phytoplankton standing crops in the
Hudson are not well defined, but growth rates have been clearly shown to be
light and temperature regulated (Malone, 1976a). The most reasonable factors
determining standing crops in the Hudson are light limitation due to the high
suspended solids content of the water and rapid removal of algal cells from
the estuary upstream of the 'Narrows by estuarine circulation. Neither of
these factors could be readily affected by changes in sewage treatment
processes since most of the suspended particulates are not supplied directly
from sewage outfalls. If a substantial reduction in suspended solid levels
during low flow summer months were to occur then algal growth rates would
increase, perhaps reaching nuisance levels. This possibility should be
carefully examined, but does not seem of primary concern for the next decade.
After leaving the Inner Harbor through the Narrows, the high nutrient
estuarine waters support increasing phytoplankton standing crops, especially
after reaching the adjacent coastal waters of the New York Bight. The
magnitude of this increased phytoplankton activity can be clearly seen in
higher chlorophyll contents in the New York Bight and even in the Lower Bay.
Growth rates of phytoplankton in the nearshore waters are primarily light and
temperature regulated and not nutrient limited (Malone, 1976b) and thus not
very sensitive to control by nutrient discharge control upstream within the
estuarine region.
165
-------
TABLE 30
Nutrient Data
M.P.
March
103
92
76
67
53
41
34
25
14
1
-2
-5
-8
-18
Depth
(m)
2-3, 1974
1
5
11
2
5
8
2
11
1
18
1
21
43
1
7
18
1
3
1
4
2
6
10
1
6
13
1
7
12
1
7
12
1
4
6
1
4
8
T
1.5
1.5
1.5
0.6
1.8
1.8
0.6
0.7
0.6
0.6
0.5
0.5
0.5
0.6
0.5
0.5
1.2
1.6
1.5
1.5
2.8
3.0
2.5.
3.8
4.1
2.8
3.6
4.2
3.6
4.1
5.0
4.1
4.5
4.9
4.7
4.8
5.4
Sal.
(° 1 1
\ loo)
O.,l
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.5
0.8
2.2
4.5
15.0
8.7
18.8
23.4
13.1
18.3
25.1
15.3
22.7
27.5
16.1
23.2
26.8
28.0
28.1
31.8
CH,
(ymol/D
0.24
0.34
0.62
0.64
0.62
0.65
0.63
0.66
0.64
0.512
0.545
0.546
0.260
0.276
0.273
0.344
0.386
0.60
1.19
1.08
1.14
0.91
0.96
0.88
0.64
0.86
0.66
0.374
0.805
0.548
0.436
0.162
0.155
0.135
PO,
(pmol/1)
0.59
0.52
1.03
1.06
0.85
0.77
0.56
0.77
0.85
0.64
0.70
0.85
1.18
1.11
1.57
1.88
1.67
2.10
1.60
1.83
2.00
1.65
1.95
1.41
1.18
1.77
1.51
1.20
0.59
0.56
0.70
Part
P
(ymol/1)
1.31
1.65
0.15
0.55
0.96
1.81
1.49
1.26
1.14
2.03
2.51
2.27
2.66
0.42
2.24
1.19
0.88
0.85
5.75
4.04
1.93
0.32
1.67
1.95
0.22
4.01
1.84
2.86
0.45
0.83
Excess
[Rn]
(dpm/1)
East River
1
9
4.0
4.1
17.8
19.0
1.03
0.90
2.42
2.90
3.84
2.36
(continued)
166
-------
Depth T
M.P. (m) (°C)
Kill Van Kull
1
11
Lower Bay
1 4.0
4 4.1
August 7, 1974
57 1
5
10
14
20
53 1
10
20
30
40
49 1
6
12
18
24
46 1
6
12
18
24
August 9, 1974
57 1
47 1
42 1
2.5
7
10.5
14
37 1
4
8
29 1
18 1
August 27, 1974
26 1
6
18 1
August 28, 1974
37 1
5
10
Sal.
14.6
16.7
20.3
20.4
0.1
0.1.
0.1
0.5
0.5
0.4
0.7
1.0
1.3
1.4
1.6
1.7
0.2
0.5
1.3
1.4
1.5
2.3
2.9
5.0
8.4
3.4
3.6
5.4
J-J — r •*-* \ *— V*.i,l.J,iJ,Ul
(pmol/^1)
1.16
1.45
0.545
0.572
0.174
0.173
0.173
0.188
0.239
0.141
0.136
0.125
0.131
0.155
0.180
0.111
0.111
0.181
0.295
0.169
0.127
0.122
0.163
0.119
0.324
0.241
0.310
0.284
0.374
0.501
0.507
0.217
0.227
0.217
0.166
0.190
0.158
0.180
0.223
0.379
0.378
0.84
-u,
PO
(pmol/1)
1.81
1.77
0.96
0.90
1.48
1.35
1.57
0.75
1.82
1.73
1.89
2.22
2.07
1.98
2.02
2.04
1.62
1.80
2.18
2.16
2.11
2.23
2.49
2.65
3.60
3.09
2.90
3.62
Part Excess
P [Rn]
(pmol/1) (dpm/1)
6.67
3.94
1.14
2.53
0.77
0.78
0.90
1.03
1.04
0.82
1.30
1.20
1.14
1.29
1.13
1.38
1.32
1.27
1.30
1.59
1.25
1.28
1.86
1.67
0.97
1.31
1.00
(continued)
167
-------
M.P.
25
13
9*
1
1**
-2
-6
-12
Depth T
(m) (°C)
1
6
11
1
5
10
0
1
5
10
1
1
6
12
1
10
20
1
6
12
TABLI
Sal.
(°/oo)
6.8
10.1
11.5
12.2
15.2
16.1
9.0
18.6
22.1
24.2
16.7
18.8
23.2
24.2
22.8
24.3
27.7
24.0
26.2
28.5
i 30 (continue
CH,
(ymol/D
0.301
0.191
0.190
0.264
0.303
0.394
1.08
0.94
1.00
1.01
1.13
1.28
1.23
1.21
1.44
0.79
0.93
0.655
0.307
:d)
(pmoltl)
3.72
4.24
5.26
4.19
5.68
5.62
7.43
5.84
6.09
5.57
5.59
6.10
5.70
6.34
5.51
3.14
5.00
3.20
1.98
Part Excess
P [Rn]
(umol/1) (dpm/1)
1.01
By 125th St. sewage outfall
4 hours after previous sample
168
-------
The total phytoplankton biomass of the coastal waters adjacent to the
Hudson is, however, related to the quantity of nutrients reaching the area.
Coastal waters adjacent to large rivers always support enhanced phytoplankton
activity produced by injection of primary nutrients, and usually support
greater fish populations as a result. Dissolved oxygen in bottom waters of
small areas of the apex of the New York Bight approach values comparable to
the Inner Harbor (< 40% of saturation). Bottom waters of the New York Bight
well beyond the zone of influence of the Hudson estuary also show substantial
oxygen depletion during the summer. The critical parameters for control of
dissolved oxygen of bottom waters immediately offshore of the Hudson are still
not well defined at this time. It is thus presently not clear if any major
detrimental effects to the coastal waters are produced by enhanced nutrient
discharge from the Hudson estuary.
In light of the current circumstances of sewage treatment, nutrient
content, and water quality parameters in the Hudson estuary, the most reason-
able policy for management of nutrient discharge during the next decade would
not appear to be the initiation of tertiary treatment for phosphate removal
(or for other primary nutrients). Considering the low summer dissolved
oxygen levels within the Inner Harbor, the major efforts in sewage treatment
should be to complete the secondary treatment facilities under construction
in New York City, and to substantially upgrade the New Jersey discharges to
include secondary treatment. Such an effort would require a minimum of a
decade considering the current mix of political interests, but should
significantly improve the dissolved oxygen content of the most heavily
impacted region of the Hudson estuary. During that period, the question of
nutrient removal and its potential effect on coastal water quality could be
more extensively examined from the viewpoint of possible construction of
tertiary sewage treatment facilities.
The conclusions presented here are based on extremely simple conceptual
models of estuarine circulation which do not include any details of
circulation or dynamics of the driving forces of the circulation. They
involve some first-order approximations of net transport rates and time
scales of flushing, as well as field data on the distribution of phosphate
and chlorophyll as a function of salinity and location within the Hudson.
More elaborate numerical models could certainly provide additional insights
into the behavior of nutrients within the Hudson, but it is unlikely from a
management viewpoint that major changes in the conclusions about phosphate
behavior within the Inner Harbor of the Hudson estuary would result from
more elaborate descriptions of the actual circulation.
169
-------
SECTION 12
WATER COLUMN TRACE METALS IN THE HUDSON
INTRODUCTION
The data reported in this section were collected by Gary Klinkhammer of
the University of Rhode Island as part of his masters thesis entitled, "The
Distribution and Partitioning of Some Trace Metals in the Hudson River
Estuary". All of the field sampling on the Hudson and laboratory work at
URI were done by Gary Klinkhammer in cooperation with his thesis advisor,
Mike Bender, a member of the faculty of the Graduate School of Oceanography
at URI. Much of the discussion in this section was taken directly from this
thesis (Klinkhammer, 1977).
Before the importance of dredge spoils and sewage sludge as sources of
trace metals can be assessed, the distribution, sources and transport
pathways of trace metals in the water column should be established. The data
discussed in this section provide much of that information for the Hudson
estuary during two periods of intensive sampling, April 1974 and October 1975.
Both of these sampling periods were characterized by high fresh water discharge
rates, early April 1974 being dominated by snow melt runoff, and October 1975
being dominated by abnormally rainy weather for several months preceding
sampling.
SAMPLE COLLECTION AND ANALYTICAL METHODS
During April 1974, subsurface samples were collected using ordinary
Niskin bottles on a"hydrowire and surface samples were taken with a plastic
bucket and polypropylene rope. In October 1975, subsurface samples were
collected with a metal-free ten liter Niskin attached to a polypropylene
rope. Samples were filtered through acid-washed 0.45 micron Millipore (T.M.)
filters as soon as possible after collection, generally within 12 hours.
Filtered samples were acidified to pH 2 with ultra-pure HC1 and stored in
acid-washed linear polyethylene bottles. Additional samples were stored cold
in the dark, and analyzed for molybdate-reactive phosphate within three days
after collection.
Soluble cadmium, zinc, copper, manganese and nickel were determined by
atomic absorption spectrophotometry (AA) after ion-exchange preconcentration.
Yields were determined by irradiating isotopically pure Cd-106, Zn-68, Cu-63,
Mn-55 and Ni-64 to produce short lived radioactive tracers (Cd-107, Zn-69M,
Cu-64, Mn-56 and Ni-65) which were added to the samples prior to passing
through Chelex-100 and then analyzed on a Ge(Li) detector after elution with
10 ml of hot 3N HNO . Quantitative analyses of the stable metals were done
170
-------
by flameless atomic absorption (graphite furnace) with a Perkin Elmer 503,
except for zinc which was done by flame AA. A great deal of attention was
devoted to blank determinations and intercalibration with other techniques
including direct injection AA of raw samples for inorganic and isotope-
dilution mass spectrometry for copper. Iron was done by direct injection into
the graphite furnace, and by preconcentration with ammonium pyrillodine
dithiocarbonate (APDC) followed by AA.
Particulate phase samples were oxidized on a low temperature asher,
transferred to a polyethylene vial with 0.2 ml of pure HF, and allowed to
digest for a week. Then 0.1 ml of pure HNO was added and digestion proceeded
for another week. Final volumes were made up to 1 ml with doubly-deionized
HO. Aliquots of 0.1 ml were dispensed onto a small filter pad, dried,
and analyzed for manganese and aluminum by neutron activation analysis
utilizing 5 minute irradiations. Dilutions of the original 1 ml quenched
digestion solutions were analyzed by flameless AA for cadmium, zinc, copper,
manganese, nickel and iron.
RESULTS
Salinity, mile point locations, pH and the soluble concentration of
trace metals are given in Table 31. Table 32 lists the concentrations in
each water sample of these metals as particulates along with the suspended
matter loads. Table 33 gives the dry weight metal abundances of the partic-
ulate matter itself. The total ranges of soluble concentrations tabulated
in Table 31 for each of the metals are compared with metal data from
Narragansett Bay, Rhode Island and the Saragasso Sea, which were also
determined by similar analytical techniques in the laboratory at the Univer-
sity of Rhode Island (Table 34).
DISCUSSION
Figure 59 shows plots of soluble zinc, copper, manganese, nickel and
reactive phosphate concentrations versus salinity for samples collected in
April 1974, and October 1974. The importance of reactive-phosphate as
a tracer for sewage input is clear from Figure 59. With the possible
exception of copper these plots show maxima in metal concentrations which
approximately coincide with those in the phosphate data. This suggests that
three source mixing was the dominant process controlling the soluble
concentrations of these metals throughout the estuary. Furthermore, in each
case, these maxima occur at salinities corresponding to the New York harbor
area.
The particulate phase also plays a major role in the transport of metals
in the Hudson. Figure 60 is a plot of the concentration of particulate
matter versus salinity. Bottom samples were consistently higher in suspended
load, as might be expected, for both upstream and downstream tidal flow.
Figure 60 also indicates that much of the suspended load of the river was
being deposited in the area of New York harbor. This observation is
consistent with the pattern of sedimentation rates based on the distribution
of Cs-137 in sediments, which indicated a substantial zone of rapid
deposition in the harbor.
171
-------
«
UJ
0.
Q_
O
O
O
0
o
O
1 1 1 1 1 1
Q
04 8 12 16 2O 24 28
APR. '740
OCT. '75 m
40
CJi
^~20
N
10
D
0
I0 O
O
o
1 1 1 1 1 1 1
16 2O 24 28
Figure 59. Sellable zinc, copper, manganese, nickel and molybdate-reactive
phosphate concentrations versus salinity for samples collected
in April 1974 and October 1975.
-------
34
0,5
-8
10
40 rA A
k
30
o>
o
<
o
20
Q
UJ
O
z
UJ
a.
en
=> 10
CO
0
i i
0.5
-8
APR 74
OCT 75
O APR 74
• OCT 75
AA BOTTOM
SAMPLES
i )• i i i* i
AO
i ii
A
CO*
10 20
SALINITY, %0
30
1000
o>
O
l-
x
C£
UJ
500
O APR 75
A OCT 75
A KNOWN LOSS
DURING ASHING
O
O
O
BEST-FIT LINE
n = 63
r = 0.949
m = 54,000 ppm
b = 140 .g
0-004 0.010 0.015
WEIGHT, grams
0.020
Figure 60. Concentration of the suspended material versus salinity and
mile point (using surface salinities for April 1974 and October 1975)
Figure 61. Weight of aluminum collected on filters versus the total weight
of suspended material - Hudson Estuary.
-------
TABLE 31
Concentrations of Dissolved Trace Metals in the Hudson River Estuary (ppb)
April 1974
Mile
Point
34
24.5
14
Salinity
Gradient
0.5
-8
-15
-18.5
Depth*
S
M
B
S
B
S
M
B
S
S
M
S
S
S
S
S
S
S
S
S
S
S
M
B
B
S
B
S
M
B
Salinity
(° / )
» / 0 0 /
0.60
0.60
0.62
1.99
2.00
3.59
3.62
3.65
1.73
1.74
3.62
1.95
2.20
2.38
2.63
3.20
3.06
3.59
5.09
5.85
6.68
7.71
12.00
15.90
22.33
21.59
21.79
29.71
30.33
32.26
Cd
.317
.309
.286
.230
.369
.194
.272
.395
.271
.230
.284
.152
.259
.279
.319
.284
.187
.237
.498
.231
-
.302
.325
.343
.339
.503
.361
.301
.091
.187
Ni
3.70
2.89
2.53
5.13
2.24
2.35
3.78
5.29
9.30
1.51
3.20
1.12
1.93
2.73
2.38
2.51
1.95
3.11
4.12
3.37
4.03
3.38
4.80
-
6.60
10.8
6.40
4.71
4.79
1.82
Zn
23.7
7.46
8.00
15.4
7.72
8.00
14.5
23.4
28.4
7.62
11.0
15.6
15.0
16.6
11.0
8.55
11.2
11.3
17.7
11.0
16.3
13.6
24.5
29.4
30.8
32.9
21.8
14.9
9.02
6.11
Cu
3.22
3.23
3.34
4.03
3.12
3.02
3.17
6.49
4.09
4.21
3.35
3.29
3.46
3.51
3.48
3.30
3.26
2.85
5.94
3.14
2.90
2.68
3.72
2.91
2.25
6.28
4.50
2.84
1.29
5.53
Mn
20.7
21.0
16.7
34.3
30.8
43.5
44.9
-
26.6
31.4
42.4
34.0
29.2
35.8
39.4
32.5
31.8
47.5
58.6
50.7
51.8
59.7
49.6
38.7
34.6
45.5
40.1
9.75
8.36
2.11
(continued)
174
-------
Mile
Point
33
31
29
27
26
25
23
21
20
18
16
14
12
10
8
6
4
3
1
0.5
0
-1
-2
Depth*
S
S
S
S
S
S
B
S
B
S
B
S
B
S
B
S
B
S
B
S
B
S
B
S
B
S
B
S
S
S
M
S
S
S
S
S
TABLE 31 (continued)
Salinity October
("/„„) Cd Ni Zn
.07
.07
.07
.89
1.03
.31
1.49
.48
1.25
.83
1.25
.89
4.90
1.10
6.29
1.68
4.31
1.98
4.06
2.66
8.22
3.33
8.17
4.19
12.40
4.48
17.63
6.83
8.11
11.47
18.46
12.40
13.67
13.72
13.87
15.43
.162
.142
.134
.139
.343
.204
.157
.129
.198
.386
.093
.292
.194
.245
.216
.443
.254
.196
.177
.145
.200
.175
.391
.184
.207
.200
.257
.179
.343
.278
.255
.627
.362
.272
1.49
1.80
1.60
1.26
2.58
2.10
1.21
2.02
2.41
2.53
1.25
2.29
2.54
2.00
2.85
2.73
2.89
2.55
2.97
2.16
2.66
2.62
3.15
1.60
3.76
2.37
4.54
2.53
3.43
3.82
4.00
4.76
4.69
4.67
5.33
4.35
3.48
7.25
4.38
2.59
5.96
3.25
3.93
2.96
3.04
8.57
4.10
4.35
6.66
4.78
3.58
5.91
5.15
5.92
4.66
3.28
6.69
6.87
7.63
9.59
9.32
6.79
14.5
6.38
9.62
9.29
14.3
14.7
11.8
14.3
30.3
10.3
1975**
Cu
2.86
2.94
3.19
3.15
3.35
3.50
2.90
3.01
3.47
3.61
3.51
3.32
2.98
3.07
2.99
3.28
3.20
3.22
3.32
3.02
2.68
3.20
2.76
3.19
2.45
3.04
2.28
2.60
3.14
2.56
2.64
3.76
3.00
3.15
3.63
2.49
Mn
1.9
1.5
1.2
1.2
2.1
1.2
2.1
1.1
1.1
2.0
2.1
1.8
19.6
2.2
23.1
4.5
15.3
6.7
17.1
9.4
28.3
14.8
40.5
20.4
42.9
21.0
45.5
28.2
33.7
41.3
43.6
39.5
42.4
40.7
44.2
40.1
Fe
91.7
20.8
16.2
37.5
23.4
30.6
14.2
26.4
33.2
21.1
26.0
13.9
43.7
10.0
57.5
22.4
23.1
32.9
26.0
18.2
51.9
16.9
30.9
16.9
22.1
24.4
13.2
10.0
15.9
13.3
51.2
15.9
12.3
14.9
39.0
13.6
(continued)
175
-------
TABLE 31 (continued)
Mile
Point
-4
-6
-8
-9
-10
Depth*
S
S
S
B
S
S
S
S
B
Salinity
(V )
» / 0 0 I
16.67
16.70
16.72
20.69
18.13
20.52
21.16
22.52
30.70
Cd
.157
.608
= 261
.136
.170
.167
.221
.234
.100
Ni
4.82
,4.83
4.29
4.19
3.99
3.22
4.08
4.29
2.10
Zn
11.6
14.4
11.8
10.3
10.8
9.84
10.2
8.75
3.60
Cu
3.13
3.08
2.59
2.32
2.33
2.37
2.35
2.32
1.08
Mn
42.9
46.9
44.8
38.8
39.9
34.4
40.5
28.2
5.8
Fe
13.3
12.3
14.6
10.6
10.6
8.3
4.7
7.0
20.3
* S = surface; M = mid-column; B = bottom
** All soluble concentrations in yg/kg
176
-------
TABLE 32
Mile Point Locations, Depths, Salinities and the Particulate Concentrations
of Six Metals for Samples Collected from the Hudson Estuary
Mile
Salinity
Point Depth* °/
34
24.5
14
Salinity
Gradient
S
M
B
S
B
S
M
B
S
S
M
S
S
S
S
S
S
S
S
S
S
S
S
M
B
B
S
B
S
M
B
0.
0.
0.
1.
2.
3.
3.
3.
1.
1.
3.
1.
2.
2.
2.
3.
3.
3.
5.
5.
5.
6.
7.
12.
15.
22.
21.
21.
29.
30.
32.
O O
60
60
62
99
00
59
62
65
73
74
62
95
20
38
63
20
06
59
09
96
85
68
71
00
90
33
59
79
71
33
26
Suspended
load
8.
7.
14.
26.
30.
27.
28.
15.
17.
10.
11.
20.
21.
20.
30.
16.
20.
11.
0.
13.
11.
11.
13.
8.
28.
10.
3.
7.
—
—
0.
mg/kg
4
0
3
8
0
7
5
6
5
9
4
3
6
6
9
5
3
9
6
9
0
8
5
8
9
9
8
7
9
Cd
.050
.121
.072
.112
.092
.074
.201
.073
.112
.057
.048
.077
.103
.106
.111
.030
.060
—
.016
.068
.030
.070
.068
.154
—
.049
.044
.027
.013
.006
.010
April 1974 **
Ni
0.8
1.2
0.8
6.4
1.2
0.7
8.0
1.2
1.5
0.7
1.1
1.2
1.3
1.1
1.6
0.2
0.7
—
0.1
2.8
und
0.9
0.6
0.6
—
0.4
0.4
und
und
und
und
Zn
__
4.7
4.2
6.7
8.5
4.2
7.7
4.0
6.0
3.3
3.7
5.8
6.4
5.7
7.9
2.0
5.2
—
0.79
1.84
0.48
3.9
4.1
3.8
—
1.77
4.1
3.00
0.20
0.04
0.29
Cu
1.65
1.61
1.21
2.8
3.2
2.5
4.8
2.7
2.9
1.19
2.1
3.3
3.9
3.7
4.5
3.6
2.3
—
und
0.82
und
1.56
2.8
3.6
—
1.89
3.2
1.98
und
und
2.5
Mn
13.8
6.9
5.8
11.5
17.1
9.7
14.9
6.0
7.0
5.6
6.1
10.5
11.8
12.0
14.1
3.5
8.2
—
1.30
2.0
0.81
3.5
5.9
5.0
—
3.4
3.1
2.7
0.91
0.22
0.40
Fe
810
850
740
1290
1360
1080
1750
810
1180
650
730
1240
1302
1280
1570
430
941
—
138
260
90
610
881
640
—
350
300
260
40
7
30
* S = Surface; M = Mid-Column; B = Bottom
** All soluble concentrations are in yg/kg
und = undetectable
(continued)
177
-------
TABLE 32 (continued)
Mile
Point
33
31
29
27
26
25
23
21
20
18
16
14
12
10
8
6
4
3
1
0.5
0
-1
-2
-4
-6
-8
-9
-10
Salinity
Depth*
S
S
S
S
S
S
B
S
B
S
B
S
B
S
B
S
B
S
B
S
B
S
B
S
B
S
B
S
S
S
M
S
S
S
S
S
S
S
S
B
S
S
S
S
B
0
.
1.
.
1.
.
1.
B
1.
>
4.
1.
6.
1.
4.
1.
4.
2.
8.
3.
8.
4.
12.
4.
12.
6.
8.
11.
18.
12.
13.
13.
13.
15.
16.
16.
16.
20.
18.
20.
21.
22.
30.
/ O O
07
07
07
89
03
31
49
48
25
83
25
89
90
10
29
68
31
98
06
66
22
33
17
19
40
48
63
83
11
47
46
40
67
72
87
43
67
70
72
69
13
52
16
52
70
Suspended
load
8
9
9
8
16
7
30
36
31
15
52
6
19
11
13
8
29
8
40
9
23
14
23
11
12
2
8
7
1
2
0
2
2
0
0
2
5
2
2
4
2
1
1
1
2
rag /kg
.8
.2
.0
.2
.4
.0
.3
.2
.2
.3
.2
.6
.0
.8
.2
.6
.0
.6
.9
.6
.5
.3
.8
.5
.5
.9
.6
.0
.7
.6
.8
.5
.8
.7
.7
.6
.5
.1
.2
.4
.5
.4
.8
.4
.9
Cd
.055
.072
.062
.136
—
.037
.090
.079
.271
.051
.132
.038
.101
.035
.034
.029
.058
.024
.066
.025
.066
.039
—
.034
.048
.030
.035
.023
.022
.022
.021
.012
.024
.030
.011
.014
.009
.095
.030
.012
.011
.005
.006
.007
.012
October 1975 **
Ni
0.75
0.72
1.03
0.84
—
0.63
1.56
1.42
1.36
0.91
2.23
1.00
1.02
0.47
0.85
0.68
1.27
0.40
0.26
0.68
1.09
0.82
1.04
0.60
0.55
0.45
0.50
0.41
0.33
0.33
0.17
0.17
0.18
0.25
0.24
0.28
0.17
0.14
0.17
0.21
0.24
1.55
0.08
0.49
0.21
Zn
4.2
4.6
7.1
4.8
—
4.1
9.0
7.6
9.1
5.3
13.9
4.7
7.5
3.1
4.6
4.5
9.3
2.7
8.1
4.6
7.4
4.9
8.0
4.5
3.4
3.9
3.1
2.8
2.9
2.3
1.50
1.31
1.38
1.62
1.71
1.64
1.12
1.13
1.57
1.38
1.99
0.73
0.83
1.07
1.71
Cu
1.52
1.83
1.34
2.2
—
1.28
3.5
3.4
3.8
2.4
5.7
2.0
2.8
0.81
2.7
1.63
4.3
1.08
5.0
1.50
4.7
2.6
4.4
2.3
2.5
2.6
2.6
1.29
0.98
1.27
0.99
0.48
0.48
0.83
0.98
1.01
0.32
0.54
0.60
0.58
1.04
und
und
0.23
0.80
Mn
11.7
12.0
19.9
15.1
23
14.5
27
26
33
22
40
17.1
18.7
14.1
16.4
11.6
18.7
10.6
15.7
16.9
12.8
13.1
13.6
9.2
8.3
6.1
4.4
8.4
3.0
5.5
3.2
2.2
2.3
2.2
2.4
2.1
3.5
2.1
4.1
4.8
2.5
2.5
1.84
2.7
5.4
Fe
690
690
1100
850
—
600
1420
1350
1400
970
2200
740
2200
690
810
530
1230
430
1360
530
1070
770
1070
—
570
390
500
470
290
280
160
160
130
150
40
130
100
120
125
100
110
50
45
80
200
178
-------
TABLE 33
Mile Point Locations, Depths, Salinities and the Metal Abundances
of Six Metals in Particulate Matter Samples Collected from the Hudson Estuary
Mile Depth
Point *
Salinity
° I
too
Cd
April 1974 **
Ni Zn Cu Mn
Fe
x 10
-3
* S = Surface; M = Mid-column; B = Bottom
** All abundances are in parts per million
und = Undetectable
Al
x 10
-3
34
24.5
14
Salinity
Gradient
0.5
-8
-15
-18.5
S
M
B
S
B
S
M
B
S
S
M
S
S
S
S
S
S
S
S
S
S
S
S
M
B
S
B
S
M
B
0.60
0.60
0.62
1.99
2.00
3.59
3.62
3.65
1.73
1.74
3.62
1.95
2.20
2.38
2.63
3.20
3.06
3.59
5.09
5.96
5.85
6.68
7.71
12.00
i R on
j-j . yu
22.33
21.59
21.79
29.71
30.33
32.26
5.9
17
'5.0
4.2
3.1
2.7
7.1
4.7
6.4
5.2
4.2
3.8
4.8
5.2
3.6
1.8
3.0
5.5
4.9
2.7
5.9
5.0
18
4.5
12
3.5
11
96
170
56
240
41
25
280
77
86
64
97
59
60
60
52
12
34
80
200
und
76
44
68
38
110
und
und
und
und
670
300
250
280
150
270
250
340
300
330
290
290
280
260
120
260
170
130
40
330
300
440
160
1090
390
340
200
230
90
110
110
90
170
170
160
110
190
160
180
180
150
220
120
70
und
60
und
130
210
400
95
830
260
und
und
und
1300
990
400
430
570
350
520
390
400
510
530
520
550
580
460
210
400
140
70
300
440
570
320
820
350
460
96.5
122.0
51.5
48.1
45.5
39.0
61.4
52.1
67.4
59.5
64.3
61.1
61.1
62.1
50.8
25.8
46.4
31.5
18.3
8.2
52.0
65.3
73.7
32.0
79.0
34.0
33.6
111
101
54.4
61.5
47.0
60.0
61.6
75.5
60.4
72.7
56.6
81.0
70.8
67.1
31.4
53.7
47.6
302
30.0
16.1
49.2
60.9
65.2
51.9
66.6
35.8
29.7
(continued)
179
-------
TABLE 33 (continued)
October 1975 **
Mile
Point
33
31
29
27
26
25
23
21
20
18
16
14
12
10
8
6
4
3
1
0.5
0
-1
-2
-4
-6
-8
-9
-10
Depth
*
S
S
S
S
S
S
B
S
B
S
B
S
B
S
B
S
B
S
B
S
B
S
B
S
B
S
B
S
S
S
M
S
S
S
S
S
S
S
S
B
S
S
S
S
B
Salinity
° /
/ o o
.07
.07
.07
.89
1.03
.31
1.49
.48
1.25
.83
1.25
.89
4.90
1.10
6.29
1.68
4.31
1.98
4.06
2.66
8.22
3.33
8.17
4.19
12.40
4.48
17.63
6.83
8.11
11.47
18.46
12.40
13.67
13.72
13.87
15.43
16.67
16.70
16.72
20.69
18.13
20.52
21.16
22.52
30.70
Cd
6.3
7.8
6.9
17
5.3
3.0
2.6
8.7
3.3
2.5
5.8
5.3
3.0
2.6
3.3
2.0
2.8
1.6
2.6
2.8
2.7
3.0
3.8
10
4.1
3.3
13
8.5
26
4.8
8.6
43
16
5.4
1.6
45
14
2.7
4.4
3.6
3.3
5.0
4.1
Ni
85
78
110
100
90
52
39
44
60
43
150
54
40
64
79
44
47
60
71
46
57
44
52
44
160
58
59
210
130
210
68
64
360
340
110
31
67
77
48
96
1100
44
350
72
Zn
480
530
790
590
580
300
220
290
350
270
710
400
260
350
520
320
310
200
480
320
340
340
270
270
1300
360
410
1700
880
1900
520
500
2300
2400
630
200
540
710
310
800
520
460
760
590
Cu
170
200
150
270
180
120
90
120
160
110
310
150
70
200
190
150
130
120
160
200
180
190
200
200
880
300
180
580
490
1200
190
170
1200
1400
390
60
260
270
130
420
und
und
160
280
Mn
1330
1300
2210
1840
2070
880
710
1070
1410
760
2590
980
1200
1240
1350
650
1230
380
1760
550
920
570
800
660
2100
510
1200
1700
2120
4000
880
820
3140
3430
810
640
1000
1860
1090
1000
1790
1000
1930
1860
F ,
x 10 '
77.7
75.2
122.0
103.0
86.2
47.1
37.4
63.0
42.3
113.0
50.6
43.5
61.6
62.3
42.6
50.1
33.3
55.1
45.8
54.1
45.3
262.0
138.0
57.6
68.0
166.0
107.0
11.1
64.4
46.1
208.0
197.0
50.7
19.1
54.9
58.3
21.8
43.0
31.4
25.1
60.8
68.8
Al
x 10
-3
109
102
210
187
83.6
138
90.0
63.6
58.0
119
75.7
181
95.0
85.5
118
99.8
75.9
91.9
57.9
90.2
69.0
87.6
69.5
107
252
98.4
267
214
350
140
83.2
395
128
535
108
93.5
62.0
128
96.5
98.4
128
109
180
-------
TABLE 34
Comparison Between Ranges of "Soluble" Metal
Concentrations Found in the Hudson Estuary,
Narragansett Bay, Rhode Island, and the Sargasso Sea
:tal Hudson
Estuary
(yg/kg)
Cd
Zn
Cu
Mn
Ni
.10 -
2.5 -
2.0 -
15 -
1.0 -
.50
35
7.0
70
10
Narragansett
Bay
(yg/kg)
.05 -
5,0 -
1.0 -
10 -
3.0 -
1.0
30
7.0
30
10
Sargasso
Sea
(yg/kg)
.005 -
.10 -
.01 -
.12 -
.025
.12
.15
.17
Unpublished data of G. Klinkhammer.
2
Cd, Cu and Ni values were taken from Bender and Gagner (1976)
Mn data was taken from Bender et al (1976).
181
-------
CO
APR.'74
OCT. '75
R«nge of
cruslal rallos
o
i i i i i i i rn i i i i T
7~P /111 i 111 rr 11 i i~r
0 4 8 12 16 20 24 28 0 4 8 12 16 20 24 28
SALINITY, "t
048 12 16 20 24 28
Figure 62. Metal to aluminum weight ratios of six metals in particulates
filtered from the Hudson Estuary plotted versus salinity.
-------
From the trends in soluble metals and suspended loads, it is clear that
New York harbor is a zone of major perturbation, involving large addition of
soluble metals from sewage and substantial decrease in suspended load due to
changes in estuarine geometry and/or salting out effects (flocculation).
Reactive-phosphate is apparently a useful tracer for sewage input of soluble
metals, but an additional chemical tracer for the suspended load would be
valuable for normalizing the metal abundances of the particulate matter. From
Figure 61, it is apparent that particulate aluminum is a good indicator of the
mass of suspended load. The metal abundances of the particulates have been
normalized to aluminum and the weight ratios plotted versus salinity in
Figure 62. Also shown are ranges in average crustal ratios of trace metals
to aluminum as given by Turekian and Wedepohl (1961).
The concentrations of metals in sewage discharged from New York City are
given in Table 35. One of the data sets (Klein et_ al., 1974) represents some
approximation of total metals in New York City sewage from samples composited
daily over a period of a month. Samples were taken unfiltered, and 15 ml of
concentrated HC1 and 5 ml of concentrated HNO were added to each 100 ml
aliquot (1.8 N in HC1 and 0.8 N in HNO,,) . The samples were then autoclaved
and analyzed by flame atomic absorption spectrometry.
The plot of soluble manganese versus salinity in Figure 59 shows
distinctive trends for the two collection periods. During April 1974,
soluble manganese in the fresh water endmember was about 17 yg/kg. The
concentration increased sharply toward Manhattan, reaching a maximum at a
salinity of about 8°/00 (in the East River). Above this salinity the levels
decreased gradually toward the Bight where the lowest concentration found was
about 1 yg/kg at a salinity of 32°/00. The same trend is apparent for samples
collected in October 1975, although the levels found were strikingly different.
Here the average concentration of 14 fresh water samples was about 2 ug/kg.
Again these levels increased sharply downstream reaching a maximum of about
45 yg/kg at a salinity of 14°/00 (also in the East River). Toward the Bight
the levels again decreased to a minimum of about 6 yg/kg at 31°/00. The large
maximum in the harbor is most likely due to sewage discharge, and based on a
three source mixing model a sewage manganese concentration on the order of
1000 yg/kg is suggested for both sampling periods. The difference in the
fresh water dissolved manganese during the two sampling periods is most likely
due to a difference in the partitioning of Mn between the particulate and
soluble phases, since the total Mn entering the estuary is similar during
both surveys (Figure 63). We have not determined the cause of the difference
in partitioning but several plausible explanations are the somewhat longer
fresh water residence time (lower fresh water flow return) to allow time for
oxidation to less soluble Mn(II) during the October period, and the higher
water temperature during October.
There is evidence for desorption of manganese from suspended matter in
the saline waters of the Hudson in the decrease in Mn to Al ratios downstream
(Figure 62) and of loss of Mn from bottom sediments resulting in lower average
Mn concentrations in deep water suspended matter, which contains a considerable
amount of resuspended sediment (Figure 64). We have not measured directly
the rate of remobilization of Mn from sediments deposited in the harbor, but a
mass balance calculation for the f,low conditions of the October survey
indicates a rate of 1-2 yg/cm /cm /day (Klinkhammer, 1977), in good agreement
183
-------
TABLE 35
Reported Metal Concentrations of Sewage
Being Discharged into the Lower Hudson Estuary
Interstate
Sanitation
Commission
Report (1972)
Number of
Determinations
* Cd Zn
491
578
Cu
583
Mn
Ni
503
398
Fe
269
yM
522
Average
Concentration
<20
120
100
120 <100
600
60
Klein, et al.
(1974)
Number of
Determinations
Average
Concentration
Range in
Concentrations
Assumed in
this Study
240
10
240
260
240
150
240
100
16+ 180+ 120+ 825+ 100+ 600+
8 90 60 410 50 310
60
*A11 metal concentrations are in yg/kg.
184
-------
eoi-
O
o>
60
00
>,O
CO
O
O
O
r w
.'*
O APR 74
• OCT 75
o
1 1 1
1 I 1 1 1 I 1 1 I 1 1 p 1
0
10 20
SALINITY, %o
30
Figure 63. Plot of total manganese versus salinity for samples collected
in the Hudson Estuary.
185
-------
4000 p-
E
a,
a.
~3000
LU
fe
_J
^
O
H-
CC
< 2000
a.
UJ
o
z
^ 1000
GQ
OCTOBER, 1975
O SURFACE
• BOTTOM
I I I I I j I I I I I I III I ill L JL Li ..II. I I I i I III I I I
30 25 20 15
MILE POINT
0
Figure 64. Manganese abundances of surface and subsurface particulates
collected from the Hudson Estuary during October 1975 plotted
versus mile point. The average manganese abundance of the
surface particulates was 1500 ppm while it was 750 ppm for the
subsurface samples.
186
-------
with direct Mn flux measurements made in Narragansett Bay (Graham et al.,
1976).
CONCLUSIONS
The partitioning of manganese between solid and soluble phases at the
fresh water extremity of the estuary is apparently related to some temperature
and/or time dependent process. An oxidation has been suggested here although
other explanations cannot be eliminated by these data. In any case, the total
manganese at low salinities was essentially constant during both collection
periods at about 20 yg/kg. Downstream, manganese concentrations were
dominated by a large anthropogenic input of soluble metal. This sewage Mn
apparently remains soluble for the residence time of water in this section of
the estuary (several days). The bulk of the particulate fraction was being
deposited in New York harbor. These data further suggest that some desorption
and/or reduction at the water-sediment interface may be occurring across the
salinity gradient. However, the extent of any such process appears to be
minor when compared to the direct supply of Mn from sewage.
The averaged partitioning of five metals between soluble and suspended
load phases is shown for three regions of the Hudson in Figure 65. Except for
iron, which was predominantly (^ 99%) on particles, most of the transport of
metals in the Hudson was occurring in solution. The general trends were
consistent for both sampling periods, except for manganese which was discussed
earlier.
187
-------
100
so
Q
<
o
Z w
si50
or z
LU
0.
± 50
o
I-
o
H
UJ
-
o 8
CO
CO
LU
< 50-
o
APRIL 1974
HUDSON
*Q29
3.5
15
4.0
37
UPPER
BAY
0.40
31
5.2
40
LOWER
BAY
0.20
3.8
10
1.9
6.8
OCTOBER 1975
HUDSON
0.20
3.0
3.3
16
UPPER
BAY
0.24
4.7
2.6
42
LOWER
BAY
0.17
3.2
1.7
17
^Average Concentration of the
Soluble Phase for each Zone, ^ag
Figure 65. The partitioning of five trace metals in the Hudson Estuary.
Iron data is not shown but about 99% of this metal was
present in the particulate phase in all three boxes during
October 1975.
183
-------
SECTION 13
SUMMARY OF HUDSON FIELD RESEARCH RESULTS
In Section 4 we discussed the distribution and sources of several metals
in Hudson estuary sediments in terms of the background trace metal composition
of fine-grained sediments and the levels of pollutant metals added from diffuse
sources throughout the Hudson estuary and from the concentrated sources in the
New York City area. The observed concentrations of Zn, Cu and Pb were
generally 3-6 times background levels with the highest values, especially for
Cu, in New York harbor/ We have not attempted to derive the total burden of
pollutant metals in New York harbor sediments, although that would probably be
feasible to do from the information we are developing about recent sediment
deposition pattersn in the Hudson based on the distribution of Cs-137.
Considerable attention has been directed to the elevated levels of trace metals
in New York Bight sediments due to disposal of dredge spoils and sewage sludge.
Although the levels of trace metals in the sediments of the dumping area are
very high relative to those of more distant sandy shelf sediments and thus
make very dramatic and useful mapping tools, the concentrations of metals in
most of the surface sediments of the Hudson, and especially in New York harbor
are comparable to those in the dumping areas. Thus if the pollutant metal
concentrations in the sediments of the Bight apex are considered as a serious
environmental problem, then similar concern should also be directed to the
condition of much of the sediments of the Hudson estuary.
The area of most intense metal pollution that we have observed in the
Hudson is in a small cove about 60 km north of New York City. Effluents of a
battery factory released over several decades have contaminated sediments of
a small portion of Foundry Cove to levels above a percent for both Cd and Ni.
We have completed a survey of the Cd and Ni distribution in the least
contaminated portion of this cove (<_ 1000 ppm in Cd and Ni) adjacent to the
main channel of the Hudson. Of the total burden of Cd (and Ni) in Foundry Cove
of 25-50 tons, we estimate that at least 2 tons have been transported away
from the area of most contamination, through a narrow connecting channel, and
accumulated in the outer area of the cove. We believe most of this transport
probably occurred by resuspension and deposition of fine particles in the
sediments over a number of years, mobilized by tidal currents, and perhaps to
some extent by a dredging operation in the most contaminated area of the inner
cove. The covariation of Cd and Ni in outer cove sediments and the pattern of
reactor radionuclide distribution in the sediments argue against transport of
Cd and Ni in solution to the outer cove, and against a single episode of Cd
and Ni deposition in the outer cove (i.e. dredging remobilization) which has
been spread through the upper 20 cm of sediments by biological or physical
mixing processes.
189
-------
Our studies of interstitial water (pore water) Cd in Foundry Cove have
been promising, but have not progressed sufficiently to reach many conclusions.
The new pore water sampling technique developed in our laboratory appears to
be quite a valuable addition to the study of sediment chemical processes,
especially in relatively shallow waters accessible by divers. The concentra-
tions of dissolved Cd in Foundry Cove sediments appear to be much higher
than water column Cd concentrations, even in the most heavily impacted area
of the Hudson. Thus these sediments definitely appear likely to be a source
of soluble Cd, as opposed to the sediments of Narragansett Bay, which do not
appear to be a source of soluble Cd. We cannot make a good estimate of the
strength of the source of Cd from these sediments, based on our experiments
to date, but can place some probable limits on flux estimate budget calcula-
tions by relating the pore water Cd data to that of Mn. In Foundry Cove the
most contaminated sediment area which we studied had pore water Cd concentra-
tions of approximately 30 ppb, about two orders of magnitude less than Mn in
the same samples. The flux of Mn from Narragansett Bay sediments has been
directly measured to be 1-2 yg/cm /day. If the same flux were to characterize
Foundry Cove sediments the total flux of Mn per year over the area of the cove
(^ 0.5 km ) would be ^ 3 tons. If the flux of Cd were in the same proportion
to Mn as the ratio of pore water^concentrations (^ 1/100) the annual flux of
Cd would be ^ 0.3 T or about 10 of the sediment burden of Cd in the Cove.
There are several reasons why this might be expected to be an upper limit
of the Cd flux from Foundry Cove sediments, but without more field experiments
we cannot be more definite at this time. It does appear likely that Foundry
Cove sediments during fresh water conditions that existed during the
experiments described here are a net source of soluble Cd and not a net sink
as indicated for the estuarine conditions of Narragansett Bay sediments.
We have found the distribution of a number of reactive pollutants,
including Zn, Cu, Pb, Cd, Ni and polychlorinated biphenyls (PCB's) and
Pu-239,240 to be strongly correlated with the distribution of Cs-137 in
recent Hudson sediments. This covariation indicates that fine particles in
estuaries are a transport vector for a large spectrum of pollutants and greatly
simplifies the task of mapping the distribution of reactive pollutants in
coastal zone sediments.
We have explored the use of C-14 as an indicator of the source of
pollutant carbon in Hudson sediments, and found that the presence of recent
sewage carbon in New York harbor overwhelms the presence of fossil fuel
contaminated carbon which produce old apparent C-14 ages of organic matter in
a number of environments contaminated with recent pollutants (Baltic Sea
coastal sediments, Narragansett Bay, and Lake Washington).
In addition to exploiting C-14 measurements to learn about the source of
pollutant carbon in contaminated sediments, we have developed a new type of
analytical approach based on enzymes for quantitative determination of
cellulose in coastal zone sediments. A preliminary survey indicates that
this technique could provide substantial new insights into the diagenesis of
large organic molecules in the coastal environment.
On the basis of a large number of water column and sediment measurements
of Rn-222, a radioactive daughter of Ra-226, we were able to establish
quantitatively the rate of transfer of a chemically-inert natural tracer from
190
-------
the sediment to the water column in the Hudson estuary. One major
conclusion was that the flux of Rn-222 from the sediments could be computed
to within a factor of two based on a simple model of molecular diffusion
within the sediments. The observed flux was approximately a factor of two
greater than predicted from molecular diffusion and could be attributed to
a combined effect of bioturbation, sediment stirring by currents, and other
poorly defined processes which might enhance the flux of material from
sediment interstitial waters.
We also explored the use of dissolved methane as an indicator of
sediment-water exchange rates. The supply from the sediments appears to be
dominated by partial solution of methane bubbles which are produced in the
sediments and then pass up through the water column. Thus methane is not a
useful analog for trace metal fluxes, but some new insights were gained. An
extremely large range of dissolved methane concentrations within the sediments
was observed (>_ x 100(5), indicating that the rate of anaerobic fermentation
of organic matter and/or bacterial consumption of methane in sediments in the
Hudson must be greatly variable. Since sulfate reduction (and sulfide
production) is to some extent a competing process with that of methane
production for destruction of organic matter in anaerobic sediments, and since
trace metal solubilities in sediments may be strongly influenced by sulfide
concentrations, the relationship between methane concentration in sediments
and metal mobility should probably be explored.
We have invested substantial effort to establish the amounts of sewage
components discharged to the Hudson near New York City and to develop simple
model descriptions of the flushing characteristics of the Hudson estuary.
This is essential for interpreting water column metal data in terms of
transport processes and fluxes to the adjacent coastal waters. We have found
dissolved phosphate to be a valuable indicator of both sewage loading rate
and residence time of water in the Hudson near New York City. Thus phosphate
is valuable as an indicator of the rate of sewage metal discharge and of the
transport rates of dissolved metals out of the estuary. The rapid transport
of phosphate from the Hudson by physical mixing compared with removal rates by
phytoplankton activity indicates that nutrient removal in the New York City
area probably would have little immediate benefit to water quality in the
Hudson estuary, and thus tertiary treatment of sewage should not be intro-
duced until clear evidence of beneficial results from such treatment can be
developed.
Extensive surveys (by Gary Klinkhammer of the University of Rhode Island)
of dissolved and suspended phase metals in the Hudson have been completed for
two high fresh water flow periods. Most of the water column transport of
metals out of the Hudson is accomplished by soluble metals, while deposition
of large amounts of suspended particles causes accumulation of large amounts
of particulate phase trace metals in New York harbor. Using the measured
concentrations of soluble metals and phosphate in New York harbor, and our
estimate of the discharge rate of phosphate through the Narrows from simple
transport models (^ 6 moles/sec) we can calculate the discharge rate of
soluble metals (Mn, Fe, Zn, Cu, Cd and Ni) from the Hudson estuary to the
New York Bight apex (Table 36). From the estimated metal deposition rates by
rain and dry fallout to the New York Bight (Duce _e_t _al. , 1976), we derived
maximum estimates of the delivery of total Fe, Zn and Cd from the atmosphere
191
-------
TABLE 36
Soluble Metal Fluxes to the New York Bight Apex
(first order estimates - grams per second)
Hudson Discharge „
April 1974 October 1975
Mn
Fe
Zn
Cu
Cd
Ni
104
-
81
14
1
21
76
36
22
5
0.4
9
Air Deposition Benthic
to Apex „ „ Flux ,
(2 x 10 km ) (150 km )
+ 15 + 5
47 +1.5+1
10
+ 0.1 + 0.2
0.2 -0.01 + 0.02
0 + 0.1
Derived from data in Sections 11 and 12
2
Derived from data in Sections 11 and 12
3
Derived from atmospheric deposition rate estimates by Duce et_ al., 1976,
4
Derived from studies of Narragansett Bay at the University of Rhode
Island by M. Bender.
192
-------
(Table 36). Assuming all of the atmospheric metals are rapidly dissolved,
it is possible that this flux could be significant for soluble Fe, Zn and Cd
relative to the dominant estuarine discharge of soluble metals. These
estimates of atmospheric deposition have a very large degree of uncertainty
(Duce et al., 1976) and probably are less significant in the New York Bight
apex relative to other fluxes than indicated in Table 36. We have also
attempted to derive first order estimates of the flux of?metals from waste
area sediments (Gross, 1976), assuming an area of 150 km of fine-grained,
metal-contaminated sediments. The flux estimates per unit area we used were
derived from studies in Narragansett Bay by the University of Rhode Island.
Except for Mn, where the benthic flux approaches 15-25% of the estuarine
discharge rate of soluble Mn, the waste area sediments do not appear to be
significant sources of metals (Fe, Cu and Cd). Extrapolation of the
Narragansett Bay results to the New York Bight waste area was done because
these data are the best available data of which we are aware for organic rich,
reducing sediments in a coastal environment.
Our conclusions about the lack of importance of soluble phase mobliization
of metals from the waste area in the Bight, except for Mn, is consistent with
the observations of Segar and Cantillo (1976), where the only consistent
difference between dump site water column dissolved metals and a distant
shelf central site was also for Mn.
There do appear to be considerable differences in published New York Bight
metal concentrations far from zones of contamination (Segar and Cantillo, 1976)
compared with those obtained by Windom (1977) for southeastern coastal waters,
especially for Zn and Cu. Despite the uncertainty of the absolute concentra-
tions of soluble metals in the New York Bight, the data do not seem incon-
sistent with the interpretation that the waste area sediments are a relatively
insignificant source of soluble metals in the New York Bight apex, compared
with the flux of soluble metals from the Hudson.
Our opinion is that, although more effort should be devoted to the
understanding of trace metal transport pathways in coastal waters impacted by
dredge spoil and sewage discharges, it is at least equally important to
intensify studies of the environmental pathways and significance of pollutants
such as PCB's and other toxic organic compounds, especially for the New York
City area.
193
-------
REFERENCES
1. Abood, K.A. Circulation in the Hudson Estuary. In: Hudson River
Colloquium, Annals of the New York Academy of Sciences, 250:36-111,
1974.
2. Artem'yev, V.Y. Carbohydrates in the bottom sediments of the Central
Pacific. Oceanology, 10:508-513, 1970.
3. Atkinson, L.P., and F.A. Richards. The occurrence and distribution
of methane in the marine environment. Deep-Sea Research, 14:673-684,
1967.
4. Bender, M.L. Does upward diffusion supply the excess manganese in
pelagic sediments? Journal of Geophysical Research, 76:4212-4215,
1971.
5. Berner, R.A. Principles of Chemical Sedimentology. McGraw-Hill,
New York, 1971. 240 pp.
6. Biggs, R.B., and C.D. Wetzel. Concentration of particulate carbohydrate
and the halocline in Chesapeake Bay. Limnology and Oceanography, 13:
169-171, 1968.
7. Bondeitti, E.A., F.H. Sweeton, T. Tamura, R.M. Perhac, L.D. Hulet, and
T.J. Kneip. Characterization of cadmium and nickel contaminated
sediments from .Foundry Cove, New York. In: Proceedings of the
National Science Foundation Trace Contaminants Conference, 1973.
8. Boyle, E.A., J.M. Edmond, and E.R. Sholkovitz. The mechanism of iron
removal in estuaries. Geochimica et Cosmochimica Acta, 41:1313-1324,
1977.
9. Bowen, V.T., and W. Roether. Vertical distributions of strontium-90,
cesium-137 and tritium near 45° North in the Atlantic. Journal of
Geophysical Research, 78:6277-6285, 1973.
10. Bower, P.M. Cadmium and nickel in the sediments of Foundry Cove.
Master's Thesis, Queens College of the City University of New York,
Flushing, New York, 1976. 162 pp. with appendicies.
11. Broecker,- W.S., and E.A. Olson. Lamont radiocarbon measurements, VI.
Radiocarbon 1:111-132, 1959.
194
-------
12. Broecker, W.S., and A. Walton. The geochemistry of C in fresh
water systems. Geochemica et Cosmochimica Acta, 16:15-38, 1959.
13. Broecker, W.S., and E.A. Olson. Lament radiocarbon measurements, VIII.
Radiocarbon, 3:176-204, 1961.
14. Broecker, W.S. Radiocarbon dating: A case against the proposed link
between river mollusks and soil humus. Science, 143:596-597, 1964.
15. Broecker, W.S. The application of natural radon to problems in ocean
circulation. In: Symposium on Diffusion in Oceans and Fresh Waters,
Lament Geological Observatory, Palisades, New York, pp. 116-145, 1965.
16. Broecker, W.S., E.R. Bonebakker, and G.G. Rocco. 'The vertical
distribution of cesium-137 and strontium-90 in the oceans, 2.
Journal of Geophysical Research, 71:1999-2003, 1966.
17. Broecker, W.S., and T.H. Peng. Gas exchange rates between air and
sea. Tellus, 26:21-35, 1974.
18. Bruland, K.W., K. Bertine, M. Koide, and E.D. Goldberg. History of
metal pollution in Southern California coastal zone. Environmental
Science and Technology, 8:425-432, 1974.
19. Buehler, K., and H.I. Hirshfield. Cadmium in an aquatic ecosystem:
Effects on planktonic organisms. In: Trace Contaminants in the
Environment, Proceedings of the Second Annual NSF-RANN Trace Contaminants
Conference, Asilomar, Pacific Grove, California, 1974.
20. Cain, W.F-, and H.E. Suess. Carbon-14 in tree rings. Journal of
Geophysical Research, 81:3688-3694, 1976.
21. Cappenberg, T.E. Ecological observations on heterotrophic, methane
oxidizing and sulfate reducing bacteria in a pond. Hydrobiologia,
40:471-485, 1972.
22. Cappenberg, T.E. A study of mixed continuous cultures of sulfate-
reducing and methane-producing bacteria. Microbial Ecol., 2:60-72, 1975.
23. Carmody, D.J., J.B. Pearce, and W.E. Yasso. Trace metals in sediments
of New York Bight. Marine Pollution Bulletin, 9:132-135, 1973.
24. Chow, T.J., K. Bruland, K. Bertine, A. Soutar, M. Koide, and E.D.
Goldberg. Lead pollution: Records in Southern California coastal
sediments. Science, 181:551-552, 1973.
25. Claypool, G.E., and I.R. Kaplan. The origin and distribution of
methane in marine sediments. In: Natural Gases in Marine Sediments,
I.R. Kaplan, ed.. Plenum Press, New York, 1974. 324 pp.
195
-------
26. Craig, H. Isotopic standards for carbon and oxygen and correction
factors for mass spectrometric analysis of carbon dioxide. Geochimica
et Cosmochimica Acta, 12:133-149, 1957-
27. Deevey, E.S., M.S. Gross, G.E. Hutchinson, and H.L. Kraybill.
The natural C-14 content of materials from hard-water lakes.
Proceedings National Academy of Science, 40:285-288, 1954.
28. deGroot, A.J., W. Salomons, and E. Allersma. Processes affecting
heavy metals in estuarine sediments. In: Estuarine Chemistry,
J.D. Burton and P.S. Liss, eds., Academic Press, London, pp. 131-157,
1976.
29. Deindoerfer, R.N., and D.E. Humphrey. Mass transfer from individual
gas bubbles. Ind. Eng. Chem., 53:755-759, 1961.
30. Doelle, H.W. Bacterial Metabolism. Academic Press, New York, 1969.
31. Dreywood, R. Qualitative tests for carbohydrate materials. Ind. Eng.
Chem., 18:499, 1946.
32. Dubois, M., K.A. Gilles, J.K. Hamilton, P.A. Rebers, and F. Smith.
Colorimetric method for determination of sugars and related substances.
Analytical Chemistry, 28:350-356, 1956.
33. Duce, R.A., G.T. Wallace, Jr., and B.J. Ray. Atmospheric trace metals
over the New York Bight. NOAA Technical Report, ERL 361-MESA 4,
1976, 17 pp.
34. Edgington, D.N., J.J. Alberts, M.A. Wahlgren, J.O. Karttunen, and
C.A. Reeve. Plutonium and americium in Lake Michigan sediments.
In: Transuranium Nuclides in the Environment, IAEA-SM-199/47,
IAEA, Vienna, pp. 493-516, 1976.
35. Elderfield, H. Manganese fluxes to the oceans. Marine Chemistry,
4:103-132, 1976.
36. Emerson, S.R. Gas exchange rates in small Canadian shield lakes.
Limnology and Oceanography, 20:754-761, 1975.
37. Emerson, S.R. Early diagenesis in anaerobic lake sediments.
Geochimica et Cosmochimica Acta, 40, 1976.
38. Emery, K.O. Estuaries and lagoons in relation to continental shelves.
In: Estuaries, G.H. Lauff, ed., American Association for the Advancement
of Science, Pub. 83:9-11, 1967-
39. Emery, K.O. A coastal pond, studied by oceanographic measurements.
American Elsevier, New York, 1969.
196
-------
40. Erlenkeuser, H., E. Suess, and H. Willkomm. Industrialization
affects heavy metal and carbon isotope concentrations in recent
Baltic Sea sediments. Geochimica et Cosmochimica Acta, 38:823-842,
1974.
41. Evans, D.W., and N.H. Cutshall. Effects of ocean water on the
soluble-suspended distribution of Columbia River radionuclides.
In: Radioactive Contamination of the Marine Environment, International
Atomic Energy Agency, Vienna, pp. 125-138, 1973.
42. Evans, D.W., N.H. Cutshall, F.A. Cross, and D.A. Wolfe. Manganese
cycling in the Newport River estuary, North Carolina. Estuarine and
Coastal Marine Science, 5:71-80, 1977.
43. Fairbridge, R.F., and W.S. Newman. Postglacial crustal subsidence of
the New York area. Zeitschrift fur Geomorphologie, 12 (3):296-317, 1968.
44. Fanning, K.A., and-M.E.Q. Pilson. The diffusion of dissolved silica
out of deep-sea sediments. Journal of Geophysical Research, 79:1293-
1298, 1974.
45. Farmer, J.G., V.T. Bowen, and V.E. Noshkin. Long-lived artificial
radionuclides in Lake Ontario, I. Supply from fallout, and concentrations
in lake water of plutonium, americium, strontium-90 and cesium-137-
Submitted to Limnology and Oceanography, 1977.
46. Feely, H.W., and J.L. Kulp. Origin of Gulf coast salt-dome sulfur
deposits. Bulletin of American Association of Petrol. Geol.,
41:1802-1853, 1957.
47. Folsom, T.R., N. Sreekimaran, Hansen, J.M. Moore, and R. Grismore.
Some concentrations of Cs-137 at moderate depths in the Pacific
1965-1968. USAEC Report HASL-217, 1970, 109 pp.
18 13
48. Fritz, P., and S. Poplawski. 0 and C in the shells of freshwater
molluscs and their environments. Earth and Planetary Science Letters,
24:91-98, 1974.
49. Garside, C., T.C. Malone, O.A. Roels, and B.C. Sharfstein. An
evaluation of sewage-derived nutrients and their influence on
the Hudson Estuary and New York Bight. Estuarine and Coastal Marine
Science (in press).
50. Garrels, R.M., and C.L. Christ. Solutions, minerals and equilibria.
Freeman, Cooper and Company, San Francisco, California, 1965.
51. Gerchakov, S.M., and P.G. Hatcher. Improved technique for analysis
of carbohydrates in sediments. Limnology and Oceanography, 17:938-943,
1972.
197
-------
52. Goldhaber, M., R.A. Berner, R.C. Allen, and J.K. Cochran.
Sulfate reduction, diffusion and bioturbation in Long Island Sound.
In: Symposium on Marine Chemistry in the Coastal Environment,
American Chemical Society Middle Atlantic Region, 1975 (abstract).
53. Graham, W.F., M.L. Bender, and G.P. Klinkhammer. Manganese in
Narragansett Bay. Limnology and Oceanography, 21:665-673, 1976.
54. Gross, M.G. Geologic aspects of waste solids and marine waste
deposits, New York metropolitan region. Geological Society of America
Bulletin, 83:3163-3176, 1972.
55. Gross, M.G. Waste disposal. MESA New York Bight Atlas Monograph,
26, 1976, 32 pp.
56. Gross, M.G., J.A. Black, R.J. Kalin, J.R. Schramel, and R.N. Smith.
Survey of marine waste deposits, New York metropolitan region.
Marine Science Research Center Technical Report No. 8, State
University of New York at Stony Brook, 1971, 72 pp.
57. Guinasso, N.L., and D.R. Schink. A simple physiochemical acoustic
model of methane bubbles rising in the sea. Unpublished report No.
73-15T, Texas ASM University, College Station, Texas, 1973, 24 pp.
58. Hammond, D.E., and H.J. Simpson. Methane distribution and sediment
flatulence in the Hudson River Estuary. Limnology and Oceanography,
revisions submitted, 1977.
59. Hammond, D.E., H.J. Simpson, and G. Mathieu. Methane and radon-222
as tracers for mechanisms of exchange across the sediment-water
interface in the Hudson River Estuary. In: Marine Chemistry in
the Coastal Environment, T. Church, ed., ACS Symposium Series 18:
119-132, 1975.
222
60. Hammond, D.E., H.J. Simpson, and G. Mathieu. Rn distribution and
transport across the sediment-water interface in the Hudson River
Estuary. Journal of Geophysical Research, in press, 1977.
61. Handa, N. Examination of the applicability of the phenol sulfuric
acid method to the determination of dissolved carbohydrates in sea
water. Journal of Oceanogr. Soc. Japan, 22(3):l-8, 1966.
62. Hardy, E.P. Depth distributions of global fallout Sr-90, Cs-137 and
Pu-239,240 in sandy loam soil. USAEC Report HASL-286, pp. 2-10, 1974.
63. Hardy, E.P., and L.T. Alexander. Rainfall and deposition of Sr-90
in Clallam County, Washington. Science, 136:881-882, 1962.
64. Harley, N., I. Fisenne, L.D.Y. Ong, and J. Harley. Fission yield
and fission product decay. USAEC Report HASL-164, 1965.
198
-------
65. Hatcher, P.G., and L.E. Keister. Carbohydrates and organic carbon
in New York Bight sediments as possible indicators of sewage
contamination. In: The Middle Atlantic Shelf and New York Bight,
Limnology and Oceanography Special Symposium, 2, M.G. Gross, ed.,
in press.
66. Hesslein, R.H. An in situ sampler for close interval pore water
studies. Limnology and Oceanography, 22:913-915, 1976.
67- Higbie, R. The rate of absorption of a pure gas into a still liquid
during short periods of exposure. Trans. Am. Inst. Chem. Eng.,
35:365-389, 1935.
68. Hunter, J.V., and H. Heukelekian. The composition of domestic sewage
fractions. Journal of Water Pollution Control Fed., 37:1142-1163, 1965.
69. Hutton, W.E., and C.E. Zobell. The occurrence and characteristics of
methane oxidizing bacteria in marine sediments. Journal of Bacteriology,
58:463-473, 1949.
70. Keith, M.L. and G.M. Anderson. Radiocarbon dating: fictitious results
with mollusk shells. Science, 141:634-637, 1963.
71. Keith, M.L., G.M. Anderson, and R. Eichler. Carbon and oxygen
isotopic composition of mollusk shells from marine and freshwater
environments. Geochimica et Cosmochimica Acta, 28:1757-1786, 1964.
72. Klein, L.A., M. Lang, N. Nash, and S. Kirschner. Sources of metals
in New York City waste waters. Department of Water Resources Report,
City of New York, Journal of Water Pollution Control, 46 (12):2653-2662,
1974.
73. Klinkhammer, G.P- The distribution and partitioning of some trace
metals in the Hudson River Estuary. Master's Thesis, University
of Rhode Island, Kingston, Rhode Island, 1977, 136 pp.
74. Kneip, T.J. Cadmium in an aquatic ecosystem: distribution and
effects. First Annual Progress Report to the National Science
Foundation, New York University Institute for Environmental Medicine,
New York, New York, 1974.
75. Kneip, T.J. Cadmium in an aquatic ecosystem: distribution and
effects. Second Annual Progress Report to the National Science
Foundation, New York University Institute for Environmental Medicine,
New York, New York, 1975.
76. Kneip, T.J., T. Hernandez, and G. Re. Cadmium in an aquatic
ecosystem: transport and distribution. In: Trace contaminants
in the Environment, Proceedings of the Second Annual NSF-RANN Trace
Contaminants Conference, Asilomar, Pacific Grove, California, 1974.
199
-------
77. Kolle, W., H. Ruff, and L. Stieglitz. Die belastung des Rheins
mit organischen schadstoffen. Naturwissenschaften, 59:299-305, 1972.
78. Koyama, T. Gaseous metabolism in lake sediments and paddy soils and
the production of atmospheric methane and hydrogen. Journal of
Geophysical Research, 68:3971-3973, 1963.
79. LeBlond, P.H. Gas diffusion from ascending gas bubbles. Journal
of Fluid Mechanics, 35:711-791, 1969.
80. Levich, V.G. Physiochemical Hydrodynamics, Prentis Hall, New York,
1962.
81. Lewis, G.J., and N.W. Rakestraw. Carbohydrates in sea water. Journal
of Marine Research, 14:253-258, 1955.
82. Li, Y.H., J. Bishoff, and G. Mathieu. The migration of manganese in
the Artie Basin sediment. Earth and Planetary Science Letters, 7:
265-270, 1969.
83. Lovegren, J.R. Paleodrainage history of the Hudson Estuary. Master's
Thesis, Columbia University, New York, New York, 1974. 152 pp.
84. Lowman, F.G., O.K. Phelps, R. McClin, V.R. DeVega, I.O. DePodavani,
and R.J. Garcia. Interactions of the environmental and biological
factors on the distribution of trace elements in the marine environment.
In: Disposal of Radioactive Wastes into Seas, Oceans and Surface
Waters, International Atomic Energy Agency, Vienna, pp. 249-265, 1966.
85. Lynn, D.C., and Bonatti, E. Mobility of manganese in diagenesis of
deep sea sediments. Marine Geology, 3:457-474, 1965.
86. Malone, T.C. Environmental regulation of phytoplankton productivity
in the lower Hudson Estuary. Estuarine and Coastal Marine Science,
5:157-171, 1977a.
87. Malone, T.C. Phytoplankton productivity in the Apex of the New York
Bight: environmental regulation of productivity/chlorophyll a.
In: The Middle Atlantic Continental Shelf and New York Bight7 M.G.
Gross, ed., Limnology and Oceanography Special Symposium, 2, 1977b,
260 pp.
88. Martens, C. Control of methane sediment-water bubble transport by
macroinfaunal irrigation in Cape Lookout Bight, North Carolina.
Science, 192:998-1000, 1976.
89. Martens, C.S., and R.A. Berner. Methane production in the interstitial
waters of sulfate-depleted marine sediments. Science, 185:1167-1169,
1974.
200
-------
90. McCrone, W.W. The Hudson River Estuary: sedimentary and geochemical
properties between Kingston and Haverstraw, New York. Journal of
Sedimentary Petrology, 37:475-486, 1967.
91. Meade, R.H. Landward transport of bottom sediments in estuaries of
the Atlantic coastal plain. Journal of Sedimentary Petrology, 39:
222-234, 1969.
92. Mechalas, B.J. Pathways and environmental requirements for biogenic
gas production in the ocean. In: Natural Gases in Marine Sediment,
I.R. Kaplan, ed., Plenum Press, pp. 11-25, 1974.
93. Mook, W.G., and Vogel, J.C. Isotopic equilibrium between shells and
their environment. Science, 159:874-875, 1968.
94. Morel, F.M.M., J.C. Westall, C.R. O'Melia, and J.J. Morgan.
Fate of trace metals in Los Angeles County waste water discharge.
Environmental Science and Technology, 9:756-761, 1975.
95. Morris, L.D. Quantitative determination of carbohydrates with
Dreywood's anthrone reagent. Science, 107:254-255, 1948.
96. Newman, W.S., D.L. Thurber, H.S. Zeiss, A. Rokach, and L. Musich.
Late Quaternary geology of the Hudson River Estuary: A preliminary
report. Trans. New York Academy of Science, 31 (2) :548-570, 1969.
97. Nissenbaum, A., B.J. Presley, and I.R. Kaplan. Early diagenesis in
reducing fjord, Saanich Inlet, British Columbia, I. Chemical and
isotopic changes in major components of interstitial water.
Geochimica et Cosmochimica Acta, 36:1007-1027, 1972.
98. Noshkin, V.E., and V.T. Bowen. Concentrations and distributions of
long-lived fallout radionuclides in open ocean sediments. Radioactive
Contamination of the Marine Environment, IAEA, Vienna, pp. 671-686,
1973.
99. O'Connor, D.J. Water quality analysis for the New York Harbor complex.
In: Water Pollution in the Greater New York Area, A.A. Johnson, ed.,
Gordon and Breach, New York, 1970. 121 pp.
100. O'Connor, D.J. and W.E. Dobbins. Mechanism of reaeration in natural
streams. Trans. Am. Soc. of Civil Eng., 123:641-666, 1958.
101. Oliver, B.J. Heavy metal levels at Ottawa River sediments.
Environmental Science and Technology, 7:135-137, 1973.
102. Olsen, C.R., H.J. Simpson, R.F. Bopp, S.C. Williams, T.H. Peng,
and B.L. Deck. A geochemical analysis of the sediments and
sedimentation in the Hudson Estuary, Journal of Sedimentary Petrology,
1978 (in press).
201
-------
103. Owens, J.P., K. Stefansson, and L.A. Sirkin. Chemical, mineralogic
and palynologic charater of the Upper Wisconsin-Lower Holocene fill
in parts of Hudson, Delaware and Chesapeake estuaries. Journal of
Sedimentary Petrology, 44:390-408, 1974.
104. Pamatmat, M.M., and K. Banse. Oxygen consumption by the seabed.
II. In situ measurements to a depth of 180 m. Limnology and
Oceanography, 14:250-259, 1969.
105. Panuzio, F.L. Lower Hudson River siltation. In: Proceedings
of the Federal Inter Agency Sedimentation Conference, Agricultural
Research Service, Miscellaneous Publication No. 970, pp. 512-520, 1963.
106. Peng, T.H. Determination of gas exchange rates across sea-air
interface by the radon method. Ph.D. Thesis, Columbia University,
New York, New York, 1973. 324 pp.
107. Quirk, Lawler and Matusky, Engineers. Hudson River water quality
and waste assimilative capacity study, a final report submitted to
the State of New York, Department of Environmental Conservation,
1970, 198 pp.
108. Rafter, T.A. and B.J. O'Brien. Radiocarbon in the atmosphere and in
the South Pacific ocean. In: Proceedings of the 8th International
Conference on Radiocarbon Dating, The Royal Society of New Zealand,
Wellington, pp. 241-266, 1972.
109. Rapaire, J.L., and G. Hugues. Monaco radiocarbon measurements, V.
Radiocarbon, 19:44-61, 1977.
110. Reeburgh, W.S. Observations of gases in Chesapeake Bay sediments.
Limnology and Oceanography, 14:368-375, 1969.
111. Reeburgh, W.S. Methane consumption in Cariaco Trench waters and
sediments. Earth and Planetary Science Letters, 28:334-337, 1976.
112. Reeburgh, W.S., and D.T. Heggie. Microbial methane consumption
reactions and their effect on methane distributions in freshwater
and marine environments. Limnology and Oceanography, 22:1-9, 1977.
113. Reese, E.T., and M. Mandels. In: Methods in Carbohydrate Chemistry,
Vol. Ill, Academic Press, New York, 1963.
114. Riel, G.K. The distribution of fallout cesium-137 in the Chesapeake
Bay. In: Proceedings of the Second International Symposium on the
Natural Radiation Environment, J.A.S. Adams, W.M. Lowder, and T.F.
Gesell, eds., pp. 883-896, 1972.
115. Ritchie, J.C., P.H. Hawks, and J.R. McHenry. Deposition rates in
valleys determined using fallout cesium-137- Geol. Soc. Amer. Bull.,
86:1128-1130, 1975.
202
-------
116. Rogers, M.A. Carbohydrates in aquatic plants and associated
sediments from two Minnesota lakes. Geochimica et Cosmochimica Acta,
29:183-200, 1965.
117- Rosen, A.A., and M. Rubin. Natural carbon-14 activity of organic
substances in streams. Science, 143:1163-1164, 1964.
118. Rubin, M., and D.W. Taylor. Radiocarbon activity of shells from
living clams and snails. Science, 141:632, 1963.
119. Rudd, J.W.M., R.D. Hamilton, and N.E.R. Campbell. Measurement of
microbial oxidation of methane in lake water. Limnology and
Oceanography, 19:519-524, 1974.
120. Sackett, W.M., and W.S. Moore. Isotopic variations of dissolved
inorganic carbon. Chemical Geology, 1:323-328, 1966.
121. Sanders, J.E. Geomorphology of the Hudson Estuary. In: Hudson
River Colloquium, Ann. New York Academy of Science, 250:5-38,
1974.
122. Schink, D.R., N.L. Guinasso, and K.A. Fanning. Processes affecting
the concentration of silica at the sediment-water interface of the
Atlantic Ocean. Journal of Geophysical Research, 80:3013, 1975.
123. Segar, D.A., and A.Y. Cantillo. Trace metals in the New York
Bight. Am. Soc. Limnology and Oceanography Spec. Symp., 2:171-198,
1976.
124. Simmons, H.B., and F.A. Herrmann. Effects of man-made works on the
hydraulic, salinity and shoaling regimes of estuaries. In:
Environmental Framework of Coastal Plain Estuaries, B.W. Nelson,
ed., Geological Society of America Memoir, 133:555-570, 1972.
125. Simpson, H.J., R. Bopp, and D.L. Thurber. Salt movement in the
lower Hudson, Ch. 9. In: Third Symposium on Hudson River Ecology,
Hudson River Environmental Society, New York, 1974.
126. Simpson, H.J., and D.E. Hammond. Application of one-dimensional
transport models to circulation in the Hudson Estuary. Submitted
to the Journal of Estuarine and Coastal Research, 1977.
127. Simpson, H.J., D.E. Hammond, B.L. Deck, and S.C. Williams. Nutrient
budgets in the Hudson River Estuary. In: Marine Chemistry in the
Coastal Environment, Amer. Chem. Soc. Sym. Series, T.M. Church, ed.,
18:616-635, '1975.
128. Simpson, H.J., C.R. Olsen, T.H. Peng, and S.C. Williams. Radiocarbon
dating of Hudson Estuary cores. In preparation.
203
-------
129. Simpson, H.J., C.R. Olsen, R.M. Trier, and S.C. Williams. Man-made
radionuclides and sedimentation in the Hudson River Estuary. Science,
194:179-183, 1976.
130. Simpson, H.J., S.C. Williams, C.R. Olsen, and D.E. Hammond.
Nutrient and particulate matter budgets in urban estuaries.
Estuaries, Geophysics and the Environment, National Academy
of Sciences, Washington, B.C., pp. 94-103, 1977-
131. Solozono, L. Determination of ammonia in natural waters by the
phenolhypochlorite method. Limnology and Oceanography, 14:799, 1969.
132. Spiker, E.G. Carbon isotope variations in San Francisco Bay.
Trans. American Geophysical Union, 4:255, 1976.
133. Spiker, B.C., and M. Rubin. Petroleum pollutants in surface and
groundwater as indicated by the carbon-14 activity of dissolved
organic carbon, 61-64, 1975.
134. Standard Methods, 13th edition. American Public Health Association,
New York, 1971.
135. Stommel, H. Computation of pollution in a vertically mixed estuary.
Sewage and Industrial Wastes, 25:1065-1071, 1953.
136. Stookey, L.L. Ferrozine - A new spectrophotometric reagent for iron.
Analytical Chemistry, 42 (7):779-781, 1970.
137. Strickland, J.D.H., and T.R. Parsons. A practical handbook of
sea water analysis. Bull. Fisheries Research Board of Canada,
167, 1968, 311 pp.
138. Swinnerton, J.W., V.J. Linnenbom, and C.H. Cheek. Determination of
dissolved gases in aqueous solutions by gas chromatography.
Analytical Chemistry, 34:483-485, 1962.
139. Taylor, S.R. Abundance of chemical elements in the continental crust:
A new table. Geochimica et Cosmochimica Acta, 28:2173-2185, 1964.
140. Thatcher, M.L., and D.R.F. Harleman. A mathematical model for the
prediction of unsteady salinity intrusion in estuaries. Technical
report No. 144, Ralph M. Parsons Laboratory for Water Resources and
Hydrodynamics, Department of Civil Engineering, Massachusetts
Institute of Technology, 1972.
141. Thomas, R.L. The distribution of mercury in the sediments of Lake
Ontario. Canadian Journal of Earth Sciences, 9:636-651, 1972.
142. Thomas, R.L. The distribution of mercury in the surficial sediments
of Lake Huron. Canadian Journal of Earth Sciences, 10:194-204, 1973.
204
-------
143. Tsivoglou, E.G. Relative gas transfer characteristics of krypton
and oxygen. In: Symposium on Direct Tracer Measurement of the
Reaeration Capacity of Streams and Estuaries, U.S. Environmental
Protection Agency, Water Pollution Control Research Series,
16950:19-30, 1972.
144. Tsou, J.L., D. Hammond, and R. Horowitz. Interstitial water studies,
Leg 15 - Study of CO released from stored deep sea sediments.
In: Initial Reports of the Deep Sea Drilling Project,, B.C. Heezen,
I.D. MacGregor, eds., Washington, D.C., 20:851-863, 1973.
145. Turekian, K.K., and K.H. Wedephol. Distribution of elements in the
earth's crust. Geological Society of America Bulletin, 72:174-192,
1961.
146. Turner, W.R. Physics of microbubbles. Report Tech. Note 01654-01-2.
Vitro Laboratories, Silver Springs, Maryland, 1963.
147- Volchok, H.L. The global strontium-90 budget. Journal of Geophysical
Research, 71:1515-1518, 1966.
148. Volchok, H.L., and M.T. Kleinman. Global Sr-90 fallout and
precipitation: Summary of the data by 10 degree bands of latitude,
USAEC Rep. HASL-245, 2-83, 1971.
149. Wakeham, S.G., and R. Carpenter. Aliphatic hydrocarbons in sediments
of Lake Washington. Limnology and Oceanography, 21:711-723, 1976.
150. Wahlgren, M.A., and J.S. Marshall. The behavior of Plutonium and
other long-lived radionuclides in Lake Michigan: I. Biological
transport, seasonal cycling and residence times in the water column.
In: International Symposium on Transuranium Nuclides in the
Environment, IAEA-SM-198/39, IAEA, Vienna, 1975, 227 pp.
151. Weiss, D. Late Pleistocene stratigraphy and paleoecology of the lower
Hudson River Estuary. Bull. Geological Society of America, 85:1561-
1570, 1974.
152. Weiss, R.F., and H. Craig. Precise shipboard determination of
dissolved nitrogen, oxygen, argon, and total inorganic carbon by
gas chromatography. Deep Sea Research, 20:291-303, 1973.
153. Whittenbury, R., K.C. Phillips, and J.F. Wilkinson. Enrichment,
isolation and some properties of methane-utilizing bacteria. Journal
of General Microbiology, 61:205-218, 1970.
154. Windom, H.L., K.C. Beck, and R. Smith. Transport of trace metals
to the Atlantic Ocean by three southeastern rivers. Southeast
Geology, 12:169-181, 1971.
155. Windom, H.L. Preprint. 1977.
205
-------
156. Wong, K.M. Radiochemical determination of plutonium in sea-water,
sediments and organisms. Anal. Chim. Acta, 56:355-364, 1971.
157. Worzel, J.L., and C.L. Drake. Structure section across the Hudson
River at Nyack, New York, from seismic observations. Annals
New York Academy of Science, 80:1092-1105, 1959.
158. Wrenn, M.E., J.W. Lentsch, M. Eisenbud, G.J. Laver, and G.P. Howells.
Radiocesium distribution in water, sediment and biota in the
Hudson River estuary from 1964 through 1970. In: Proceedings of
the Third National Symposium on Radioecology, Oak Ridge, Tennessee,
1971.
159. Wyman, J., P.P. Scholahder, G.A. Edwards, and L. Irving. On the
stability of gas bubbles in seawater. Journal of Marine Research,
11:47-62, 1952.
160. Zafiriou, O.C. Petroleum hydrocarbons in Narragansett Bay, 2.
Chemical and Isotopic Analysis. Estuarine Coastal Marine Science,
1:81-87, 1973.
161. Zein-Eldin, A.P. and B.Z. May. Improved N-ethylcarbazole determination
of carbohydrates with emphasis on sea water samples. Analytical
Chemistry, 30:1934-1941, 1958.
206
-------
TECHNICAL REPORT DATA
(Please read /nztiiictions on the reverse before completing)
. REPORT NO.
EPA-600/3-79-029
3. RECIPIENT'S ACCESSI ON- NO.
4. TITLE AND SUBTITLE
DREDGE SPOILS AND SEWAGE SLUDGE IN THE TRACE
METAL BUDGET OF ESTUARINE AND COASTAL WATERS
5. REPORT DATE
March 1979 issuing date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO
H. James Simpson
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Lamont-Doherty Geological Observatory
of Columbia University
Palisades, New York 10964
10. PROGRAM ELEMENT NO.
1BA819
11. CONTRACT/GRANT NO.
803113
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Research Lab. - Narragansett, RI
Office of Research and Development
U.S. Environmental Protection Agency
Narragansett, Rhode Island 02882
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/600/05
15. SUPPLEMENTARY NOTES
16. ABSTRACT ~ ' ~~—~ ~~~
Many reactive pollutants, such as Zn, Cu, Pb, Cs-137, Pu-239, 240 and PCB's
appear to be transported and accumulated together in association with fine-grained
particles in the Hudson River estuary. Anthropogenic increases of 3-6 times natural
levels of Zn, Cu, and Pb were found for Hudson sediments. Mobilization of Cd and Ni
in the sediments of a small embayment of the Hudson with very high contamination
levels appears to be primarily by resuspension of fine particles, although elevated
concentrations of Cd in pore waters were also observed. Radiocarbon measurements
indicate the predominant source of organic carbon in New York harbor sediments is
recent sewage and not petroleum hydrocarbon contamination. A new enzymatic tech-
nique was developed to trace the distribution of cellulose, a significant component
of sewage sludge, in coastal sediments. Radon-222, a natural radioactive gas dis-
solved in the Hudson, is supplied primarily from the sediments at approximately twice
the rate predicted by molecular diffusion. Methane measurements provided additional
information on the flux of materials from sediments. The behavior of phosphate and
trace metals derived from sewage was examined on the basis of field data and the
use of simple models to examine management alternatives. The most reasonable course
appears to be completion of secondary sewage treatment plants in New York City and
major upgrading of primary treatment in New Jersey.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Geochemistry
Sediments
Radioactive isotopes
jredge spoil
Coastal waters
b.IDENTIFIERS/OPEN ENDEDTERMS
COSATI Field/Group
08/A
08/D
08/H
3. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
223
20. SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
/O/
• U. S. GOVERNMENT PRINTING OFFICE: 1979 — 657-060/1™
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Environmental Research Information Center
Cincinnati, Ohio 45268
OFFICIAL BUSINESS
PENALTY FOR PRIVATE USE. S3OO
AN EQUAL OPPORTUNITY EMPLOYER
POSTAGE AND FEES PAID
US ENVIRONMENTAL PROTECTION AGENCY
EPA-335
Special Fourth-Class Rate
Book
If your address is incorrect, please change on the above label
tear off, and return to the above address.
If you do not desire to continue receiving these technical
reports, CHECK HEREL~H, tear off label, and return it to the
above address.
EPA-600/3-79-029
------- |