MARINE
ENVIRONMENTAL
RESEARCH
ELSEVIER
Marine Environmental Research 51 (200 I) 389-415
www.elsevier.comfloca tefmarenvrev
Changes in the metal content of surficial
sediments of Boston Harbor since the cessation
of sludge discharge
C. Zago a,*, A. E. Giblin b, A. Bergamasco c
•unil•ersiry of Venice, Department of Environmental Sciences, Sr. Marra 2137, Venice. Jraly
bTJre Ecosystems Center, Marine Biological Laboratory, Woods Hole, MA , 02543 USA
clsriruro per lo Studio della Dinamica delle Grandi Masse, C.N.R .. S. Polo 1364, Venice, Italy
Received 3 February 1999; received in revised fonn 18 April 2000; accepted 25 May 2000
Abstract
\
Temporal trends of metals in surficial sediments (1991- 1998) at two sites in Boston Harbor
were a nalyzed to evaluate the effect of stopping sludge dumping in December 1991. Metal
contents of sediments from the o ld sludge disposal site were higher than those of a station in
the central Harbor. Since 1991, carbon, copper, and lead contents have significantly decreased
in sed iments from the disposal site. Chromium and Zn have shown smaller decreases while Fe,
a nd Mn, have remained relatively constant. Metal content in the central Harbor station,
located in an area of sediment reworking, has been quite variable, but, with the exception of
Zn which has shown a large decrease relative to iron, the changes seemed to be well correlated
with changes in the organic carbon content at this site due to resuspension. Ratios of metals in
the sediments are fairly similar at both sites and similar to those of sewage-derived particles,
with the exception of Cr, which appears to be enriched in the sediments.
200 I Elsevier
Science Ltd. All rights reserved.
Keywords: Metals; Sediments; Boston Harbor: Contamination; Benthic environment
l. Introduction
Sediments of coastal environments play an important role in nutrient and metal
cycling. Coastal sediments may act as a sink of pollutants but the eventual fate of
• Corresponding author.
E-mail address: [email protected] (C. Zago).
0141-1136/01 /$- see front mauer
Pll: SO 141-1136(00)00 135-5
2001 Elsevier Science Ltd. All rights reserved.
390
C. Zago et al. / Marine Environmental Research 51 (200 1) 389-415
contaminants depends upon a number of factors including sediment resuspension
and lateral transport, biogeochemical transformatio n processes, bioturbation, and
partitioning between sediment and pore water (Bothner, Buchholtz ten Brink &
Manheim, 1998; Shea & Kelly, 1992; Williamson, van Dam, Bell, Malcolm & Kim,
1996). Heavy metal contamination in aquatic and sediment environments is of critical concern, due to the toxicity of metals and their accumulation in the aquatic
habitat (Forstner & Wittman, 1979). Ecosystem managers are sometimes faced with
the question as to how long polluted estuarine systems take to recover after contaminant inputs are controlled or reduced. Heavy metals, in contrast to most pollutants, are not biodegradable and they accumulate in coastal sediments. It is the
rate of self-cleansing mechanisms of the polluted sediments, such as burial and
mobilization/dispersal that governs this recovery time (Williamson et al., 1996).
Boston Harbor, Massachusetts (Fig. I), has received high rates of metal loading
from a variety of sources including, municipal wastewater effluents, combined sewer
overflows, runoff from an airport, tributaries, and atmospheric sources (Alber &
Chan, 1994; Hunt, West & Peven, 1995). Sewage sludge was also directly disposed in
the Harbor for more than 100 years until the practice was stopped in December of
1991. This high contaminant loading has resulted in some of the highest sediment
~
A
Deer bland
Effluent Outfalb
,
ut Island
Slud Outfall
.,_
*
**
Nut IslaDd
.,
Effluent Outfalb
•
....
·-
.I
<
I
..
~-
Fig. I. Map of Boston Harbor showing the two sampling sites and the location of Lhe sewage discharge
C. Zago et al./ Marine Environmental Research 51 (2001) 389-415
.
-~
\.
391
metal concentrations in the USA, high levels of disease in some fish populations and
an altered benthic community (Shea & Kelly, 1992).
In Boston Harbor, organo-metallic compounds and iron oxide coatings on suspended particles and bottom sediment appear important in the transport and retention of metals coming from different sources (Fitzgerald, 1980). Calculations suggest
that at least 3% of the metals discharged into the outer harbor are retained. By
trace-metal profiles in the bottom sediments Fitzgerald (1980) found strong evidence
for an increasing storage of metals in sediments during the past 100 years.
Metal inputs in Boston Harbor began to decrease in the 1980s due to improved
industrial treatment, and more recently to an improved sewage treatment. Bothner
et al. (1998) examined sites throughout Boston Harbor and found that the concentrations of metals in surface sediments have decreased from 1977 to 1993. The
similarity of metal profiles in cores from different areas of Boston Harbor suggested
to the authors that contaminant accumulation processes are similar throughout the
region, although their rates may be different. Bothner et al. (1998) concluded that
the cause of the decreasing concentrations in sediments was presumed to be the
decreasing flux of metals to the harbor from industrial point sources, sewage, street
runoff, combined sewer overflows, and the atmosphere. A major reduction of metal
inputs occurred after 1991 when sludge discharge to the harbor ceased, but the
sampling interval used by Bothner et al. (1998) did not allo~ them to assess how
the changes were related to this event.
Wallace et al. (1991) examined sub- and intertidal sites in a more localized area of
Boston Harbor, Savin Hill Cove, to determine sources of metals in the sediments
and temporal changes. They concluded that this site was highly influenced by the
deposition of contaminated particles of sewage origin. The concentrations of most
metals, excluding those in alumino-silicate minerals, were highly correlated with the
organic carbon content of the sediments. Vertical profiles of metals in sediments
from rapidly sedimenting areas indicated that there may have been some modest
decreases in metal inputs but only Cu and Cd showed strong evidence of reductions
over the last 15 years at this site.
There have been few studies which have closely followed the short-term response
of metals in sediments to a cessation of loading. Zdanowicz, Cunneff and Finneran
(1995) examined the reduction of metals in sediments at and near the 12-mile
dumpsite in the New York Bight Apex. They found that metal concentrations in
the dumpsite area decreased by a factor of two within one month of the cessation of
dumping, while nearby moderately-contaminated stations did not significantly
change over the study period.
The purpose of this study was to observe how metals and organic matter in the
surface sediments of Boston Harbor changed over time following the cessation of
sludge discharge to Boston Harbor. We sampled two sites, both with muddy sediments, but with different sedimentary environments, intensively over 8 years. Station
BH03 was the site of sludge disposal until December 1991. This station is located in
a depositional environment off Long Island, Massachusetts (Knebel & Circe, 1995),
and within 1.4 km of a combined sludge and effluent discharge site from the Deer
Island treatment plant (Fig. 1). Sediments are fine-grained and flocculent. This site
392
C. Zago et a/.j Marine Environmental Research 51 (2001) 389-415
was selected because historically the major inputs of metals and organic matter to
this site were from direct sludge disposal. Since the cessation of sludge disposal,
from December 1991, large numbers of benthic amphipods and other benthic
organisms have been routinely observed at this site (Giblin, Hopkins & Tucker,
1993, 1997; Kropp and Diaz, 1995). An analysis of the sediment using 8 15N indicated that the sludge contribution of the sediments to nitrogen at this station was
44% in 1991 and decreased to about 19% in 1994 (Tucker, Sheats, Giblin Hopkinson & Montoya, 1999). The decrease was attributed to the loss of N through
mineralization of organic matter and to the incorporation on non-sewage particles.
The other Station BH02, also with fine-grained sediments, is located 3.5 km from the
Deer Island discharge site (Fig. 1) in an environment of sediment reworking (Knebel
& Circe, 1995) close to Governor Island Flats. This station receives heavy metals
from a variety of non-point sources as well as sewage-derived particles from sludge
and effluent disposal. Station BH02 is closer than BH03 to combined sewer overflows which carry raw sewage during periods of heavy rain. During the study period
the site was often devoid of macrofauna, although amphipods were observed here in
summer 1994. The percentage of nitrogen derived from sewage at this station,
determined by 815N analysis, decreased less than at BH03, from 45 to 35% from
1991 to 1994 (Tucker et al., 1999).
2. Materials and methods
2.1. Sampling
Stations (Fig. 1) were located using a differential global positioning system. Station locations using this system should be accurate to within 10m.
Samples were collected by SCUBA divers. At each station several sized cores were
taken. Replicate 2.5-cm cores were collected for porosity, metal content, and total
organic C and N content in surficial sediments from 1991 to 1998. Cores for porewater sampling in July 1994 were 6.5 em in diameter and 1 m long. Separate 6.5 em
diameter and 20 em long cores were simultaneously taken, for redox potential (Eh),
metals content, and HCl extractable acid volatile sulfides (A VS; Di Toro et al., 1992).
At Station BH03, two replicate cores were collected in July 1994 at a distance of
about 3 m apart, in order to examine spatial variability in pore waters.
2.2. Sediment analysis
Heavy metals content (HC1/HN0 3 digestion), and carbon and nitrogen in sediments were analyzed from samples collected from the two sites during 1991-1998.
Surface sections of 0-2 em depth were collected for a total of 12 metal samples and
14 C and N samples in BH02, and 16 metal samples and 19 C and N samples in
BH03. Core for metal profiles were sectioned in 1 em intervals to a depth of 10 em
and then in 2 em intervals to the bottom of the core. Samples for metals and C and
N analyses were dried, ground, and stored in small glass vials until analysis. C
C. Zago eta!./ Marine Environmental Research 51 (2001) 389-415
393
and N analysis was performed within a few weeks of sampling. Heavy metals content in dried sediments collected from 1991 to 1994 was measured in 1995, samples
collected in 1998 were analyzed in 1999. Metal content of dried sediments does not
change over time.
To measure metals present in different fractions, cores collected in July 1994
were analyzed with cold HCl and with hot HN0 3 /HC1 digestion. The HCl extraction should include the metals found in metal oxides, monosulfides and metal
loosely bound to clay. The combined hot HN0 3 /HC1 digestion extracts pyrite and
metals bound to clays and organic matter in addition to the fractions extracted by
cold HCI. This digestion will not get all metals strongly bound within clay lattices,
but does largely remove the metals of anthropogenic origin (Forstner & Salomons,
1980). Bothner et a!. (1998) found that for fine-grained sediments from urban
environments, such as Boston Harbor sediments, a concentrated nitric-acid leach
yielded similar concentrations to complete digestion using HF, HN0 3 , and HC10 4
acids for most metals of anthropogenic origin.
For the cold HCl digestion, 0.1 g of dried sediment for Fe and Mn analysis and
0.5 g for Cr, Cu, Pb and Zn were added with 7.5 ml of 5.5 N HCl and shaken
overnight. The HN0 3 /HC1 digestion was performed by adding 5 ml of concentrated
HN0 3 and then heating the samples in a water bath to about 70°C for 2 h;
when samples cooled down, 5 ml of concentrated HCl were added to all tubes and
again samples were heated for 2 h. Samples were then diluted and filtered. All samples were analyzed with a Perkin-Elmer 2380 atomic absorption spectrometer
(AAS) at the Marine Biological Laboratory.
The A VS concentrations of the sediments collected in 1994 were analyzed by the
Di Toro eta!. (1992) method. For the analysis, 20 g of wet sediments were added to
a distillation apparatus and N 2 was allowed to flush for 20 min. Acid was added
to the wet sediment samples to achieve a final concentration of 0.5 M HCI. The
A VS was distilled over and trapped with Zn acetate. To achieve better mixing of
acid with the sediment, we modified the extraction technique of Di Toro eta!. (1992)
by adding the acid to the sediment and heating for 1 h.
Organic carbon and nitrogen analysis was performed on a Perkin-Elmer 2400 CHN
elemental analyzer following carbonate removal. To remove carbonates, sediments
were placed in a dessicator over fuming HCl (Kristensen & Andersen, 1987). The
sediment was then dried, weighed and ground. The percentage of C and N measured
on the sediment was corrected for the weight change due to the procedure.
2.3. Pore water sampling and analysis
To avoid metal contamination, all cups and bottles were kept in HN0 3 diluted
( 1: 100) in distilled water for at least 24 h, and then rinsed with distilled water.
Finally, each bottle was sealed in plastic bags until the moment of analysis. To avoid
contact of the sample with atmospheric oxygen (Lyons, Gaudette & Smith, 1979), all
treatments were carried out in nitrogen atmosphere utilizing a glove bag.
For pore water metals concentrations analysis along the vertical profile, cores
were cut in 1-2 em sub-samples (corresponding to the following depths: 0-1, 1-2,2-4,
394
C. Zago et al.f Marine Environmental Research 51 (2001) 389-415
4--6, 6-8, 8-10, 10-12, 16-18, 18-22 em). Pore waters were removed from muddy
sediments by centrifuging sections for 15 min at 8000 rpm (International Clinical
centrifuge) in centrifuge tubes capped under nitrogen. Pore water was filtered with
0.4 Jlm polycarbonate filters (Nuclepore) in nitrogen atmosphere. Filtered pore
water was immediately acidified with HN0 3 Suprapur, and subdivided to compare
three different types of analysis: particle induced x-ray emission (PIXE), inductively
coupled plasma atomic emission spectrometry (ICP-AES), or atomic absorption
spectrophotometry with graphite furnace (GFAAS). Samples were stored acidified
at 4°C until the moment of analysis.
PIXE analysis is a multi-elemental technique that allowed analyzing metal traces,
utilizing a 2 MeV Van de Graaf accelerator of the INFN Laboratory in Legnaro
(Padua, Italy). Thin targets prepared by heavy metals precipitation as carbamates
on Nuclepore filters were exposed to a 1.8 MeV proton beam (Cecchi, Ghermandi
& Zonta, 1990). Emitted x-ray spectra were detected and analyzed to obtain concentration data with errors generally less than 10%. The ICP-AES instrument used
was a Perkin-Elmer Optima 3000 of the Battelle Ocean Sciences laboratories,
Duxbury, MA. The GFAAS was a Perkin-Elmer model 5100 of the Battelle Ocean
Sciences laboratories, Duxbury, MA. In contrast to PIXE, these latter two techniques do not need any sample treatment after acidification. In order to test PIXE
accuracy at low-metal concentrations, the comparison among the three different
techniques was performed in site BH02. Samples from BH03 were only analyzed
by PIXE.
To analyze dissolved sulfides and pH, cores were cut at the following depths: 0-l,
1-2, 2-4, 4--6, 6-8, 8-10, 11-13, 15-17 em. Pore water was immediately sub sampled
for the different analysis after centrifugation. For sulfide analysis, 0.1 ml of filtered
sample was instantaneously fixed in 2% zinc acetate to prevent hydrogen sulfide
losses and analyzed within 12 h using the method of Cline (1969). Samples for pH
were analyzed immediately.
Sediment Eh was measured using a platinum electrode (Bohn, 1971). The electrode was pushed into the sediment at l.0-3.0 em increments and allowed to stabilize for 15 min (Giblin et al., 1997).
2.4. Chemical equilibrium model
We utilize the computer chemical equilibrium model MINEQL+ (Schecher &
McAvoy, 1994) to calculate the activity of the metals in pore waters. The most fundamental assumption of the chemical equilibrium calculation is that equilibrium
exists in the system chosen for that study. Therefore, saturation index computation
can only indicate the tendency for the precipitation or dissolution of a mineral
(Nordstrom et al., 1979). The presence of saturation of a mineral phase, calculated
by this model, does not mean that the mineral is present in the sediment, but we can
consider it as thermodynamically favored to precipitate.
The sulfide, carbonate and hydroxide saturation indexes, defined as:
SI = logQ/Ks
C. Zago eta/. f Marine Environmental Research 51 (2001) 389-415
395
where Q is the ion activity product and Ks is the solubility constant for the solid,
were calculated to test whether or not metals should be expected to be precipitated
in sediments. The ion activity coefficient was corrected for the ionic strength within
the program using the Davidson formula.
Because we did not analyze pore water concentrations ofNa, Cl, K, Mg, Ca, these
values were obtained by salinity values (Burton & Liss, 1976; Olausson & Cato,
1980) and kept constant with depth. The metal complexes considered and the stability constants used are those of the model MINEQL +.
3. Results and discussion
3.1. Temporal trends of metals in surficial sediments
Metal concentrations in sediments at the two stations measured were high and
generally typical of those of highly-contaminated environments (Angelidis &
Aloupi, 1995; Bothner et al., 1998; Feng, Cochran, Lwiza, Brownawell & Hirschber,
1998; Forstner & Wittman, 1979; Haynes, Raymen & Toohey, 1996; Olausson &
Cato, 1980; Scarponi et a!., 1998 Tam & Wong, 1995; Williamson et al., 1996),
although, a direct comparison of sediment metal data of different environments is
difficult because heavy metals concentrations vary from area to area depending on
sediment characteristics and particle size distribution (Barreiro, Real & Carballeira,
1994; Pardue, De Laune & Patrick, 1992). Our average metal values for Station
BH02 were very similar to those measured by Bothner et a!. (1998) taken in 1993
from two nearby stations in Boston Harbor.
The heavy metals concentrations (Cr, Cu, Fe, Mn, Pb, and Zn) measured in surficial samples of the two sites from 1991 to 1998 are shown in Fig. 2 and Table 1.
Prior to the cessation of sludge disposal in 1991, the concentrations of all the metals
measured at Station BH03, the old sludge disposal area, were higher than at Station
BH02. After the discharge ceased, the concentrations Fe and Mn at Station BH03
stayed relatively constant while Pb and Cu, and to a lesser extent, Cr and Zn
decreased. Metal concentrations at Station BH02 showed large fluctuations on a
month to month basis, and metal concentrations from August 1992 until May 1993
were higher than in September 1991, 3 months before discharge had ceased.
Sediment organic carbon and nitrogen show a similar pattern to metals (Fig. 3).
The carbon content at Station BH02 shows large temporal fluctuations without a
clearly defined trend. This is an area of sediment reworking and the high variability
we observed is presumably due to changes in sediment composition caused by sediment resuspension and deposition from storms. Even though organic carbon from
sewage decreased by 20% after sludge discharge ceased (Bothner et a!., 1998; Hunt
eta!., 1995), we did not see a consistent decrease in the organic carbon content in the
sediment at this site. Perhaps a more long-term analysis will be able to observe a
decrease of organic carbon in areas of sediment reworking. Station BH03, with
a higher organic carbon content, is located in a depositional area and does not
experience sediment reworking. The decrease in carbon and nitrogen content since
----------
396
---
------
C. Zago et al./ Marine Environmental Research 51 (2001) 389-415
Table 1
Heavy metal concentrations (11molejkg d.w.) obtained from 1991 to 1998 surficial samples in the two
stations
Fe
Mn
Cu
Cr
Zn
Pb
BH02
Sept. 91
Aug. 92
Feb. 93
May 93
July 93
Aug. 93
Oct. 93
Mar. 94
May94
July 94
Oct. 94
Oct. 98
221684
365 835
311782
388 457
273 958
392943
292 937
318 823
345 555
307 314
208 778
411 144
2858
5056
4429
5782
3601
5591
4067
4411
4645
4636
3128
5289
507
711
1278
1147
403
922
499
771
760
647
326
777
815
869
1652
1647
683
1483
956
1197
1192
1019
632
1350
1127
1748
2068
1760
1086
1687
1215
1254
1325
1122
653
1694
189
171
405
365
144
330
197
224
241
179
94
234
BH03
Sept. 91
Apr. 92
May 92
Jun. 92
Aug 92
Nov. 92
Feb. 93
May 93
July 93
Aug. 93
Oct. 93
Mar. 94
May 94
July 94
Oct. 94
Oct. 98
485 880
547 326
405960
470443
625 379
480117
438 394
518 792
458 873
496796
447097
472224
509 730
455070
443 821
553 443
5903
5863
5508
5944
5735
7422
6512
8117
7413
9407
6085
5937
6313
5464
6196
7675
2318
2610
2181
2493
2798
2317
1755
1664
1757
1709
1720
1483
1438
2026
1870
1349
2147
2281
2188
2168
2382
2213
2070
2116
1975
1939
1837
1642
1459
1943
1995
1983
2456
2527
2309
2236
2725
2254
1970
1990
2213
2088
2022
2361
1839
2081
627
679
646
628
733
598
571
556
601
655
637
639
601
602
593
489
2497
1991 is almost certainly due to the cessation of sludge disposal which is consistent
with the sediment 8 15 N data (Tucker eta!., 1999), which showed that the amount of
nitrogen from sludge had decreased by half over the same period. Similarly to carbon, metals such as Ph and Cu (Fig. 2), known for their affinity with particulate and
organic material, decreased in BH03 respectively of about 20 and 40% from September 1991 to October 1998. Zn and Cr showed smaller decreases ranging from 15
to 20% in the first 3 years of sampling; after that they stayed almost constant or
slightly increased (Zn) until late 1998.
The mean ratios among anthropogenic metals (Cu, Cr, Zn and Ph) and the metal
to carbon ratios calculated from 1991 to 1998 are fairly similar at the two stations
(Fig. 4). There is no significant difference between stations in the ratio of Cr, Zn,
Cu and Ph to carbon although metal carbon ratios at BH02 are higher for Cr and
Zn and lower for Cu and Ph than at Station BH03. When the metal/metal ratios are
compared, Cr and Zn are quite enriched relative to Ph at Station BH02 and Cr is
C. Zago eta!./ Marine Environmental Research 51 (2001) 389-415
397
2500
BH02
2000
~ 1500
'C
Cl
lll::
Iii
0 1000
E
:I
500
0
10
10
N
..,..,
0
N
..,..,"'
10
..,..,
00
G>
0
10
..,
~
10
....
....
....
..,
0
00
..,
0
10
..,~10
0
....
00
10
....
....
..,
..,"'
10
3000
2500
~
2000
-o
Cl
lll::
Iii
0
E
:I
1500
1000
~:
500
;
0
10
10
N
..,..,
0
N
..,..,"'
10
..,..,
00
G>
..,....
..,
0
10
10
....
..,........
0
00
..,
0
10
10
..,3
10
0
....
00
..,
10
1--cr ---cu --.-zn --Pb --iiE-Fe/1000 --Mn/10 I
10
....
....
"'..,
Fig. 2. Heavy metals concentrations measured in surface sediments of site BH02 and BH03 from 1991 to
1998. Values are t-tmole/kg dry weight.
somewhat enriched relative to Cu. Presumably, both stations were highly influenced
by sludge discharge as several studies have shown that sewage-derived particles
are transported throughout the harbor (Tucker et a!., 1999; Wallace et a!., 1991).
However, at Station BH02, mean metal/metal and metal/carbon ratios show a larger
variability (error bars in Fig. 4) than at Station BH03. Particularly, mean ratios
obtained for Cr and Zn have the largest error bars. This may reflect the reworking of
sediments occurring in the area around Station BH02 (Knebel & Circe, 1995) and the
greater variety of sources of metal loading affecting this area (Alber & Chan, 1994).
C. Zago eta!./ Marine Environmental Research 51 (2001) 389-415
398
4
··--·--~-·-·~-------------·----------·-·--·-·-··-·--·--··---------
.
BH02
0.4
0.35
3
0.3
0.25
C%2
0.2 N%
0.15
0.1
0.05
0
....
~
c
Ill
....,
N
~
c
Ill
....,
...,
..,.
0
II)
~
~
~
Ill
....,
Ill
....,
....,Ill
c
c
c
CD
~
c
....,Ill
......
~
c
Ill
....,
GO
~
c
Ill
....,
en
~
c
....,Ill
5 ..---------------------·--··· -·--------···-----------··-··-····--·-----····------··· 0.6
0.5
4
0.4
3
C%
0.3N%
2
0.2
0.1
o+---,---,----,-----,-----,----,----.---+o
....
N
CD
......
GO
~
~
~
~c
~c
~
~
~c
c
c
c
c
c
Ill
Ill
Ill
Ill
....,Ill
....,Ill
....,Ill
....,
....,Ill
....,
....,
....,
1-c%·-•··N%1
Fig. 3. C% and N% values measured in. sediments from 1991 to 1998 at Stations BH02 and BH03.
The mean ratio of metals in sediments can be compared to those of both the
sludge being directly discharged to Station BH03 (Nut Island) and the sludge plus
effluent loading of the harbor (Fig. 5) to better determine the likely importance of
sources other than sludge and other sewage-derived particles. Until sludge disposal
ceased in 1991 we assumed that inputs at BH03 would be dominated by sludge from
Nut Island while Station BH02, would be more strongly influenced by the total
loading (sludge plus effluents). Loadings of 1979-1984, 1991 and 1993-1994 are
from MWRA data. For some 1979-1984 samples the Cr data are influenced by the
detection limits of the analysis (Murray Hail, personal communication). When
C. Zago eta/. / Marine Environmemal Research 51 ( 2001 ) 389-415
399
1000
0
800
'f
g 600
~
:::IE
400
200
Cr/C
Cu/C
Pb/C
Zn/C
7
6
0
~
5
!!•
4
i
3
~
:IE
2
Cr/Cu
Cr/Zrl
Cr/Pb
Cu/Zn
Cu/Pb
Zn/Pb
Fig. 4. Ratios of anthropogenic metals to carbon (top) and metal/metal ratios in surface sediments from
Stations BH02 and BH03. Error bars are the 95% confidence interval calculated from a ll 1991 to 1998
samples where metals were measured.
values were close or lower than the detection limit, loadings were calculated assuming the concentration equals the detection limit. For this reason, Cr concentrations,
and hence ratios involving Cr, calculated for the 1979-1984 period may be somewhat overestimated. However, with the exception of Cr ratios for the 1979- 1984
period, the ratio of most metals in sludge does not appear to be that different from
the ratio of metals found when effluent is included (Fig. 5). From 1979 to 1984
inputs of Cr were considerably higher than in the following years (the difference is
much greater than the possible error due to the detection limit), and the total loading due to effluents is 2- 3 times more enriched in Cr than in sludge.
The ratios of metals found in sewage inputs are very similar to those found in the
sediments, again, with the exception of Cr that is highly enriched in the sediments of
both stations. The Cr ratios in sediments are at least 10 times higher than one would
predict based upon ratios measured in sludge and effluent in 1991- 1994. It is possible
400
C. Zago et al. f Marine Environmemal Research 51 ( 2001) 389-4 15
8
.
6
G)
4
0
~
s
·::
:::E
:::::
!!
Ql
:::E
:
2
0
rp~t102
C BJ:10_3 _ ---,--- • J>Iudge 1979-84
CTotal1979-84
p.~iudge 1991
t:lTotal1991
Cl Effluent 1993-94
: j~
~
CrtCu
1.60
1.05
0.43
0.90
0.07
0.11
0.14
~
0.81
0 .89
0.20
0.51
0.05
0.08 0.12
~
~
CrfZn
CrfPb
Cu!Zn
I
5.05
3.30
0.80
2.83
- o .33
tt------o.35
-0.86
- --r-
-
-
--
'-----
--
-
'--
0.51
0.88
0.45
0.56
0.68
-t0.70
- 0:82
~ j'
CutPb
~
·_
:i:
<:
Zn/Pb
3.19
6.4
-~
3.18
1.86
4.10
-r- - 5.61
3.17
4.67
6 .87
3.09
4 .39
6.14
7.50
-
Fig. 5. Ratios among anthropogenic metals from both stations (data from this paper) and from sewage
sources (MWRA data). The data from sludge discharged from Nut Island, which is close to station BH03
(see Fig. I) and the total amount of metals discharged by both effluen ts and sludge (coming from Deer
Isla nd and Nut Island) are given from 1978- 1984, and 1991 , the last year of sludge discharge. The 19931994 data are only for total effluent since the disposal of sludge had ceased. Error bars shown for BH02
and BH03 are the 95% confidence interval calculated from 1991 to 1998 surficial sediment ratios.
the sediments reflect earlier loadings but the values are still 2- 5 times higher than the
ratios reported for sludge during 1979-1984. The sediment ratios with Cr are also
higher than the historical ratios from total loading from effluent and sludge, which
as previously discussed, probably overestimate Cr in the effluent. For this reason,
the enrichment of C r in sediments cannot be explained by changes in sewage inputs.
This suggests that either there is a Cr source in the Boston Harbor in addition to
sludge a nd/or effluents, or that differential scavenging and/or diagenesis accounts
for the Cr enrichment.
In order to enhance the anthropogenic signal and to correct the metal concentration data to variations due to sediment reworking, we exam ined the changes in the
anthropogenic metai/ Fe ratio over time at each station beginning in 1992 when
sludge disposal had ceased. The metal/Fe ratio has been used in many coastal
environments and , as suggested by other authors (Bothner et aL , 1998), Fe was
preferred to AI as a normalizing constituent because AI is rich in aluminum-rich
feldspars, which are common in glacially-derived sands in the Boston area.
Although Fe itself is a constituent of sewage, its concentration is very low compared
to the anthropogenic metals (Bothner et aL, 1998). Some caution needs to be used in
this approach as Fe may not be completely conserva tive at these two stations. Our
data suggests some migration of Fe in pore waters may occur (as shown in the following paragraph). Keeping in mind these considerations, metai/Fe ratios are a
useful instrument to observe changes in the metal content over time.
C. Zago eta/./ Marine Environmental Research 51 (2001) 389-415
0.6
~-··--·~·-··--------·-·~··~-·-·--~
0.5
• •
Jan-92
0.4
'
•'!
l
Jan-94
Jan-96
Jan-98
•
0.14
•
0.08
006
0.04
0.2
Jan-92
Jan-94
•
1.0E-03
Jan-92
•
Jan-98
I
Jan-96
Jan-98
Pb/Fe%.
••
•
•• •
Jan-92
Jan-94
•
Jan-96
Jan-98
Corg/Fe%
·-
• •••
2.0E-04
Jan-96
i
I
•'
I
•
•
;
•I
Jan-94
•
0.10
0.3
6.0E-04
0.2
0.12
>,
I
i
••
• • '•
•
Zn/Fe%
••
I
•
0.3
Jan-92
---
·
•
i
0.1
·-·-·-·--·------------··-·------~·-·
0.6
0.5
i
!
• •
0.3
0.7
Cu/Fe%
••
0.4
0.2
···-·-·-·----·--·--··--·-··l
0.4
•
0.4
0.3
0.5
Cr/Fe%
0.5
401
•
Jan-94
•
Jan-96
Jan-98
Fig. 6. Metals/Fe and organic C/Fe ratios over time from 1992 to 1998 at station BH02.
At Station BH02 (Fig. 6) there is only a weak decrease of the organic carbon/Fe
ratio over time suggesting that iron and organic carbon are changing at the sites in
similar ways. This would be expected if resuspension of sediments were the major
process determining sediment composition. In contrast, a strong linear decrease is
seen in the Zn/Fe ratio which decreases almost 50% from February 1993 to October
1994, followed by a period of little change. Similarly the ratios of CufFe, Pb/Fe
exhibit some decrease during the same period, but the decrease was not linear and
the pattern is not as strong. A Cr/Fe decrease is not evident. Using Fe normalized
data there is a suggestion that Zn inputs to Station BH02 have decreased and some
decrease may have also occurred in Cu and Pb. This result would not be seen using
only the metal concentration data over time (compare Fig. 6 to Fig. 2).
-------------------------------
-----
----------
C. Zago eta/./ Marine Environmental Research 51 (2001) 389--415
402
0.6
0.5
0.4
•
••• •
• •
• ••
• •
0.6
0.5
0.4
Jan-96
-,
Zn/Fe% j
,..
.. •
1.0E-03
8.0E-04
7.0E-04
6.0E-04
••
I
I
I
I
••
.... ,.
•
5.0E-04
I!
• ••
0.2
Jan-92
0.14
Jan-94
·I
Jan-96
0.12
----,
•
......
••• •
0.10
I
I
.I
•
0.08
Jan-94
Jan-98
Pb/Fe% 1
•
0.06
Jan-92
Jan-98
C/Fe%
...• ••
0.18
I
Jan-96
C~e% 1
•'••
0.3
0.16
el
1.1E-03
9.0E-04
0.4
Jan-98
•
•••• •
Jan-94
0.5
I
•II
•
0.3
Jan-92
I
I
I
j
•
Jan-94
0.6
I
0.3
Jan-92
1
Cr/Fe%
Jan-96
Jan-98
I
II
I
•
•
I
4.0E-04
Jan-92
Jan-94
Jan-96
Jan-98
Fig. 7. Metals/Fe and organic CfFe ratios over time from 1992 to 1998 at station BH03.
At Station BH03 (Fig. 7), the organic carbon(Fe decreased over time. This was
due to a loss of carbon at the site, as Fe concentrations remained relatively constant
throughout the study period (Fig. 2). A small decrease is observed in the Cr(Fe ratio
(about 20%), while larger decreases are observed in the CufFe ratio (about 50%)
and the Pb(Fe ratio (about 30%). In contrast to Station BH02, Zn(Fe ratios showed
great variability and no consistent decrease in concentration with time. The absence
of a decreasing Cr/Fe ratio at BH02 and the relatively small decrease at BH03 after
stopping sludge discharge again suggests that the high Cr values found in sediments
(Fig. 5) are not solely related to sewage inputs, or that Cr and Fe have similar postdepositional mobility.
To determine if the temporal patterns observed for all metals were linked with
changes in organic matter content in sediments, we performed a linear regression
analysis on all the time points from both stations (Fig. 8). Although all of the
regressions were highly significant, a defined separation between the two stations is
C. Zago et a!./ Marine Environmental Research 51 ( 2001) 389-415
403
Table 2
Pearson rank correlation matrix for organic carbon and metal concentrations in superficial sediments
(HCl digestion) collected in the two stations from 1991 to 1998
BH02
N sample= 12
Cu
Zn
Cr
Fe
Mn
Pb
c
Cu
Zn
Cr
Fe
Mn
0.46
0.57"
0.46
0.58"
0.58"
0.46
0.88b
0.95"
0.65b
0.70b
0.96b
0.8lb
0.76b
0.73b
0.83b
0.7lb
0.75b
0.96b
0.96b
0.57"
0.62b
c
Cu
Zn
Cr
Fe
Mn
0.58"
0.80b
0.27
-0.43
0.67b
0.59"
0.55"
-0.27
0.39
0.27
-0.07
0.29
0.15
0.23
-0.37
c
Cu
Zn
Cr
Fe
Mn
0.89b
0.83b
0.88b
0.78b
0.68b
0.87b
0.89b
0.94b
0.82b
0.59b
0.93b
0.9[b
0.89b
0.68b
0.90b
0.85b
0.73b
0.93b
0.79b
0.86b
0.71b
BH03
N samples= 16
Cu
Zn
Cr
Fe
Mn
Pb
0.79b
0.33
0.78b
-0.06
-0.!3
0.44
All sediment data
N samples= 28
Cu
Zn
Cr
Fe
Mn
Pb
95% confidence interval.
b 99% confidence interval.
a
evident, with low organic C concentrations in BH02. When individual stations are
regressed, there is a higher (or at least similar) correlation at Station BH02 than in
BH03 between C versus Fe, Mn, Zn, Pb and a lower correlation only between C
versus Cr and Cu. When we examine the correlation between individual metals in a
multiple correlation a similar pattern emerges (Table 2). In the overall data-set all
metals are very strongly correlated with %C, and nearly all metals are strongly
correlated with one another at 99% confidence intervals. When the data are separated
by station, the pattern among metals is much stronger at Station BH02, and correlation
of metals with organic C show higher values than in BH03, again with the exception
of Cr/C and Cu/C. The higher correlation observed among all the metal species in
Station BH02 than in BH03 and the higher metal and carbon variability over time
suggest that at the BH02 site reworking of sediments governs the metal and carbon
content of the area. A high correlation also suggests that there have not been large
changes in the ratio of metal inputs. The one exception maybe Zn, which appears to
be decreasing faster (from 1993 to 1995) than can be accounted for by changes in
C. Zago et al./ Marine Environmental Research 51 (2001) 389-415
404
3000
...~
2000
f
3000 . - - - - - Cu(aii)IC r'= 0.89
Cu(03)/C r'= 0.79
Cu(02)/C r'= 0.46
Cr(aii)/C r'= 0.88
1000
Cu
~
2000
f~
1000
Cr(03)/C r'= 0.78
Cr(02)/C r'= 0.46
2 C% 3
600000
8000
~
~
...
...
f::::::
"!
~
4000
~
2
C%
800
...~
~
1
3000
600
~
·~
Zn(all)? = 0.83
Zn(03) r' = 0.33
Zn(02)
r' =~57
2000
~
400
..
~
~
~
200
2 C% 3
4
1000
4
C%
..
Zn
2
I
I
C%
Fig. 8. Linear regression analyses among metals and carbon in sediments of both stations. + = BH03
values, D.= BH02 values. Correlation coefficient for the overall data set and for each station is shown.
carbon content. At Station BH03, strong correlation values (at 99% confidence
interval) are evident for ratios involving Cu or Cr (CujC, Cr/C, Cr/Cu and Pb/Cu).
3.2. Vertical pore water and sediment profiles
To facilitate the discussion of metal cycling at each site, we present first a brief
discussion on the redox conditions observed in the two sites in July of 1994. Dissolved sulfides analyzed in pore waters increased with depth at both stations (Fig. 9)
with higher concentration at site BH02 than at site BH03. In July 1994, Station
BH02 had a more negative Eh in surface samples than did Station BH03. More
reduced conditions at Station BH02 than at BH03 were generally measured during
1991-1994 (Giblin et al., 1997). During a 1991-1994 differences in the dissolved
sulfide concentrations between the two stations were frequently greater than we
observed in July 1994 (Giblin et al., 1997). The more oxidized sediment surface layer
C. Zago eta/./ Marine Environmental Research 51 (2001) 389-415
0
0
0
0
<D
en
"' ...
:.::
E
::>
'
··•····•·
Gl
0
,A
::t
0
0
0
0
405
0
0
0
0
N
0
0
N
CD
~
~
depth (em)
a:>
-~
:z::
Q.
I
I
.,
...:
0
N
...
a:>
0
depth (.;-m)
8N
>
Eo
0
0
":'
depth (em)
---l·
-----·
N
en
N
:z::
I
~E
E
. ....... .
0
0
N
...
<D
a:>
$!
depth (em)
~
;:!:
....
~
----BH02
~
• • •· • BH03
Fig. 9. Dissolve sulfides, redox potential (Eh), pH in pore water and acid volatile sulfide (AVS) in sediment in sites BH02 and BH03 taken in July 1994.
406
C. Zago et al./ Marine Environmental Research 51 (2001) 389-415
LL
Ql
l~c% -lr-Water%
0
0
0
0
0
<0
..
0
0
:::J "'
I
0
~
"'
0{2'
Oo
Eo
2."'
8~
Jg
(!!-
c
2c::
';JI.N
~5
oc
OQ)
8E
c::
5l 8
0
-e
t'l
~'U
I
~
;JI.
'"
0
a.
<n
0
0
~
0
~
0
"'
0
"'
0
"'
0
0
Q;
:::J
Cl
15
..
.
$
3:<0
1!!0
0
E
0
0
0
2.
0
0
'6
"' c:E
~
a.
"'
~
~
"
3:
0
0
E
2.
0
....
*"'
3:
·:···=*·:~
80
~0
"'
:;;
;;
3:~
0
0
i!!ci
a.
~
0
~
depth (em)
~
~
2.
f
0
E
2.
~
E
'6
~
0
0
0
co
::;;
Cl
0
0
0
~~
...
2.
0
a.
~
15
E
2.
~
3:o
"E
~
0
Q;
!!!-
E
"'
<:
0
0
0
N
..
·:·. ·..
',',)(
'
-~~-lil;.
..
~
~
depth (em)
0
"'
•••
0
0
0
:J'
ill
0
~
0
"'
"'
~
0~
"'
0
0
0
0
Q;
15
0
0
a.
"'
Cl
"'
"'
"'0 c:
(!!
0
"'ci
N
N
Ql
0
0
0
0
- - - - - - - - - 00...
<:
<D
g ~
~~
2.
~
0
0
....0
:J'
~
c:
"E
'6
5:
;)(
0
"'
• ·Porewater metals (ICP)
metals (PIXE)
• • !)(· • •Porewater
• - a- - ·Porewater metals (AAS)
--&-Sediment metals
Fig. 10. Heavy metals in pore water, sediment metals, water content, and% carbon in sediments at station BH02. Dissolved metals analyzed by particle induced x-ray emission (PIXE), atomic absorption
spectrometer (AAS), and/or inductively coupled plasma atomic emission spectrometry (ICP-AES). Sediments metals were obtained by a HC1/HN0 3 digestion except for Cr which was an HCl digestion. See text
for details.
at Station BH03 was apparently associated with very high densities of tube-building
amphipods that irrigated the sediment. From 1991 to 1998, the densities of building amphipods at Station BH03 were usually greater than at BH02, although
densities were highly variable over time. The A VS concentrations (Fig. 9) also
increase with depth in the sediments at both stations indicating higher concentrations of metal sulfides at depth. In contrast to dissolved sulfides, A VS was higher in
sediments from BH03 than from BH02.
407
C. Zago et al. /Marine Environmental Research 51 (2001) 389--415
1-+-c% -.!r-Water%
0
_,
~~
0
0
-~
Q;
0
E
0
2-
§~
~o
iU1
~'E
~
0.
··
(I)
0
;:::·...
...
0
"'
~
~
~
0
0
0
"'
~
g~
~
2-
2-
c
o<D
8.5
I
~
0
0.
~
§~
Oo
EN
::.::
""E
2-
c
oa>
g
&]
C;
::.::
Q;
0
.,
gE
o=>
... ::;-
~0
.,_
cQ)
50.
'5
Q)
~
"'
0
depth (em)
E
0
0
"'
0
...
...
~
depth (em)
• - •
• •
~-
0
"'
2-
Q)
~
"'"
Q;
~~
~
~
"E
'5
g
"' C;
I
"'C:
•·x
~"'
~
~
~
0
oE
o=>
o-
0
(.)
0
0
0
;!
~
0
0
0
0
"'
,.._
"'
r--;·----
~
"' ·-----c
N
~"rt
"'""'"
...
~
~"'
"'
hee
I
"#~$
-e
"'
oo
""E
~
...
0
o
~
0
S!
0
E
,------·---I
<!'
c
'5
(I)
• .X
I
"'
"'
,)(
~
0
"'
• · Porewater metals (Core A)
- · Porewater metals (Core B)
~Sediment
metals
Fig. II. Heavy metals in pore water, total sediment metals, water content, and% carbon in sediments at
station BH03. Dissolved metals were analyzed on replicate cores (A and B). Sediment metals were
obtained by a HCl/HN0 3 digestion except for Cr which was an HCI digestion. See text for details.
Pore water and sediment metals concentrations from sediments (HCI/HN0 3) collected in July 1994 in the two stations are shown in Figs. 10 and 11. Results obtained
by the three different techniques in site BH02 (Fig. 10) show that PIXE, AAS and
ICP-AES analysis give comparable results for metals with concentrations higher
than I IJ.mole/L (Fe, Mn, Zn). Some contamination probably occurred for Mn
in surface sediment samples analyzed by PIXE. For concentrations lower than
1 IJ.mole/L (as in the case of Cr and Cu), PIXE values were lower than those of other
techniques probably because they were closer to the PIXE detection limit. For this
reason PIXE values of Cr and Cu were not considered. Metals at Station BH03 were
only analyzed by PIXE so pore water concentrations of Cr and Cu are not shown.
408
C. Zago eta!./ Marine Environmental Research 51 (2001) 389-415
The comparison between the pore water analysis of the replicate cores collected in
site BH03 shows that for Fe and Mn spatial variability was fairly low (Fig. 11 ).
Different trends among metals were observed in pore water at the two sites. At
Station BH02 there is strong evidence for an upward migration of Fe and Mn (Fig.
10), where the concentrations increase to the sediment-water interface. Given the
resolution of our sampling interval (1 em) the data suggests that Fe and Mn are
fluxing from the sediments and are not being precipitated as oxide or hydroxides
here. In the deepest core sections below the oxidized zone, where H 2S concentrations
become higher (below 5 em depth) and redox potential more negative, Fe and Mn
are removed from pore water presumably forming metal sulfides in sediments. At
Station BH03 (Fig. 11) the Fe and Mn concentrations reach maximal values at 1-2
em and then Fe decreases toward the surface of the sediment. This suggests significant precipitation at the surface as Fe (and probably Mn) oxides and hydroxides.
The profiles of Fe and Mn in the solid phases supports this interpretation; Fe and
Mn are enriched in the surface sediment layers at BH03 but not appreciably at
BH02. Pore water concentrations of Zn, Cr and to a lesser extent, Cu, show similar
trends to Fe and Mn, as has also been observed in pore water of other coastal
environments (Zago et al., 1994).
The metal content with depth measured in sediments shows a greater variability
for BH02 than for Station BH03, similar to that obtained from the temporal data.
Concentration peaks are present at 4.5 and 13 em depth in BH02, showing strong
similarities with the percentage water and organic carbon content along vertical
profiles suggesting events where higher contents of fine particles were deposited.
Matrix correlation performed on the BH02 vertical profiles (Table 3) shows
Table 3
Correlation matrix for organic carbon, and metal concentrations in vertical sediment cores collected in the
two stations. Cr data are obtained by HCI sediment digestion
BH02
N sample=13
c
Cu
Zn
Cr(HCI)
Fe
Mn
0.7Jb
0.55
0.7Jb
0.70b
0.35
BH03
N samples=9
c
Cu
Zn
Cr(HCI)
Fe
Mn
0.75b
-0.35
0.89b
0.77b
0.84b
a 95% confidence interval.
b 99% confidence interval.
Cu
Zn
0.76b
0.87b
0.21
-0.10
0.77b
0.40
-0.02
Cu
Zn
-0.02
0.7oa
0.64
0.71a
-0.07
-0.29
-0.34
Cr(HCI)
0.37
-0.12
Fe
0.74b
Cr(HCi)
Fe
0.45
0.55
0.97b
C. Zago et al.f Marine Environmental Research 51 (2001) 389-415
409
significant correlation coefficients for Cu/C, Fe/C , Cr/C, Zn/Cu, Cr/Cu, Zn/Cr and
Fe/ Mn. Generally, the correlation coefficients obtained by vertical profile data in
this site are lower than those from the superficial temporal data (with the exception
of Cu/C , Fe/C, Cr/C and Zn/ Cu) suggesting some diagenetic and remobilization
processes may be occurring, or reflecting changes in metal loading over time. The
low correlation of Zn in the depth profiles, along with the linear decrease of Zn/ Fe
ratio with time from surface sediments (Fig. 6), and the generally high Zn content in
BH02 sediments (Fig. 4) all suggest that there may be substantial remobilization and
recycling of Zn from the sediments at Station BH02. At Station BH03, Fe, Mn, Cu
and organic C are strongly correlated with one another in the vertica l profiles while
Zn is negatively correlated with both carbon or the other metals. A high correlation
between vertical profiles of metals and ca rbon suggest organic- metal bounding.
The extent to which metals in sediments were partitioned between dissolved and
solid phases was quantified with calculations of their distribution coefficient, K 0 .
This distribution coefficient, which is not an equilibrium constant, is derived from
the equation
Ko
= Cs/Cpw
where C5 is the concentration of metal in the sediments (!!mole/ Kg) and Cpw is the
concentration of dissolved metal in pore water (11mole/ L).
The log K0 was calculated for metals in the two sites and with the two different
sediment digestions (HCI/ HN03 and HCI) in order to study which mineral form
accounts for the metal trapping in sediments (Fig. 12). The cold HCI digestion a nalyzes the metal fraction including metal oxides, monosulfides a nd metal loosely
bound to clay. The HCI/ H N0 3 digestion analyzes the metal fraction including pyrite, stable sulfides and metals bound to clays. For Cpw va lues at BH02 we used the
mean concentration obtained by the different techniques and at BH03 we used mean
values from the two cores. We only have HCI/HN03 Cr concentration from superficial sediment layers.
The log K 0 of the HCI fraction increased with depth for all examined elements,
showing a tendency of dissolved metals to be transferred into solid phases in deeper
sediment layers. The increasing coefficient with depth should be primarily due to
monosulfides. Site BH02 had lower log K 0 values than did site BH03, particularly in
surface sediments, suggesting that metal partitioning between dissolved and solid
phase favors remobilization of metals at site BH02. The log K 0 values for Fe and
Mn measured at surface sediments at Station BH03 were higher than Station BH02
due to the more oxidized nature of the sediments that allows Fe a nd Mn hydroxides
to form in the top 2 em. The higher log K 0 values for the other elements at this
station arc potentially due to the presence of the Fe/Mn oxide layers, the higher
organic carbon content of the sediments that bounds metals (as also shown by
correla tion analysis in Table 3), or the formation of other mineral forms such as Zn
and Cu hydroxides.
For most metals, adding in the fraction of metals which can only be released with
a strong H C1/ HN03 digestion does not increase the log K 0 greatly (Fig. 12). Pyrite
"'"
0
2
3
4
5
e
05
2
4
5
2
6
HCI
HN01
3
5
4
~ "' II ': /~ ~
15
HCI
2
3
4
5
6
2
05
05
15
15
3
4
5
6
0
~
Q
3
3
5
5
7
7
--~...
9
~
~
HC
.!5
5
HCI
+ HN01
9
6
N
+ HNO,
3
logKd Fe
logKd Mn
05
BH02
-n
3
05
1.5
f 3
!f
logKd Cu
logKdZn
logkd Cr
----l
91
9
n
9
~
~
::·
;:;
~·
0
§
logKd zn
2
3
4
3
4
5
6
II
15 I
:1 ~ .~~
HCI
I
I
:~'__j
1 5 I II
3
5
2
3
4
5
6
~,
I
I
1.5
I
I
3
HCI
HCI
+ HN01
5
7
7
• L...___U _ _ _ J
L
!:..
.."'
...
~
05
0.5
05
BH03
2
6
~
log Kd Fe
logKd Mn
5
\.
I
Fig. 12. Log Kd values obtained by the ratio between pore waters a nd sediments concentrations analyzed with both HCI a nd HCI/H NO\ d igestions.
"'
<"'
::::v,
._
--.
8
._
.._
~
!._
v,
C. Zag a et al. /Marine Environmental Research 51 ( 2001) 389-415
411
is the major iron form extracted from sediment by the HCljHN0 3 digestion. Many
metals such as Cu, Cr and Zn, due to their low concentration in comparison with
that of iron, are frequently coprecipitated with pyrite in anoxic waters rather
than as discrete sulfide crystals (Forstner & Wittman, 1979). However, when we
sampled the sediment, pyrite accounted for only 2-3% of the Fe content, so pyrite
precipitation does not appear to be controlling the metal composition of the
sediment. At this time, monosulfides and oxides would appear to be the likely
controlling phases.
To further examine the potential controls of metal distribution in the sediments, we
used the equilibrium model MINEQL +. At Station BH02, this model predicts ironsulfide precipitation all along the depth of the core. Pyrite is the most stable form of
iron sulfide and is always predicted from equilibrium models although it may be
slow to form. At Station BH03, the model predicts pyrite precipitation beginning
at 2-4 em, and for the remainder of the core where sulfide concentrations are
higher. Zn is the only other metal that is predicted to precipitate as a discrete
monsulfide. Zn monosulfide, sphalerite, is predicted to form below 2-4 em in both
stations. Metal-hydroxide saturation indexes show a tendency to precipitate in
surface sediment.
In order to study the potential relative importance of monosulfides relative to
oxides and hydroxides in the sediments, we calculated the following ratio:
K1 = CsHci/CAvs
=([Me-hydroxides]+ [Me-monosulfides] + [Me-clay])j[monosulfides]
By this ratio, we normalize the metals in sediment versus the monosulfides
content of sediment in order to observe the contribution of hydroxides and
monosulfides precipitation. Negative log K 0 • values mean a higher monosulfides
concentration among metals extracted with HCl digestion, and so monosulfides
Table 4
LogK0 • values obtained by the ratio between metals and AVS. Values greater than one indicate that there
is not enough AVS present to bind all of the metal as a metal sulfide and suggests that metal oxides must
be present.
Depth (em)
Fe
Mn
Cu
Cr
Zn
BH02
0-2
2-4
4-6
6-8
8-10
2.37
1.47
1.11
1.01
0.95
0.30
-0.60
-0.95
-1.09
-1.13
-0.25
-1.08
-1.39
-1.55
-1.50
0.11
-0.75
-1.07
-1.19
-1.20
-0.43
-0.93
-1.01
-1.05
-1.17
BH03
0-2
2-4
4-6
6-8
8-10
1.72
1.10
0.95
0.80
0.63
-0.33
-0.93
-1.08
-1.26
-1.44
-0.55
-1.03
-1.13
-1.17
-1.35
-0.30
-0.82
-0.90
-1.02
-1.19
-0.002
-0.82
-1.15
-1.26
-1.27
412
C. Zago eta!./ Marine Environmental Research 51 (2001) 389-415
precipitation is possible. Positive log Kn• values means higher HCI extractable
metals concentrations exceed monosulfides so hydroxides must be present. Surficial
samples from both stations (Table 4) had log Kn• values greater than I for Fe
indicating that oxides must be present. However, A VS was present in the surface
sediments in concentrations more than sufficient to bind the Zn in the surface
sediments at BH03. Log Kn* values strongly decrease at 2--4 em depth where Eh
values become negative. Below this depth it is likely that Zn is largely present as
sulfides.
4. Conclusions
The concentrations of some metals in sediments from these two Boston Harbor
stations changed after sludge discharge was ended, although with different intensities. We expected that Station BH03, the old sludge disposal area, would show a
large decrease in metal content after discharge ceased, but this was only strongly
evident for Cu and Pb, metals known for the high affinity with organic material
(Burton & Liss, 1976; Forstner & Wittman, 1979). Cr and Zn did show some evidence of a decrease but not as strongly as Cu and Pb. The metal concentrations
measured in BH03 sediments from 1991 to 1998 were much higher than those of
Station BH02. Organic carbon content also was much higher than that of Station
BH02 but has been decreasing since sludge disposal ceased. At BH02, metal and
organic carbon content fluctuated greatly over time and metal and carbon concentrations were highly correlated over time with one another. Ratios of Fe, to carbon, Mn, and Cr, concentrations remained fairly similar over time at this station,
while Zn, and to a lesser extent, Pb and Cu have decreased in the sediments relative
to iron.
These data suggest that the absolute decreases in Cu, Pb, and Zn concentrations at
Station BH03, and that the relative (to Fe) decreases at Station BH02, could be
related to lower metal inputs to these station since sludge discharge ceased. At both
stations, the amount of Cr found in the sediments is considerably higher than would
be expected from the ratio of Cr to the other anthropogenic metals if sewage and
sludge were the major source of Cr. Either the sediments trap and retain Cr much
more efficiently than the other metals or there is another significant source of Cr to
the harbor.
Acknowledgements
We would like to thank Jane Tucker, David Vasiliou, and David Geihtbrock for
field assistance and for carrying out many of the chemical analysis. Thanks also to
Prof. Grazia Ghermandi of the Geophysical Observatory of Modena University for
performing PIXE analysis and Dione Lewis at Battelle for performing the AAS and
ICP analyses. Battelle furnished and operated the differential global positioning
system used to locate stations. We thank Murray Hall and Kenneth Keay for
C. Zago eta/./ Marine Environmental Research 51 (2001) 389-415
413
their assistance in obtaining the metal loading data. Frank Manheim provided
helpful comments on an earlier draft of the manuscript. Financial support was
provided to Cristina Zago by the MAS3 CT95 0021 of the European Union, and
by the Department of Environmental Science of the Venice University. Financial
support was provided to A.E. Giblin by NOAA National Sea Grant College
Program Office, Department of Commerce, under Grant No. NA46RG04700,
Woods Hole Oceanographic Institution Sea Grant Project No. R/P-56 and the
Massachusetts Water Resource Authority. The views expressed here are those of the
authors and do not necessarily reflect the views of NOAA or MWRA or any of their
subagencies.
References
Alber, M., & Chan, A. B. (1994). Sources of contaminants to Boston Harbor: revised loading estimates.
Massachusetts Water Resources Authority, Environmental Quality Department Technical Report
Series 94-1.
Angelidis, M. 0., & Aloupi, M. (1995). Metals in sediments of Rhodes Harbor, Greece. Marine Pollution
Bulletin, 31, 273-276.
Barreiro, R., Real, C., & Carballeira, A. (1994). Heavy metal horizontal distribution in surface sediments
from a small estuary (Pontedeume, Spain). Science of the Total Environment, 154, 87-100.
Bohn, H. L. (1971). Redox Potentials. Soil Science, 112, 39-45.
Bothner, M. H., Buchholtz ten Brink, M., & Manheim, F. T. (1998). Metal concentrations in surface
sediments of Boston Harbor- changes with time. Marine Environmental Research, 45, 127-155.
Burton, J.D., & Liss, P. D. (1976). Estuarine chemistry. New York: Academic Press.
Cecchi, R., Ghermandi, G., & Zonta, R. (1990). PIXE analysis to study metal diffusion and sedimentation
phenomena in Venice Lagoon. Nuclear Instrument and Methods, B49, 283-287.
Cline, J. D. (1969). Spectrophotometric determination of hydrogen sulfide in natural waters. Limnology
and Oceanography, 14, 454-458.
Di Toro, D. M., Mahony, J.D., Hansen, D. J., Scott, K. J., Carison, A. R., & Ankley, G. T. (1992). Acid
volatile sulfide predicts the acute toxicity of cadmium and nickel in sediments. Environmental Science
and Technology, 26, 96-101.
Feng, H., Cochran, J. K., Lwiza, H., Brownawell, B. J., & Hirschber, D. J. (1998). Distribution of heavy
metals and PCB contaminants in the sediments of an urban estuary: the Hudson River. Marine Environmental Research, 45, 127-155.
Fitzgerald, M. G. (1980). Anthropogenic influence of the sedimentary regime of an urban estuary Boston Harbor. PhD. thesis. Woods Hole Oceanographic Institution - Massachusetts Institute of
Technology Joint Program in Oceanography.
Forstner, U., & Salomons, W. (1980). Trace metal analysis on polluted sediments. Environmental Technology Letters, /, 494-505.
Forstner, U., & Wittmann, G. T. W. (1979). Metal pollution in the aquatic environment. Heidelberg:
Springer Verlag.
Giblin, A. E., Hopkins, C. S., & Tucker, J. (1993) Metabolism, nutrient cycling and denitrification in
Boston Harbor and Massachusetts Bay sediments. Massachusetts Water Resources Authority, Environmental Quality Department Technical Report Series 93-2.
Giblin, A. E., Hopkins, C. S., & Tucker, J. (1997). Metabolism and nutrient cycling in Boston Harbor
Massachusetts. Estuaries, 2, 346-364.
414
C. Zago eta/./ Marine Environmental Research 51 (2001) 389--415
Haynes, D., Raymen, P., & Toohey, D. (1996). Long term variability in pollutant concentrations in
coastal sediments from the Ninety Mile Beach, Bass Strait, Australia. Marine Pollution Bulletin, 32,
823-827.
Hunt, C. D., West, D., & Peven, C. S. (1995). Deer Island effluent characterization and pilot treatment plant
studies. Massachusetts Water Resources Authority, Environmental Quality Department Technical
Report Series, 9 5-97.
Knebel, H. J., & Circe, R. C. (1995). Seafloor environments within the Boston Harbor-Massachusetts Bay
sedimentary system: a regional synthesis. Journal of Coastal Research, 11(!), 230-251.
Kristensen, H. J., & Andersen, F. 0. (1987). Determination of organic carbon in marine sediments: a
comparison of two CHN-analyzer methods. Journal Experimental Marine Biology and Ecology, 109,
15-23.
Kropp, R. K., & Diaz, R. J. (1995). lnfauna/ community changes in Boston Harobr 1991-1994. Massachusetts Water Resourches Authority Environmental Quality Department, Technical Report 95-21.
Massachusetts Water Resources Authority, Boston, Massachusetts.
Lyons, W. B., Gaudette, H. E., & Smith, G. M. (1979). Pore water sampling in anoxic carbonate sediments: oxidation artefact. Nature, 277, 48-49.
Nordstrom, D. K., Plummer, L. N., Wigley, T. M. L., Wolery, T. J., Ball, J. W., Jenne, E. A., Bassett,
R. L., Crerar, D. A., Florence, T. M., Fritz, B., Hoffman, M., Holdren, G. R. Jr, Lafon, G. M., Mattigod, S. V., McDuff, R. E., Morel, F., Reddy, M. M., Sposito, G., & Thrailkill, J. (1979). A comparison of computerized chemical models for equilibrium calculations in aqueous systems. American
Chemical Society Symposium Series, 93, 857-892.
Olausson, E., & Cato, J. (1980). Chemistry and Biogeochemistry of Estuaries. Chichester: John Wiley and
Sons.
Pardue, J. H., DeLaune, R. D., & Patrick, W. H. Jr (1992). Metal to aluminium correlation in Louisiana
coastal wetlands: identification of elevated metal concentrations. Journal of Environmental Quality, 21,
539-545.
Scarponi, G., Turetta, C., Capodaglio, G., Toscano, G., Barbante, C., & Cescon, P. (1998). Chemometric
studies in the Lagoon of Venice (Italy). Part I. Environmental study of water and sediment matrices.
Journal of Chemical Information and Computer Sciences, 38(4), 552-562.
Schecher, W. D., & McAvoy, D. C. (1994). MINEQL+ a chemical equilibrium program for personal
computers. Version 3. Hallowell Maine: Procter & Gamble Company, Environmental Research
Software.
Shea, D., & Kelly, J.R. (1992). Transport and fate of toxic contaminants discharged in MWRA input to
the Massachusetts Bay. Massachusetts Water Resources Authority, Boston, MA, Technical Report
92-94.
Tam, N. F. Y., & Wong, Y. S. (1995). Spatial and temporal variations of heavy metal contamination in
sediments of a mangrove swamp in Hong Kong. Marine Pollution Bulletin, 31, 464-470.
Tucker, 1., Sheats, N, Giblin, A. E., Hopkinson, C. S., & Montoya, J.P. (1999). Using stable isotopes to
trace sewage-derived material through Boston Harbor and Massachusetts Bay. Marine Environmental
Research, 48, 353-375.
Wallace, G.T., Krahforst, C.F., Pitts, L.C., Shine, J.P., Studer, M.M., & Bollinger, C.R. (1991).
Assessment of the chemical composition of the Fox Point CSO Effluent and associated subtidal and
intertidal environments: analysis of CSO effluents, receiving water and surface sediments for trace
metals prior to CSO modification. Final report to the Department of Environmental Protection,
Commonwealth of Massachusetts. Prepared by the University of Massachusetts Environmental
Sciences Program.
Williamson, R. B., Van Dam, L. F., Bell, R. G., Malcolm, 0. G., & Kim, J. P. (1996). Heavy metal and
suspended sediment fluxes from a contaminated intertidal inlet (Manukau Harbor, New Zealand).
Marine Pollution Bulletin, 32, 812-822.
Zago, C., Costa, F., Cecchi, R., Ghermandi, G., Zaggia, L., & Zonta, R. (41994). Heavy metals
C. Zago et a!./ Marine Environmental Research 51 ( 2001) 389-415
415
distribution in porewaters of the Cona Marsh: a preliminary investigation on surface sediments. Nuovo
Cimento, 17, 503-510.
Zdanowicz, V. S., Cunneff, S. L., & Finneran, T. W. (1995). Reductions in sediment metal contamination in the New York Bight apex with the cessation of sewage sludge dumping. In A. L. Studholme,
J. E. O'Reilly, & M.C Ingham, Effects of the cessation of sewage sludge dumping at the 12 mile site
(pp. 89-100). Seattle, Washington: NOAA Technical Report NMFS 124. U.S. Department of
Commerce.