Heavy metal contamination of brown seaweed and sediments from the UK coastline between the Wear river and the Tees river Lorenzo Giusti* Department of Environmental Sciences, Faculty of Applied Sciences, University of the West of England, Frenchay Campus, Coldharbour Lane, Bristol BS16 1QY Received 1 December 1999; accepted 23 November 2000 Abstract The concentration of Fe, Mn, Zn, Cu, Pb, Ni, Cr, Cd, and Ag were determined in the brown alga Fucus vesiculosus and intertidal surface sediments from coastal locations of northeast England. Levels of heavy metals similar to those of polluted areas of the British coastline were detected. There is evidence of contamination (especially with Zn and Pb) in sediments from sites affected by colliery spoil and from the Wear estuary. The pelitic fraction ( < 63 mm) is usually more enriched in heavy metals, but it represents a very small percentage of the bulk samples. The fine-grained sand is a very important repository of contaminants especially where particles of colliery spoil, secondary mineral, and amorphous phases are present. Aqua regia-extracted Zn, Cu, and Pb in sediments are significantly correlated with those in seaweed. Despite the closure of all base metal and coal mines, and the cessation of many industrial activities in the region, sediments and brown algae are contaminated with heavy metals. The control site (Holy Island) and the Tees estuary appear to be the least affected. D 2001 Elsevier Science Ltd. All rights reserved. Keywords: Heavy metals; Seaweed; Sediment; Estuarine and coastal pollution 1. Introduction For many centuries, the economic development of the northeast of England has been strongly linked to base metal mining in the Pennines and to coal mining in the Durham and Northumberland Coalfields. Urbanised indus- trial centres grew mainly around the estuaries of the Tyne River, the Wear River, and the Tees River (Fig. 1). The production of iron and steel, and shipbuilding, were located on the Tyne and Wear estuaries, whereas the petrochemical industry developed on the Tees mouth. All coal mining, heavy metal mining, and shipbuilding activ- ities have now ceased, and steel production has been scaled down. As a result, there has been a significant reduction in industrial discharges into the local estuaries and coastal waters. Domestic sewage, previously dis- charged untreated into the rivers, has been largely diverted to treatment plants located on or near the coast. With the exception of Holy Island, which consists of sandstone intruded by a quartz–dolerite dyke, the coastline sampled in this study is characterised by magnesian lime- stone cliffs of Permian age overlain by Pleistocene glacial deposits. The beaches are mostly composed of relatively coarse material (boulders, cobbles, gravel, pebbles, and sand). Natural clastic material eroded from the cliffs is mixed with alluvium carried by the local rivers and drifted south- ward by tidal currents. At sites such as Horden, Easington, and Blackhall Rocks, millions of tonnes of coal waste were dumped along the coast (Norton, 1985). In some cases, the colliery spoil has been reworked into terraces where weath- ering processes have produced clay minerals and iron oxide coatings on sand particles (Humphries, 1996). Large volumes (presently about 300,000 m 3 day 1 ) of minewater are still being pumped into the River Wear and its tributaries in order to prevent groundwater rebound (Younger, 1995). This is equivalent to about 15% of the average Wear River flow. The alluvium of the main river systems of this region has been contaminated by historic mining, especially of minerals such as galena (PbS), spha- lerite (ZnS), cerussite (PbCO 3 ), smithsonite (ZnCO 3 ), pyrite (FeS 2 ), fluorite (CaF 2 ), and baryte (BaSO 4 ; Dunham, 1934). 0160-4120/01/$ – see front matter D 2001 Elsevier Science Ltd. All rights reserved. PII:S0160-4120(00)00117-3 Tel.: +44-117-344-2487; fax: +44-117-344-2904. E-mail address: [email protected] (L. Giusti). www.elsevier.com/locate/envint Environment International 26 (2001) 275 – 286
12
Embed
Heavy metal contamination of brown seaweed and …dzumenvis.nic.in/Microbes and Metals Interaction/pdf/Heavy metal... · Heavy metal contamination of brown seaweed and sediments from
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Heavy metal contamination of brown seaweed and sediments from the UK
coastline between the Wear river and the Tees river
Lorenzo Giusti*
Department of Environmental Sciences, Faculty of Applied Sciences, University of the West of England, Frenchay Campus, Coldharbour Lane,
Bristol BS16 1QY
Received 1 December 1999; accepted 23 November 2000
Abstract
The concentration of Fe, Mn, Zn, Cu, Pb, Ni, Cr, Cd, and Ag were determined in the brown alga Fucus vesiculosus and intertidal surface
sediments from coastal locations of northeast England. Levels of heavy metals similar to those of polluted areas of the British coastline were
detected. There is evidence of contamination (especially with Zn and Pb) in sediments from sites affected by colliery spoil and from the Wear
estuary. The pelitic fraction ( < 63 mm) is usually more enriched in heavy metals, but it represents a very small percentage of the bulk samples.
The fine-grained sand is a very important repository of contaminants especially where particles of colliery spoil, secondary mineral, and
amorphous phases are present. Aqua regia-extracted Zn, Cu, and Pb in sediments are significantly correlated with those in seaweed. Despite
the closure of all base metal and coal mines, and the cessation of many industrial activities in the region, sediments and brown algae are
contaminated with heavy metals. The control site (Holy Island) and the Tees estuary appear to be the least affected. D 2001 Elsevier Science
Ltd. All rights reserved.
Keywords: Heavy metals; Seaweed; Sediment; Estuarine and coastal pollution
1. Introduction
For many centuries, the economic development of the
northeast of England has been strongly linked to base
metal mining in the Pennines and to coal mining in the
Durham and Northumberland Coalfields. Urbanised indus-
trial centres grew mainly around the estuaries of the Tyne
River, the Wear River, and the Tees River (Fig. 1). The
production of iron and steel, and shipbuilding, were
located on the Tyne and Wear estuaries, whereas the
petrochemical industry developed on the Tees mouth. All
coal mining, heavy metal mining, and shipbuilding activ-
ities have now ceased, and steel production has been
scaled down. As a result, there has been a significant
reduction in industrial discharges into the local estuaries
and coastal waters. Domestic sewage, previously dis-
charged untreated into the rivers, has been largely diverted
to treatment plants located on or near the coast.
With the exception of Holy Island, which consists of
sandstone intruded by a quartz±dolerite dyke, the coastline
sampled in this study is characterised by magnesian lime-
stone cliffs of Permian age overlain by Pleistocene glacial
deposits. The beaches are mostly composed of relatively
coarse material (boulders, cobbles, gravel, pebbles, and
sand). Natural clastic material eroded from the cliffs is mixed
with alluvium carried by the local rivers and drifted south-
ward by tidal currents. At sites such as Horden, Easington,
and Blackhall Rocks, millions of tonnes of coal waste were
dumped along the coast (Norton, 1985). In some cases, the
colliery spoil has been reworked into terraces where weath-
ering processes have produced clay minerals and iron oxide
coatings on sand particles (Humphries, 1996).
Large volumes (presently about 300,000 m3 day ÿ 1) of
minewater are still being pumped into the River Wear and its
tributaries in order to prevent groundwater rebound
(Younger, 1995). This is equivalent to about 15% of the
average Wear River flow. The alluvium of the main river
systems of this region has been contaminated by historic
mining, especially of minerals such as galena (PbS), spha-
nd = not detected.a Bryan and Langston, 1992 (HNO3 digestion).b Wedepohl, 1995 (HNO3/H2O2 digestion).c Long et al., 1995.
Fig. 2. Relationship between the concentration of (a) Pb and Cu, and (b) Zn and Cu in F. vesiculosus. For n = 17 at P < .01, the critical r is .606. The error bars
combine analytical error and natural variations of sample concentration at each site.
L. Giusti / Environment International 26 (2001) 275±286280
abundant are usually Cd and Ag (Table 4). However, Cd and
Ag appear quite concentrated in samples from Easington
and Horden (about 9±10 and 4 mg kgÿ 1, respectively), and
Ni (about 20±70 mg kgÿ 1) in samples from Roker and
Blackhall Rocks. The concentration sequence of the most
abundant elements is usually Fe > Mn > Zn > Cu. However,
at Holy Island, Mn > Fe, and at Whitburn, Roker, Easington,
Blackhall Rock, and one site at Bran Sands (No. 17),
Zn > Mn. The highest Mn enrichment in seaweed is at Holy
Island (276±778 mg kgÿ 1), which is also characterised by
the lowest concentrations of Zn (13±18 mg kgÿ 1), Pb
(0.1±1.1 mg kgÿ 1), and Ni (0.3±0.7 mg kgÿ 1). At Bran
Sands, Fe and Mn enrichment in seaweed decreases sig-
nificantly from Sites 13 to 17, going downstream along the
southern part of the Tees estuary. There is also an apparent
parallel decrease in all other trace metals, though this is not
statistically significant.
A significant correlation ( P < .001) exists between Fe
and Ag (r =.826), Zn and Cu (r =.957), Zn and Pb (r =.887),
Zn and Cr (r =.830), Zn and Ag (r =.822), Cu and Pb
(r =.910), Cu and Cr (r =.839), Cu and Ag (r =.881), and Pb
and Cr (r =.780). No element was correlated with Mn or
with ash content. Fig. 2a,b illustrates examples of these
statistical relationships.
The possibility that seaweeds may have been contami-
nated with metals scavenged by iron oxyhydroxides present
in fine sediment particles was considered and tested in two
ways: (1) assuming that all the iron in the seaweed samples
originated from sediment contamination, and (2) comparing
metal/Fe ratios in sediment and seaweed. Potential contam-
ination from sediment particles appears theoretically possi-
ble only for Pb.
3.2. Sediment
Observations of sediment samples with a binocular
microscope revealed that the sediments were largely com-
posed of quartz, carbonates, and feldspars, some coated by
oxides, and mixed with shell debris. Opaque minerals
(mostly pyrite) were present. Coal particles were also
common in all sediments. Very small grains could not be
identified. The fine sand fraction of surface sediments was
available from eight of the sites investigated, whereas the
pelitic fraction was available in sufficient amount only at six
of the sites sampled: At Easington and Horden, only sand
and gravel were present in the sediment (Table 1).
The concentrations (mean � S.D.) of heavy metals in
sediments are given in Table 5. Iron is the most abundant
(about 1±4.5 wt.%) of the metals analysed, and, exception
made for Blackhall Rocks, it is more enriched in the pelitic
fraction. Manganese is always the most abundant of the
trace metals in the pelite, with mean values ranging from
207 mg kgÿ 1 at Holy Island to 2597 mg kgÿ 1 at Roker. In
most cases, the least abundant elements are Ag and Cd.
For comparative purposes, Table 5 lists also the heavy
metal concentrations in sediments of the Tyne estuary (Bryan
and Langston, 1992) and in the continental crust (Wedepohl,
1995). Unfortunately, data sets cannot be readily compared
as different digestion methods were used on different sedi-
ment fractions. More meaningful estimates of sediment
contamination can be made with reference to background
values from the same area investigated. Rowlatt and Lovell
(1994) reported median values of Pb (11 mg kgÿ 1), Zn (15
mg kg ÿ 1), and Cr (18 mg kg ÿ 1) for < 2-mm seabed
sediments analysed in the Joint Monitoring Group Sediment
Baseline study. They studied the shelf region of the North
Atlantic and its marginal areas, including the North Sea. Pb,
Zn, and Cr background values from a sediment core taken 16
km off the Tyne estuary were 12, 38, and 45 mg kgÿ 1,
respectively (Rowlatt and Lovell, 1994). In general, these
authors found that the seabed sediments of Tyneside and
Teesside were above these background values. Even though
a direct comparison with our data cannot be made due to the
different size range of the sediment fraction analysed, these
values can be referred to as baseline concentrations.
The sediment quality guideline values proposed by Long
et al. (1995) are also listed in Table 5, as they can be used to
estimate the probability that adverse effects to the local biota
may occur. These effects are possible within the concentra-
tion range indicated, and frequently observed at higher
concentrations. These values refer to bulk sediments
digested with strong reagents and cannot be compared with
Table 7
Correlation coefficient matrix for heavy metals in aqua regia-digested
sediments ( < 63 mm fraction)
Fe Mn Zn Cu Pb Ni Cr Cd Ag % LOI
Fe .864 .666 .515 .689 .813 .784 .716 ÿ .152 .184
Mn .916 .322 .921 .612 .926 .910 ÿ .228 ÿ .084
Zn .232 .982 .516 .856 .801 ÿ .395 .045
Cu .212 .491 .253 ÿ .110 ÿ .545 .395
Pb .549 .868 .762 ÿ .386 .032
Ni .685 .000 ÿ .400 ÿ .300
Cr .740 ÿ .332 ÿ .173
Cd .460 ÿ .268
Ag ÿ .472
% LOI
Critical r = .468 at P < .05, or r = .590 at P < .01 (n = 18).
Table 6
Correlation coefficient matrix for heavy metals in aqua regia-digested
sediments ( < 180 mm fraction)
Fe Mn Zn Cu Pb Ni Cr Cd Ag % LOI
Fe .538 .649 .761 .833 .871 .247 .358 ÿ .425 .539
Mn .089 .141 .548 .401 .881 ÿ .370 ÿ .329 .055
Zn .793 .826 .761 ÿ .134 .900 ÿ .394 .579
Cu .716 .706 ÿ .205 .651 ÿ .369 .901
Pb .829 .308 .581 ÿ .510 .453
Ni .207 .469 ÿ .567 .468
Cr ÿ .430 ÿ .145 ÿ .228
Cd .134 .000
Ag ÿ .228
% LOI
Critical r = .404 at P < .05, or r = .515 at P < .01 (n = 24).
L. Giusti / Environment International 26 (2001) 275±286 281
our data listed in Table 5. However, it is possible to calculate
minimum bulk metal concentration, taking into account the
percentage of fine-grained sand and pelite, and assuming no
metal contribution from gravel and coarse/medium sand.
This means a dilution effect of about 50±60%. The recal-
culated sediment levels of Zn, Pb, Ni, Cd, and Ag for Roker,
Easington, Horden, Blackhall Rocks, and Middleton are
either within or higher than the guideline values of Long
et al. (1995).
In our study, the surface sediment at control Site 1
(Holy Island) has lower or similar Pb, Zn, and Cr con-
centrations than the background values of Rowatt and
Lovell (1994). Background Cr is exceeded only at Roker
and Middleton. All other sites show Zn and Pb contamina-
tion. Iron is significantly ( P < .01) correlated with many
elements in the fine-grained sand, including Mn, Zn, Cu,
Pb, and Ni, and with Mn, Zn, Pb, Ni, Cr, and Cd in the
pelitic fraction (Tables 6 and 7). Some of the deviations
from a linear trend are due to metal enrichment of sedi-
ments at some of the sites investigated. More specifically,
Pb contamination in the pelitic and sand fractions can be
inferred for Roker, Easington, Horden, and only for the
pelitic fraction at Blackhall Rocks. Both fractions are
enriched in Ag at Holy Island and Bran Sands. Examples
of scatter plots of Fe concentration (expressed as percent)
vs. the concentration of other metals (in mg kgÿ 1) are
shown in Fig. 3. The scatter found for the fine-grained
sand fraction is usually due to high metal values (espe-
cially Zn, Pb, and Cd) at Easington and Horden. Similar
trends were observed when these metals are plotted against
Mn (Fig. 4). Zinc enrichment at Easington and Horden is
also shown by the trends in the Zn vs. Pb and Zn vs. Cr
scatter diagrams (Figs. 5 and 6, respectively). If the
samples from Horden and Easington are excluded from
the correlation calculations, a highly significant ( P < .001)
correlation would be observed between Fe and Zn
(r =.872), Mn and Zn (r =.874), Mn and Pb (r =.968),
and Mn and Cd (r =.714) in the sand fraction. Silver is
negatively correlated to all elements except Cd. This may
be partially due to the similar speciation of these two
metals in aquatic environments.
The organic matter content (as % LOI) in sediments is
generally in the range 0.4±3.6%. The relatively higher
percentage found at Easington, Horden, North Gare, and
Fig. 4. Scatter plot of (a) Mn vs. Pb and (b) Mn vs. Zn concentration in sediment fractions (pelite and fine sand) from the coast of northeast England.
Fig. 3. Scatter plot of (a) Fe vs. Pb and (b) Fe vs. Zn concentration in sediment fractions (pelite and fine sand) from the coast of northeast England.
L. Giusti / Environment International 26 (2001) 275±286282
especially at Blackhall Rocks are partly due to the presence
of small particles of coal, especially in the sand fraction.
Copper shows the highest positive correlation with %
LOI (r =.901, P < .01) in the fine-grained sand (Fig. 7 and
Tables 6 and 7). The strong affinity of Cu for organic
material in sediments is quite well known (e.g. Luoma
and Bryan, 1981; Davies-Colley et al., 1984; Borg and
Jonsson, 1996), although other phases such as Fe±Mn
oxides and hydroxides can also be good Cu scavengers.
Other metals positively correlated with organic matter in the
sand fraction are Fe, Zn (both at P < .01), Pb, and Ni (both at
P < .05).
3.3. Comparison between heavy metal concentrations in
sediments and marine organisms
The aqua regia-extracted Zn, Cu, and Pb in sediments
are positively correlated ( P < .05) with those of F. vesicu-
losus, as shown for example in Fig. 8 for Zn. A sig-
nificant correlation ( P < .05) was also found for Ni
extracted from the sand fraction. These relationships
indicate that some of the metals held in the sediment
may become available to the seaweeds. This may occur,
for example, when Fe and/or Mn are remobilised (together
with the scavenged trace metals) from anoxic sediments
back into the water column. The biodegradation of organic
material will also cause a release of heavy metals to
marine water. Even though Ag is widely distributed in
all sediments, and especially those of Holy Island and
Bran Sands, this metal is more accumulated in F. vesicu-
losus at sites affected by mining activity.
Metal concentrations in surface sediments (Table 5) and
in the soft tissue of the mussel M. edulis (Giusti et al., 1999)
appear to be generally unrelated. It is possible that although
Fe oxides/hydroxides and organic matter can bind high
Fig. 7. Relationship between Cu concentration and organic matter content
(as % LOI) in marine sediments from the northeast coast of England. The
regression line and the correlation factor (r =.901) refer to the sand fraction
(full circles). The empty circles represent the pelitic fraction.
Fig. 8. Relationship between Zn concentrations in F. vesiculosus and Zn in
aqua regia extracts of sediment fractions from the same sites in northeast
England (n = 14, P < .05).
Fig. 6. Plot of Zn vs. Cr concentration in sediments from the coast of
northeast England, showing the Zn enrichment of sand samples from
Easington and Horden.
Fig. 5. Plot of Zn vs. Pb concentration in sediment fractions from the coast
of northeast England. The pelite is represented by empty circles, the fine
sand by full circles. The regression line and the correlation factor (r =.965)
refer to the pelitic fraction only (n = 18). For n = 18, at P < .01, the critical r
is .590.
L. Giusti / Environment International 26 (2001) 275±286 283
concentrations of trace metals, they also cause a reduction in
trace metal bioavailability in the digestive tract of mussels
(Luoma and Bryan, 1978; Langston, 1980). This is more
likely to result in a correlation of heavy metals in mussels
with the ratio metal/Fe in sediment. In our study, we found a
significant positive correlation (at P < .01) between Pb in
mussels and the Pb/Fe ratio in the aqua regia extracts of
surface sediments.
3.4. Metal pollution index (MPI)
The overall metal burden of F. vesiculosus was compared
with the total aqua regia-extracted metal content in sedi-
ments, using the MPI calculated with the formula (Usero et
al., 1996, 1997):
MPI � �M1 �M2 �M3 � . . .�Mn�1=n
where Mn is the concentration of metal n expressed in mg
kgÿ 1 of dry weight.
Lower MPI in algal material is to be expected, as
accumulated heavy metals derive only from dissolved
species present in marine water. Sediments are normally
larger repositories of contaminants. However, the variations
in algal MPI between sites appear to be similar, though of a
different order of magnitude, to the variations of sediment
(fine sand) MPI. Fig. 9 shows that the two sets of MPI are
significantly correlated (r =.959, n = 8). The correlation
between pelite MPI and algal MPI (n = 6) is not significant.
Both monitoring methods (i.e. with sediment and with
algae) have thus proved to be quite effective in highlighting
metal concentration gradients.
3.5. Comparison between heavy metal concentrations in
seaweeds and mussels
Along the coastline studied, the concentration of Cu, Zn,
Pb, Cd, and Ag in seaweed is usually lower than in soft
tissue of mussels (Giusti et al., 1999) from the same sites.
Only at Roker was the Zn accumulated by F. vesiculosus
higher than in M. edulis. Nickel is more abundant in
seaweeds than in mussels from Holy Island, Roker, and
Blackhall Rocks, and Cr levels are higher in seaweeds at
Holy Island and Middleton.
Our seaweed samples showed a more pronounced accu-
mulation of Mn than the mussels from the same sites (Giusti
et al., 1999). This appears to be quite commonly observed
elsewhere. For example, the average North Sea and Baltic
Sea Mn concentration in F. vesiculosus is 356 and 748 mg
kg ÿ 1 dry weight, respectively, which is one order of
magnitude higher than in mussels from the same areas, i.e.
29.7 and 47.2 mg kgÿ 1, respectively (Struck et al., 1997).
High levels of Mn or Zn in water are known to suppress
the accumulation of trace amounts of dissolved Cd, Co, Ni,
Zn, or Mn in seaweed as a result of competition for
available binding sites (Bryan et al., 1985). It is thus
possible that the low Zn, Ni, and Cd values in seaweeds
at Holy Island might be caused by the high accumulation of
Mn. Also, the general lack of a significant correlation
between Mn and any of the other metals analysed in
seaweed may be partially due to the same reason. In our
study, the significant positive correlation between Zn and
Cu in seaweed rules out large competition effects between
these two metals.
4. Conclusions
Given the seasonal and short-term variability of estuar-
ine and coastal environments, large variations in dis-
solved and particulate metal concentrations should be
expected, thus making the interpretation of heavy metal
distribution gradients quite problematic. Nonetheless,
some conclusions can be drawn based on the combina-
tion of information obtained from seaweed, sediment, and
previously published data on heavy metals in mussels
from the same sites.
(i) Seaweeds from Whitburn, Roker, Easington, and
Horden, have a relatively high burden of Zn (511±1016