ARTICLE IN PRESS
Applied Geochemistry xxx (2005) xxx–xxx
www.elsevier.com/locate/apgeochem
Mercury contamination in Marano Lagoon
(Northern Adriatic sea, Italy): Source identiWcation
by analyses of Hg phases
RaVaella Piani a, Stefano Covelli
a
a,¤
, Harald Biester
b
Department of Geological, Marine and Environmental Sciences, University of Trieste, Via Weiss, 2, 34127 Trieste, Italy
b
Institute of Environmental Geochemistry, INF 236, University of Heidelberg, 69120 Heidelberg, Germany
Received 6 September 2004; accepted 5 April 2005
Editorial handling by W.B. Lyons
Abstract
Total Hg concentrations and Hg speciation were determined in bottom sediments of Marano lagoon to investigate
the consequences of Hg phases on Wsh farms and shellWsh cultivation areas. Mercury phases were separated into cinnabar (HgS) and non-cinnabar compounds, via a thermo-desorption technique, in surface and core sediments; both of
which had been contaminated by industrial wastes and mining activity residues. The former are due to an industrial
complex, which has been producing cellulose, chlor-alkali and textile artiWcial Wbres since 1940. Processing and seepage
wastewaters, which were historically discharged into the Aussa-Corno river system and therefore into the lagoon, have
been signiWcantly reduced since 1984 due to the construction of wastewater treatment facilities. The second source is the
Isonzo River, which has been the largest contributor of Hg into the northern Adriatic Sea since the 16th century due to
Hg mining at the Idrija mine (western Slovenia). Red cinnabar (HgS) derived from the mining area is mostly stable and
insoluble under current environmental conditions. In contrast, organically bound Hg, such as Hg bound to humic acids,
has the potential to be transformed into bioavailable Hg compounds (for example, methylmercury). The presence of the
two Hg forms permitted each Hg source to be quantiWed. It also allowed the areas with the highest risk of Hg contamination from Hg-rich sediment to be identiWed; thus potentially avoiding the transfer of Hg from the sediment into the
water column and eventually into living biota. The results show that Hg Enrichment Factors in bottom sediments
exceed values of 10 and cinnabar dominates the central sector near the main tidal channel where tidal Xux is more eVective. Non-cinnabar compounds were found to be enriched in Wne grained material and organic matter. In fact, up to 98%
of total Hg at the Aussa-Corno river mouth and in the inner margin of the basin occurred in an organic form. This evidence, combined with the high contents of total Hg (4.1–6.6 g g¡1 and EF > 10) measured in surface sediments, suggest
that Hg in Marano lagoon is involved in biogeochemical transformations (e.g., methylation).
2005 Elsevier Ltd. All rights reserved.
1. Introduction
¤
Corresponding author. Tel.: +39 040 5582031; fax: +39 040
5582048.
E-mail address: covelli@univ.trieste.it (S. Covelli).
0883-2927/$ - see front matter 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.apgeochem.2005.04.003
The biogeochemistry of Hg has become one of the
most important topics in environmental sciences in the
last 4 decades. This is due, in part, to Hg dispersion on a
global scale, its extreme toxicity at low levels, and its
ARTICLE IN PRESS
2
R. Piani et al. / Applied Geochemistry xxx (2005) xxx–xxx
tendency to bioaccumulate within food chains (Mastrine
et al., 1999; WHO, 1990; Porcella, 1994; Salomons,
1997). The ecological and human health eVects of Hg are
generally related to the environmental transformation of
inorganic Hg to the more toxic methylmercury and biomagniWcation in ecological systems (Compeau and Bartha, 1985). Moreover, several studies have demonstrated
that wetlands are important areas for Hg methylation
(Kierfve, 1994; King et al., 2002; Lacerda and Gonçalves,
2001) and that SO4-reducing bacteria (SRB) are the
major biological mediators of Hg methylation in sediments (Gilmour et al., 1992; King et al., 1999). Also, particulate and dissolved organic C play a signiWcant role in
the biogeochemical cycling of Hg in the aquatic environment (Gill and Bruland, 1990; Mason et al., 1993; Stordal et al., 1996). Possible remobilisation of heavy metal
compounds, and Hg in particular, from bottom sediments into the water column of coastal lagoon ecosystems may result in variable accumulation by the lagoon
biota (Lacerda et al., 1992). The presence of Hg in
aquatic systems is due to several inputs, such as atmospheric deposition (Marins et al., 1996), and/or mining
and industrial discharges (Baldi and Bargagli, 1984;
Hines et al., 2000; Biester et al., 2002). Speciation of Hg
in the water column of coastal lagoons may be of great
environmental signiWcance, since it will not only control
the eventual transfer of Hg to coastal waters but also the
capacity of water bodies to accumulate large amounts of
Hg in their sediments. As such, it is important to determine the distribution and the forms of this metal in contaminated sediments in order to establish potential risks
and appropriate remediation technologies (Kaplan et al.,
2002). Sequential extraction techniques have been widely
used in order to diVerentiate between anthropogenic and
geogenic Hg sources in soils and sediments and to evaluate the potential for remobilisation of Hg into the aqueous phase (Wallschläger et al., 1995; Bloom et al., 2003).
Alternatively, Hg-thermo-desorption techniques have
been successfully applied to distinguish metallic Hg from
Hg(II)-binding forms; for example, between humic acids
and sulphides in Xuvial and marine sediments (Biester
and Scholz, 1996; Biester et al., 2000, 2002; Sladek et al.,
2002).
The Po River and the industrial zone of the Venice
Lagoon are the primary sources of Hg contamination in
coastal sediments of the Northern Adriatic sea (Fabbri
et al., 2001; Donazzolo et al., 1981). Further east, the
Isonzo River supplies suspended material enriched in Hg
to the Gulf of Trieste from sediments and tailings of the
Idrija mining area (western Slovenia). Since late 1970
several investigations on coastal sediments of the Gulf of
Trieste have reported high concentrations of total Hg
(23.6–47.0 g g¡1), which exponentially decrease with distance from the Xuvial source (Faganeli et al., 1991;
Covelli et al., 2001). The Isonzo river inputs are dispersing westward and out of the gulf, following the main cir-
culation system, and are also contributing to the Hg load
in bottom sediments and, most importantly, to biota of
the adjacent Grado and Marano lagoons (Brambati,
2001). Biester et al. (2000) have shown that cinnabar is
the predominant form of Hg at the Isonzo River mouth,
whereas organically bound Hg forms, which are mainly
associated with Wne particles, are subject to long range
transport. In addition, the inXux of this element into the
western part of the lagoon system was signiWcantly
enhanced by an industrial complex located along a
stream Xowing into the lagoon. The industrial plant has
been producing cellulose, chlor-alkali and textil artiWcial
Wbres since the 1940s (RFVG, 1991; Marocco, 1995;
Brambati, 1997).
The aim of this study was to determine and quantify
the proportion of cinnabar (HgS) and non-cinnabar
compounds by a thermo-desorption technique in recent
sediments of the Marano lagoon. As the industrial
complex is a major source of non-cinnabar Hg compounds, attention was speciWcally given to locating
areas of prevalent organomercury accumulation in
comparison to sulphide-bound Hg. The circulation system and the relationships between Hg compounds and
sediment particle characteristics were also investigated
in order to recognise sites of potentially bioavailable
Hg and the transfer of these Hg species from lagoon
sediments into the water column and eventually into
biota. Bioaccumulation, biotransfer and bioconcentration of Hg in the local environment is of direct health
concern, as bivalves (clams and mussels), and a large
number of Wsh farms in the lagoon are of local economic importance.
2. Environmental setting
The Marano and Grado lagoons (Fig. 1(a)) are part
of the extensive system of transitional environments
developing along the northern Adriatic coast from the
Po river delta to the Isonzo river mouth in the Gulf of
Trieste. The Marano and Grado lagoons extend for
about 32 km, between the Tagliamento and Isonzo River
deltas, reaching up to 5 km for a total area of 160 km2.
The lagoon basin is characterised by semi-diurnal tidal
Xuxes (65 and 105 cm mean and spring tidal range,
respectively). Small rivers Xow into the lagoon, which
drain waters coming from the spring line and those from
irrigation canals, which are conveyed into the water
scooping machines. Particulate matter from these
streams is of secondary importance, restricted to areas
surrounding the spring river mouths. Conversely, the
primary source of suspended sediments arrives from the
sea, as the contribution of river deltas and from erosion
of the barrier islands. Dispersion of sediments into the
lagoon is controlled by tidal Xuxes through tidal inlets
(Brambati, 1970).
ARTICLE IN PRESS
R. Piani et al. / Applied Geochemistry xxx (2005) xxx–xxx
3
Fig. 1. Central sector (Buso basin) of the Marano and Grado lagoons (a) and location of sampling points (b). Circles indicate bottom
samples and black squares show core sites. Fish farm sites are also reported.
The sampling area is located in the Buso basin (Fig.
1(b)), which occupies the central position of the lagoon
system and presently has a surface area of 45 km2. The
width of the Buso inlet was modiWed and reduced to
450 m after the building of jetties from 1964 to 1969.
Commercial shipping through the main channel is maintained by periodical dredging, which ensures a channel
depth of approximately 8 m along its axis. The most
important freshwater stream Xowing into the Buso basin
is the Aussa-Corno River, which has a steady Xow in the
order of 25–27 m3 s¡1 (RFVG, 1983). The suspended
load reaches 8–10 mg L¡1 during normal run oV periods
(Piani, 1992 unpublished data). The Zellina River, which
is the second and less inXuential tributary of Buso basin,
has a Xow rate of just 1–2 m3 s¡1 (RFVG, 1983).
The general grain-size distribution observed for the
whole lagoon is also found in the Buso basin (Piani and
Covelli, 2000). Medium to Wne sands are only present in
front of the Buso inlet, along the main channel and the
two lateral secondary channels and, subordinately, at the
lower limit of the tidal Xats. The Wne pelitic component is
widely dispersed in the basin and varies from 52% to
100%. The silty fraction (52–2 m) normally comprises a
higher percentage than the clay fraction (<2 m) in the
pelitic component. In fact, the ratio Wne-silt/clay (2–16/
<2 m) identiWes areas aVected by freshwater inputs into
the lagoon: a high ratio indicates an area of fresh water
input and vice versa. The highest contents of Wne silt
occur where the Zellina and Aussa-Corno rivers enter
the lagoon. Moreover, areas aVected by fresh water
inputs and the inner part of the basin show the highest
values of organic C (0.2–3.5%) and C/N ratios (up to 14).
This suggests that organic matter of prevalently continental origin is partially preserved in those sectors where
reducing conditions (Eh 6 ¡300 mV) in bottom
sediments occur due to limited Xuvial inXuence and
ARTICLE IN PRESS
4
R. Piani et al. / Applied Geochemistry xxx (2005) xxx–xxx
interchange of marine waters through tidal cycles
(Brambati, 1997).
This area of the lagoon has high vulnerability due to
the substantial human impact caused by urban settlements and industrial activities. Porto Nogaro, located
1 km north of the Aussa-Corno mouth is the most
important industrial site in the region. In addition, the
Torviscosa industrial complex produced cellulose from
cane (Arundo donax sp.) from the 1940s until 1992. It is
estimated that about 20 kg day¡1 of Hg has discharged
into the Aussa River since 1949 (when the chlor-alkali
plant became operative). As reported in RFVG (1991),
soda production was increased from 4500 tons in 1950 to
20,000 tons during the following 10 a. The decrease of
Hg discharged amounted to 6–7 kg day¡1 in 1970 and it
presumably stopped in 1984 when wastewater treatment
systems were installed. At the present time, industrial
eZuents Xow through submarine pipes that discharge
processing eZuent into the open sea; about 6.5 km
oVshore from the main tidal inlet. According to the
report mentioned above, an estimated total amount of
186,000 kg of Hg was released into the river. It is likely
that a signiWcant volume of this Hg reached the lagoon
environment. This site is currently under intense investigation and some corrective actions (e.g., soil removal
and groundwater treatment) are in progress. Although
high levels of Hg in bottom sediments of the Gulf of
Trieste revealed a very strong supply of suspended material from the Isonzo River due to the long-term mining
activity in the Idrija region (Brambati, 2001; Covelli
et al., 2001), no estimation of the contributions from this
source to the total amount of Hg into the lagoon has
been attempted.
3. Materials and methods
A total of 41 bottom sediment samples were recovered by a Van Veen grab and two cores (sampling stations 27 and 40) were hand collected from both sides of
the main navigation channel (Fig. 1(b)), which connects
the tidal inlet to the Aussa-Corno river mouth. Organic
debris and shelly fragments were removed by sieving
(<2 mm) each sample prior to grain-size analysis. Subsamples from each core and grab sample were subjected
to 24 h of H2O2 treatment to remove organic matter, and
subsequently wet sieved using a 52-m sieve to separate
sand from the pelitic fraction. Additionally, pelite was
treated with H2O2 (10 v/v) prior to being analysed by a
Micromeritics Sedigraph 5000 ET.
The organic C and total N were determined on 40 °C
dried and homogenised samples, after acidiWcation with
progressively increasing concentration of HCl (Hedges
and Stern, 1984), using a Perkin–Elmer C–H–N Elemental Analyser at a combustion temperature of
920 °C.
Sediments were digested with HNO3 and H2SO4 solution (IRSA-CNR, 1985) and total Hg was measured following a modiWed form of CV AAS procedure by using
a Perkin–Elmer Mercury Hydride System coupled to an
AAS Perkin–Elmer mod.380.
Determination of Hg phases by solid-phase-Hgthermo-desorption is based on the speciWc thermal
decomposition of Hg compounds from sub-samples of
dry sediment at diVerent temperatures. A detailed description of the method is found elsewhere (Biester and Zimmer, 1998; Biester et al., 2000). Results are represented by
Hg-thermo-desorption curves showing release of Hg(0)
versus increasing temperature. Mercury desorbed from
matrix-bound Hg, such as Hg associated with humic
acids, is in general normally distributed, in contrast to cinnabar which is irregularly released during thermo-desorption. Mercury peaks were then quantiWed by peak
integration. Since non-cinnabar and cinnabar peaks usually partially overlap, non-cinnabar Hg was calculated by
doubling the Wrst half of the corresponding peak (150–
250 °C). Therefore, Hg bound in cinnabar was obtained as
the diVerence between total Hg and non-cinnabar Hg.
Geochemical data for the overall Buso basin surface
area were processed using the Surfer software program
(Kriging algorithm).
4. Results and discussion
4.1. Mercury occurrence in the lagoon sediments
The high vulnerability of coastal lagoon ecosystems is
related to the shallowness and the limited water
exchange of these semi-enclosed basins. Due to these features, the residence time of water, sediments and, consequently, contaminants, in the lagoon is much longer than
in river mouths and in open coastal areas. Heavy metals,
and in particular, Hg accumulated in sediments may be
subjected to burial and/or to biogeochemical processes,
which aVect their distribution, speciation and bioavailability to the lagoon biota. The high potential risk of Hg
remobilisation from sediments into the water column
and its subsequent bioaccumulation in the trophic chain
is also enhanced by the abundance of animal and plant
species in this ecosystem. Human activities such as Wshing, collection of clams and mussels and cultivation of
edible Wsh species in Wsh farms increase chances of Hg
uptake by living organisms and, ultimately, by human
residents in the surrounding area.
Mercury levels in aquatic plants and organisms from
the trophic chains (Brambati, 1997) of the Marano and
Grado lagoons suggest that tissue Hg is correlated with
bottom sediment Hg concentrations, which show a progressive westward decrease from about 10 mg kg¡1
(Grado lagoon) to 1 mg kg¡1 (Marano lagoon). However, diVerent concentrations of Hg found in aquatic
ARTICLE IN PRESS
R. Piani et al. / Applied Geochemistry xxx (2005) xxx–xxx
5
Fig. 2. Distribution of total Hg (a) in g g¡1 and organic C (b) in % for the surface sediments of the lagoon basin.
species are related to their trophic characteristics. For
example, mussel (Mytilus) and common cockle (Cardium
edule) which are Wlter feeders show low concentrations;
higher concentrations were found in prawn (Palaemon) a
coarse detritus feeder; whereas very high levels (average
0.5–0.7 mg kg¡1, maximum 2–3 mg kg¡1) were detected in
the carnivorous goby (Zosterissessor ophiocephalus) and
silverside (Atherina boyeri). The same high Hg anomalies
were observed in the Wsh farms where edible Wsh (e.g.,
gilt-head bream, Sparus auratus and bass, Dicentrarchus
labrax), which are at the top of the food chain, showed
levels far higher (up to about 5 mg kg¡1) than the same
species collected in the natural environment. Unfortunately, Hg biomagniWcation increases the total Hg levels
in marine species of commercial value (Brambati, 2001),
and the magniWcation process seems to be more eVective
in the Wsh farms. Brambati (2001) also found high Hg
concentrations in the hair of people having a Wsh-based
ARTICLE IN PRESS
6
R. Piani et al. / Applied Geochemistry xxx (2005) xxx–xxx
Table 1
Relative abundances of main textural and geochemical parameters in surface sediment samples
Sample
Sand (%)
Silt (%)
Clay (%)
2–16 m (%)
C org. (%)
Hg (g g¡1)
RP1
RP2
RP3
RP4
RP5
RP6
RP7
RP8
RP9
RP10
RP11
RP12
RP13
RP14
RP15
RP16
RP17
RP18
RP19
RP20
RP21
RP22
RP23
RP24
RP25
RP26
RP27
RP28
RP29
RP30
RP31
RP32
RP33
RP34
RP35
RP36
RP37
RP38
RP39
RP40
RP4A
0.7
2.5
16.4
0.4
1.5
3.3
1.0
3.0
19.2
13.4
11.2
5.6
98.3
47.8
30.5
29.3
8.7
0.9
0.4
0.5
5.0
30.8
4.4
6.1
5.6
7.4
9.6
18.0
16.9
12.0
16.2
3.1
2.1
8.7
21.1
18.6
5.0
7.2
13.6
5.3
0.2
71.3
70.5
62.6
70.1
80.5
75.7
72.0
72.0
55.8
64.1
71.8
74.4
1.7
33.7
36.0
36.2
67.3
62.1
61.6
65.5
71.5
42.2
70.6
67.9
71.4
74.1
65.4
57.0
59.1
66.0
62.8
74.9
79.4
73.8
60.9
56.4
63.0
76.8
57.4
71.2
66.8
28.0
27.0
21.0
29.5
18.0
21.0
27.0
25.0
25.0
22.5
17.0
20.0
0
18.5
33.5
34.5
24.0
37.0
38.0
34.0
23.5
27.0
25.0
26.0
23.0
18.5
25.0
25.0
24.0
22.0
21.0
22.0
18.5
17.5
18.0
25.0
32.0
16.0
29.0
23.5
33.0
42.0
51.0
45.5
51.5
49.5
31.5
46.0
33.0
36.5
35.5
24.0
28.0
0
23.5
39.5
42.0
36.0
48.5
39.5
40.0
28.5
32.0
36.5
32.5
32.0
43.5
30.0
32.0
31.0
27.5
27.0
31.0
33.5
22.5
24.0
33.0
41.0
36.0
38.0
44.5
58.5
1.92
2.18
1.81
3.06
2.51
1.61
2.62
1.81
1.29
1.40
1.28
1.33
0.21
0.75
1.17
1.22
1.68
3.48
2.06
2.08
1.29
1.66
1.53
1.41
1.59
1.64
1.36
1.60
1.30
1.27
1.23
1.42
1.60
1.04
1.07
1.41
2.21
1.64
1.63
2.03
3.20
3.61
3.52
3.51
4.70
4.77
5.48
4.62
3.98
3.03
3.66
3.69
3.57
0.13
3.07
3.53
3.89
5.73
4.89
5.24
5.62
4.14
5.63
6.19
4.95
4.56
4.09
3.45
4.57
2.74
4.11
3.20
6.58
3.21
3.55
4.81
2.14
4.50
3.57
3.57
4.69
5.50
diet (up to 20 mg kg¡1) and Hg levels showed a direct
positive correlation with age. Furthermore, high levels of
Hg were also found in birds, demonstrating that the top
consumers in the area may bioaccumulate Hg.
Results of textural and geochemical parameters of
the study area reported by Piani and Covelli (2000)
revealed that major and trace elements (Al, Fe, K and
Zn, Cr and Ni) as well as organic C, are signiWcantly correlated with Wne particles (<16 m) and they reXect the
distribution pattern of this textural component in the
lagoon basin.
The anomalously high concentrations of Hg occur
near the Zellina and Aussa-Corno river mouths and on
tidal Xats in the eastern part of the basin (Fig. 2(a) and
Table 1). Due to the double source of contamination
mentioned above, Hg concentrations in sediments, ranging between 0.1 and 6.6 g g¡1, are on average
(4 § 1.2 g g¡1) far higher than natural background values proposed in previous studies for the northern Adriatic sea (0.1 g g¡1, Faganeli et al., 1991; 0.2 g g¡1, Covelli
et al., 2001). Higher concentrations of organic C are found
in the inner and western parts of the basin where bottom
sediments are mostly pelites rich with Wne silt (Fig. 2(b)).
The two short cores do not show noticeable grainsize variations, although a slight increase in the coarse
sandy fraction was detected at the top of core 27. This
core was collected in the western part of the basin where
salt marshes are absent and tidal Xuxes are more eYcient
ARTICLE IN PRESS
7
R. Piani et al. / Applied Geochemistry xxx (2005) xxx–xxx
Table 2
Relative abundances of main textural and geochemical parameters in core samples
Depth (cm)
Sand (%)
Silt (%)
Clay (%)
2–16 m (%)
C org. (%)
Hg (g g¡1)
Core 40
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0–3
3–5
5–7
7–9
9–12
12–14
14–16
16–18
18–20
20–22
24–26
28–30
33–35
37–39
41–43
44–46
48–50
54–55
58–60
62–64
1.1
0.8
0.5
0.4
1.0
0.0
0.1
0.5
0.5
0.4
0.4
1.2
0.2
1.1
0.3
0.3
0.3
1.2
0.8
0.6
74.5
68.8
73.4
65.8
67.2
60.2
62.8
63.4
64.4
63.0
65.3
70.0
67.3
76.9
75.9
71.6
68.7
75.2
73.0
73.2
24.4
30.4
26.1
33.8
31.8
39.8
37.1
36.1
35.1
36.6
34.3
28.8
32.5
22.0
23.8
28.1
31.0
23.6
26.2
26.2
45.2
40.6
44.5
44.8
42.4
41.1
44.4
44.5
43.5
42.5
43.4
41.3
48.9
42.2
45.3
47.0
51.2
39.1
42.1
42.7
2.07
1.98
1.91
1.75
1.66
2.03
1.88
1.84
1.96
1.80
1.62
1.63
1.45
1.76
1.53
2.12
2.87
1.41
–
–
5.80
7.19
7.43
8.41
6.55
9.84
5.99
4.23
4.13
3.58
3.55
5.13
3.40
3.42
2.09
2.14
1.10
0.52
0.63
0.41
Core 27
1
2
3
4
5
6
7
8
9
10
12
13
0–3
5–7
9–11
13–15
17–19
21–23
25–27
29–31
33–35
37–39
41–43
45–47
0.7
1.8
2.3
1.9
0.1
0.4
0.6
0.9
0.4
0.1
1.3
0.4
70.3
73.4
76.7
74.7
69.7
70.6
72.3
71.2
72.0
68.1
71.1
64.9
29.0
24.8
21.0
23.4
30.2
29.0
27.1
27.9
27.6
31.8
27.6
34.7
41.9
37.0
34.3
35.9
45.1
44.7
44.4
45.1
45.0
46.3
44.9
46.9
1.38
1.27
1.06
1.27
1.22
1.24
1.28
1.28
1.19
1.38
1.41
1.57
2.15
2.74
2.24
2.66
2.46
0.85
0.39
0.42
0.41
0.63
0.32
0.30
Fig. 3. Calculated contribution of cinnabar and non-cinnabar Hg compounds in surface sediment samples as estimated by peak integration. The cumulative bar charts provide the amount of both fractions related to total Hg content.
ARTICLE IN PRESS
8
R. Piani et al. / Applied Geochemistry xxx (2005) xxx–xxx
Fig. 5. Areal distribution of the amount of non-cinnabar compounds, expressed as percentage of the total Hg content in the
surface sediments of the lagoon.
Fig. 4. Typical Hg-thermo-desorption curves show surface sediment samples where Hg is predominantly present as non-cinnabar (sample 18) and cinnabar (sample 14) compounds, and
where cinnabar and non-cinnabar compounds are almost
equally mixed (sample 24).
on the mudXats. In contrast, core 40 was sampled on a
tidal Xat near a salt marsh area, which is typically a low
energy environment. Major and trace elements do not
show a signiWcant vertical trend along the core proWles
(Piani and Covelli, 2000). The only exception is Hg that
is detected at a constant value of about 2 g g¡1 in the
Wrst 20 cm of core 27. Higher concentrations are found in
core 40 where subsurface maximum is reached at 13 cm
depth (9.9 g g¡1) and Hg decreases downcore approaching background values at 50 cm depth (Table 2).
Many trace elements are known to be preferentially
associated with Wne particles in aquatic systems. Therefore, a common approach is to normalise metal concentrations to account for grain-size variability prior to
assessing the degree of contamination (Loring and Rantala, 1992). The degree of enrichment of a metal is then
calculated by dividing its ratio to the normalising element by the same ratio found in the baseline (e.g., Middleton and Grant, 1990). A non-dimensional enrichment
factor EF D (Me/N)sample/(Me/N)baseline is obtained,
where Me is the concentration of the potentially
enriched metal and N is the concentration of the normalising element. A value of unity denotes no enrichment or
depletion relative to the pre-industrial datum. Iron was
chosen as the normalising element due to its signiWcant
correlation with the <16 m fraction (r D 0.87; p < 0.001)
and the local baseline was assumed to be the average
concentration of the two elements in the basal levels of
the two cores (Febaseline D 2.11%, from Piani and Covelli,
2000).
Mercury enrichment factors indicate that the centralsouthern sector of the basin along with the inner coastline extending from the Aussa-Corno river mouth westwards are preferential accumulation areas (EF > 10). In
the latter, the higher Hg enrichment (maximum 18) can
be explained by freshwater input, whereas in the Wrst
ARTICLE IN PRESS
R. Piani et al. / Applied Geochemistry xxx (2005) xxx–xxx
9
Fig. 6. Calculated contribution of cinnabar and non-cinnabar Hg compounds in core 27 and 40 sediment samples as estimated by peak
integration. The cumulative bar charts provide the amount of both fractions related to the total Hg content downcore.
area periodic dredging operations in the main channel
could be considered as a factor that contributed to
increase metal concentration in sediments. Also, dumping of excavated material in the mudXats, immediately
adjacent to tidal channels, was a common procedure followed in the past. These operations were recently ceased
due to introduction of severe national legislation rules.
Therefore, subsequent reworking and dispersion of
dredged sediments through tidal Xuxes cannot be
excluded and EF areal distributions may reXect this
complex situation.
The constant Hg concentrations in the upper part of
core 27 (Table 2) appear to be related to mixing eVects
due to bioturbation in the Wrst 20 cm, which would have
homogenised recent metal supplies. This evidence was
conWrmed by X-ray radiography of the core, which
shows concentrations of small tube-like burrows at the
core top. Conversely, contamination in the upper section
of core 40, extending downcore to about 50 cm, shows
two apparent maxima (Table 2). The relevant diVerence
between the two cores, in terms of thickness of contaminated sediment suggests two hypotheses.
The core 40 proWle could be explained as the original
sedimentary sequence, which was altered by anthropogenic activities, although dumping of excavated muds
seems to be improbable due to the distance of the coring
site from the main navigation channels. A second
hypothesis, supported by 137Cs proWles in the same cores
(Piani and Covelli, 2000), suggests that the two sampling
sites are characterised by diVerent local sedimentation
rates. This has also been reported for example by Pavoni
et al. (1987) and Donazzolo et al. (1981) for the Venice
lagoon and depends on variable energy hydrodynamic
conditions from the lower to the upper mudXats.
4.2. Mercury phases in bottom sediments
Results of Hg thermo-desorption measurements performed on bottom samples demonstrate that both cinnabar and non-cinnabar compounds are present in lagoon
sediments (Fig. 3) and that the relative contribution to
total Hg contents can be assessed. The speciWc Hg phase
in the sediments is the most important factor in terms of
potential bioavailability of Hg to the aquatic food chain
ARTICLE IN PRESS
10
R. Piani et al. / Applied Geochemistry xxx (2005) xxx–xxx
Fig. 7. Distribution of non-cinnabar compounds, expressed as percentage of the total Hg content, in the sedimentary sequence of core
40 and 27. Hg-thermo-desorption curves for each core level are also reported.
rather than the total Hg concentration. Presence and
abundance of the non-cinnabar component in bottom
sediments of some sampling sites may identify those
areas in the lagoon where Hg is present in chemical forms
that may be more mobile and/or involved in methylation
processes.
Almost all samples show curves with two peaks. A
Wrst maximum was detected between 230 and 260 °C,
which represents Hg released from non-cinnabar compounds such as humic acids, according to the curves of
standard Hg compounds reported by Biester et al.
(2000). Similarly, the second peak (300–350°C) is related
to red cinnabar. As noted by the same authors, several
high narrow Hg peaks usually appear in the curves due
to slight diVerences in the crystallinity and grain-size of
cinnabar, which inXuences Hg release temperatures.
Three types of Hg-thermo-desorption curves were
obtained for surface sediments (Fig. 4). Just one sample
taken upstream in the Aussa-Corno River is almost
entirely characterised by the non-cinnabar peak. Due to
the low temperature of this Wrst peak, Hg could be associated with organosulphides or occurs as metacinnabar
as a result of absorption of Hg(II) or Hg(0), by the prevailing silty-clay fraction with higher speciWc surface
area, and subsequent transformation. In contrast,
thermo-desorption curves of samples (e.g., 14 and 36)
collected near the lagoon inlet and in the eastern part of
the basin show Hg release temperatures mostly related
to red crystalline cinnabar. All other samples from the
lagoon show curves with two peaks, thus representing an
evident mixing of both Hg compounds in surface sediments as a result of tidal dynamics. The spatial distribution of non-cinnabar compounds (Fig. 5 and also Fig. 3)
is clearly inXuenced by the freshwater inputs (AussaCorno and Zellina rivers). Non-cinnabar Hg is relatively
more abundant, especially in the lower course of the tributary streams and in the lagoon just in front of the river
mouths. This evidence conWrms that most of the Hg dis-
ARTICLE IN PRESS
R. Piani et al. / Applied Geochemistry xxx (2005) xxx–xxx
11
Fig. 8. Relations between non-cinnabar Hg and organic carbon, and the 2–16 m fraction.
charged into the Aussa River from the industrial complex preferentially accumulates in the river itself and in
the inner part of the lagoon.
In contrast, cinnabar, although present at all locations, is mostly concentrated near the southern and eastern sectors of the basin. This fact testiWes to the role of
tidal Xuxes in controlling dispersion and accumulation of
cinnabar bound to suspended matter from Isonzo River
and entering the lagoon through at least 3 of the 4 inlets.
4.3. Mercury phases in core sediments
Occurrence of Hg phases in core 27 and 40 (Fig. 6)
shows an almost equal distribution of cinnabar and noncinnabar compounds in the upper section of the sedimentary sequence. In contrast, Hg sulphides decrease downcore. Only non-cinnabar compounds are present in the
basal part of the cores where Hg concentrations become
very low (Fig. 7). The sediment core from the Gulf of Trieste (Covelli et al., 2001) was mostly composed of Wne
grained sediments, which have recorded the historical evolution of Hg production at Idrija mine through an exponential increase of metal concentrations since 1800
(although mining activity began in 1496). Between 1850
and 1970 the amount of excavated ore increased as much
as the Hg production although the ore grade decreased
from several tens to less than 1%. This activity produced
an increasing amount of mining residues along with an
improved eYciency of the roasting processes, including
intensive crushing of the ore, which would have signiWcantly increased the amount of Wne grained material in the
mining residues (Biester et al., 2000). The large amount of
Hg enriched Wne particles (silt-clay) transported by freshwaters of the Isonzo river into the Gulf of Trieste would
ARTICLE IN PRESS
12
R. Piani et al. / Applied Geochemistry xxx (2005) xxx–xxx
have also entered the lagoons following the anticlockwise
circulation system characteristic of this part of the northern Adriatic sea and Xood tidal Xuxes.
Comparing Hg compounds with other textural and
geochemical parameters (Fig. 8), the best correlations
were found between Hg and organic C (r D 0.83) and the
2–16 m fraction (r D 0.86), suggesting that both the Wne
particles and organic matter play an important role in
transferring and accumulating the metal in the lagoon
sediments. Although Hg behaviour in this system
depends on its chemical form introduced into the
aquatic environment, Hg associated with industrial eZuents has been shown to be positively correlated with Wne
grain sizes such as <63 m (Rae and Aston, 1981) or
<20 m (Barghigiani et al., 1996); Hg also has a strong
aYnity for organic matter (Smith and Loring, 1981;
Baldi and Bargagli, 1984; Gagnon et al., 1997). In areas
where chlor-alkali plants are responsible for Hg contamination of sediments, up to 70–90% of the total metal in
sediments can be associated with easily oxidizable
organic matter, which is easily degraded and may then
be released from the substrate (Smith and Loring, 1981;
Gagnon et al., 1997; Biester et al., 2002).
Distinction between detrital Hg, i.e. cinnabar particles,
settling in deltaic sandy sediments in front of the Isonzo
river mouth and non-cinnabar compounds accumulated
in the central sector of the Gulf of Trieste was made on
the basis of correlation with the <16 m grain size fraction
(Covelli et al., 2001). According to these authors, Hg is
preferentially adsorbed to the Wnest particles in the form
of Hg(II), Hg(I) and Hg(0), due the higher speciWc surface
area of each particle; although the presence of micro-crystalline cinnabar cannot be excluded.
forms as a consequence of recent industrial inputs. The
distinction between the two Hg forms allows the areas at
most risk from Hg remobilisation from bottom sediments
to the water column and, possible interaction with lagoon
biota to be identiWed. The main risk areas occur near some
Wsh farms, which are protected from high energy regimes.
For example, the areas of most concern are in front of the
Zellina river mouth and the eastern part of the basin, as
well as the whole Aussa-Corno River mouth. The relevant, although extremely variable, thickness of sediments
contaminated by Hg, the high percentage of non-cinnabar
compounds, and the extension of the basin make it diYcult to devise an appropriate, in situ, remediation strategy.
Due to the importance of local Wsh farm production, the
most suitable choice would be to concentrate the eVorts
on the bottom sediments and remediation of restricted
pools. For example, excavation and removal of the contaminated layer and a reduction of new Hg supplies
bound to suspended matter through tidal Xuxes Xowing
periodically into the Wsh farms could be the most Wtting
procedures.
Acknowledgements
This work has been supported by Regione Autonoma
Friuli-Venezia Giulia (Direzione Regionale dell’Istruzione e Cultura) in the project “Rimobilizzazione e biodisponibilità del mercurio in ambienti lagunari e Xuviali
regionali” (resp. Prof. A.Brambati). A special thanks to
Ben Gildfedder for his valuable editing contribution.
References
5. Conclusions
Marano and Grado lagoons have experienced signiWcant historic Hg inputs from mining and industrial point
sources. The Buso basin, located in the central sector of
this lagoon, has been severely aVected both by long-term
Hg contamination through Isonzo river suspended load
and by untreated eZuents discharged into the AussaCorno river from a chlor-alkali plant since 1950. This
study has shown that the Hg-thermo-desorption technique is a useful method to discriminate between organically bound Hg and sulphide (principally cinnabar) bound
Hg. Using this method, the relative contribution of each
river system to the inXux of Hg into the lagoon can be
determined. Cinnabar from Isonzo River dominates the
central sector of the lagoon, near the main tidal channel
where tidal Xux is more eVective. Non-cinnabar compounds (avg. 49 § 22%) were found to be enriched in Wne
grained material and organic matter in the inner margin
of the basin, particularly at the Aussa-Corno River mouth
where up to 98% of total Hg is bound in non-cinnbar
Baldi, F., Bargagli, R., 1984. Mercury pollution in marine sediments near a chlor-alkali plant: distribution and availability
of the metal. Sci. Total Environ. 39, 15–26.
Barghigiani, C., Ristori, T., Lopez Arena, J., 1996. Mercury in
marine sediment from a contaminated area of the northern
Tyrrhenian Sea: <20 m grain-size fraction and total sample
analysis. Sci. Total Environ. 192, 63–73.
Biester, H., Scholz, C., 1996. Determination of mercury binding
forms in contaminated soils. Mercury pyrolysis versus
sequential extractions. Environ. Sci. Technol. 31, 233–239.
Biester, H., Zimmer, H., 1998. Solubility and changes of mercury binding forms in contaminated soils after immobilization treatment. Environ. Sci. Technol. 32, 2755–2762.
Biester, H., Gosar, M., Covelli, S., 2000. Mercury speciation in
sediments aVected by dumped mining residues in the drainage area of the Idrija mercury mine, Slovenia. Environ. Sci.
Technol. 34, 3330–3336.
Biester, H., Müller, G., Schöler, H.F., 2002. Binding and mobility of mercury in solid contaminated by emissions from
chlor-alkali plants. Sci. Total Environ. 284, 191–203.
Bloom, N.S., Preus, E., Katon, J., Hiltner, M., 2003. Selective
extraction to assess the biogeochemically relevant fraction-
ARTICLE IN PRESS
R. Piani et al. / Applied Geochemistry xxx (2005) xxx–xxx
ation of inorganic mercury in sediments and soils. Anal.
Chim. Acta 479, 233–248.
Brambati, A., 1970. Provenienza, trasporto e accumulo dei sedimenti recenti nelle lagune di Marano e di Grado e nei litorali tra
i Wumi Isonzo e Tagliamento. Mem. Soc. Geol. It. 9, 281–329.
Brambati, A., 1997. Metalli pesanti nelle lagune di Marano e
Grado. Piano di studi Wnalizzato all’accertamento di sostanze persistenti nelle Lagune di Marano e Grado ed al loro
risanamento. (RFVG) Direzione Regionale dell’Ambiente,
Servizio dell’Idraulica, Trieste.
Brambati, A., 2001. Coastal sediments and biota as indicators of
Hg contamination in the Marano and Grado Lagoons.
RMZ – Material Geoenviron. 48, 165–171.
Compeau, G.C., Bartha, R., 1985. Sulfate reducing bacteria:
principle methylators of mercury in anoxic estuarine sediments. Appl. Environ. Microbiol. 50, 498–502.
Covelli, S., Faganeli, J., Horvat, M., Brambati, A., 2001. Mercury contamination of coastal sediments as the result of
long-term cinnabar activity (Gulf of Trieste, northern Adriatic Sea). Appl. Geochem. 16, 541–558.
Donazzolo, R., Hieke Merlin, O., Menegazzo Vitturi, L., Orio,
A.A., Pavoni, B., Perin, G., Rabitti, S., 1981. Heavy metal
contamination in surface sediments from the Gulf of Venice,
Italy. Mar. Pollut. Bull. 12, 417–425.
Fabbri, D., Gabbianelli, G., Locatelli, C., Lubrano, D., Trombini, C., Vassura, I., 2001. Distribution of mercury and other
heavy metals in core sediments of the northern Adriatic sea.
Water Air Soil Pollut. 129, 143–153.
Faganeli, J., Planinc, R., Pezdic, J., Smodis, B., Stegnar, P., Ogorelec, B., 1991. Marine geology of the Gulf of Trieste (northern
Adriatic): geochemical aspects. Mar. Geol. 99, 93–108.
Gagnon, C., Pelletier, E., Mucci, A., 1997. Behaviour of anthropogenic mercury in coastal marine sediments. Mar. Chem.
59, 159–176.
Gill, G.A., Bruland, K.W., 1990. Mercury speciation in 1994,
distribution and Xux of particulate mercury in four stratiWed
seepage lakes, surface freshwater system in California and
other areas. Environ. Sci. Technol. 31, 489–495.
Gilmour, C.C., Henry, E.A., Mitchell, R., 1992. Sulfate stimulation of mercury methylation in freshwater sediments. Environ. Sci. Technol. 26, 2281–2287.
Hedges, J.I., Stern, J.H., 1984. Carbon and nitrogen determinations in carbonate-containing solids. Limnol. Oceanog. 29,
657–663.
Hines, M.E., Horvat, M., Faganeli, J., Bonzongo, J.C.J., Barkay,
T., Major, E.B., Scott, K.J., Bailey, E.A., Warwick, J.J., Berry
Lyons, W., 2000. Mercury biogeochemistry in the Idrija
River, Slovenia, from above the Mine into the Gulf of Trieste. Environ. Res. A 83, 129–139.
IRSA-CNR, 1985. Istituto di Ricerca sulle Acque. Metodi
analitici per i fanghi: parametri chimico-Wsici, vol. 3, pp.
1–5.
Kaplan, D.I., Knox, A.S., Myers, J., 2002. Mercury geochemistry in wetland and its implications for in situ remediation. J.
Environ. Engin. 128, 723–732.
Kierfve, B., 1994. Coastal lagoons. In: Kjerfve, B. (Ed.), Coastal
Lagoon Processes. Elsevier, Amsterdam, pp. 1–8.
King, J.K., Harmon, S.M., Fu, T.T., Gladden, J.B., 2002. Mercury removal, methylmercury formation, and sulfate-reducing bacteria proWles in wetland mesocosms. Chemosphere
46, 859–870.
13
King, J.K., Saunders, F.M., Lee, R.F., Jahnke, R.A., 1999. Coupling mercury methylation rates to sulfate reduction rates in
marine sediments. Environ. Toxicol. Chem. 18, 1362–1369.
Lacerda, L.D., Gonçalves, G.O., 2001. Mercury distribution and
speciation in waters of the coastal lagoons of Rio de Janeiro,
SE Brazil. Mar. Chem. 76, 47–58.
Lacerda, L.D., Fernandez, M.A., Tanizaki, K.F., Calazans, C.F.,
1992. Bioavailability of heavy metals in sediments of two
coastal lagoon in Rio de Janeiro, Brazil. Hydrobiol. 228, 65–70.
Loring, D.H., Rantala, R.T.T., 1992. Manual for the geochemical analyses of marine sediments and suspended particulate
matter. Earth-Sci. Rev. 32, 235–283.
Marins, R.V., Lacerda, L.D., Silva Filho, E.V., 1996. Atmospheric mercury deposition over Sepetiba Bay, Brazil. J.
Braz. Chem. Soc. 9, 77–88.
Marocco, R., 1995. Sediment distribution and dispersal in
northern Adriatic lagoons (Marano and Grado paralyc system). Geologia 57, 77–89.
Mason, R.P., Fitzgerald, W.F., Hurley, J., Hanson Jr., A.K.,
Donaghay, P.L., Sieburth, J.M., 1993. Mercury biogeochemical cycling in a stratiWed estuary. Limnol. Oceanog. 38,
1227–1241.
Mastrine, J.A., Bonzongo, J.C.J., Lyons Berry, W., 1999. Mercury concentrations in surface waters from Xuvial systems
draining historical precious mining areas in southeastern
USA. Appl. Geochem. 14, 147–158.
Middleton, R., Grant, A., 1990. Heavy metals in the Humber
estuary: scrobicularia clay as a pre-industrial datum. Proc.
Yorkshire Geol. Soc. 48, 75–80.
Pavoni, B., Donazzolo, R., Marcomini, A., Degobbis, D., Orio,
A.A., 1987. Historical development of the Venice lagoon:
contamination as recorded in radio-dated sediment cores.
Mar. Pollut. Bull. 18, 18–24.
Piani, R., 1992. Il particellato sospeso nella Laguna di Grado e
Marano: aspetti tessiturali e composizionali. Università di
Trieste, Istituto di Geologia e Paleontologia, Tesi di Laurea
in Sedimentologia (unpublished).
Piani, R., Covelli, S., 2000. Contributo antropico di metalli
pesanti e 137Cs nei sedimenti del bacino di Buso (Laguna di
Marano e Grado, Italia settentrionale). Studi Trentini di Scienze Naturali – Acta Geolog. 77, 169–177.
Porcella, D.B., 1994. Mercury in the environment: biogeochemistry. In: Watras, C.J., Huckabee, J.W. (Eds.), Mercury Pollution: Integration and Synthesis. Lewis Publishers, Boca
Raton, FL.
Rae, J.E., Aston, S.R., 1981. Mercury in coastal and estuarine
sediments of the northeastern Irish Sea. Mar. Pollut. Bull.
12, 367–371.
RFVG Regione Autonoma Friuli Venezia Giulia – USL n. 8
Bassa Friulana, 1991. La qualità delle acque della Laguna di
Marano. Regional Technical Report.
RFVG Regione Autonoma Friuli Venezia Giulia, 1983. Progetto del piano di risanamento del bacino idrograWco della
Laguna di Marano e Grado. Rapporto conclusivo. Regional
Technical Report, Trieste.
Salomons, W., 1997. Biogeochemical of contaminated sediments and soils: perspective and future research. J. Geochem. Explor. 62, 37–40.
Sladek, C., Gustin-Sexauer, M., Biester, H., Kim, C., 2002. Assessment of three methods for determining mercury speciation in
mine wastes. Geochem. Explor. Environ. Anal. 4, 369–375.
ARTICLE IN PRESS
14
R. Piani et al. / Applied Geochemistry xxx (2005) xxx–xxx
Smith, J.N., Loring, D.H., 1981. Geochronology of mercury pollution in the sediments of the Saguenay Fjord, Quebec.
Environ. Sci. Technol. 15, 944–951.
Stordal, M.C., Gill, G.A., Wen, L.S., Santschi, P.H., 1996. Mercury
phase speciation in the surface waters of three Texas estuaries:
importance of colloidal forms. Limnol. Oceanog. 41, 52–61.
Wallschläger, D., Desai, M.V.M., Spengler, M., Wilken, R.-D.,
1995. Mercury in contaminated sediments and Xoodplains
of the river Elbe – assessment of major species and mobilization behaviour. In: Wilken, R.-D., Forstner, U., Knochel, A. (Eds.), International Conference Heavy Metals in
the Environment, vol. 2, Hamburg, September 1995, pp.
52–56.
WHO (World Health Organization), 1990. Environmental Criteria 101: Methylmercury. World Health Organization.
Geneva, Switzerland.