Relocation effects of dredged marine sediments on mercury geochemistry: Venice lagoon, Italy

Estuarine, Coastal and Shelf Science 93 (2011) 7e13

Contents lists available at ScienceDirect

Estuarine, Coastal and Shelf Science journal homepage: www.elsevier.com/locate/ecss

Relocation effects of dredged marine sediments on mercury geochemistry: Venice lagoon, Italy Seunghee Han a, b, *, Joris Gieskes a, Anna Obraztsova c, Dimitri D. Deheyn a, Bradley M. Tebo a, d a

Scripps Institution of Oceanography, University of California San Diego, La Jolla, CA 92093, USA School of Environmental Science and Engineering, Gwangju Institute of Science and Technology, Gwangju 500-712, Korea c Synthetic Genomics, La Jolla, CA 92037, USA d Division of Environmental & Biomolecular Systems, Oregon Health & Science University, Beaverton, OR 97006, USA b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 28 January 2010 Accepted 7 March 2011 Available online 17 March 2011

Understanding the biogeochemical process of Hg is critical in the overall evaluation of the ecological impacts resulting from the reuse of Hg-contaminated dredged sediment. Sediment banks (V1 and V2) were constructed with freshly dredged sediments from a navigational channel in Venice Lagoon, Italy, with the goal of clarifying potential differences in the biogeochemistry of Hg between the reused dredged sediments and those from surrounding sites (SS1 and S2). Toward this purpose, Hg and monomethylmercury (MMHg) concentrations, and Hg methylation rates (MMRs) in the surface 2.5 cm sediments were monitored, along with ammonium, iron, sulfate and sulfide concentrations in the pore waters of banks and surrounding sites from November 2005 to February 2007. Pore water analyses indicate that the bank sediments are characterized by lower levels of sulfate and iron, and by higher levels of ammonium and sulfide compared to the surrounding sediments. With respect to Hg speciation, the fractions of MMHg in total Hg (%MMHg/Hg) and the MMRs were significantly lower in the bank V1 compared to those in the reference site SS1, whereas the %MMHg/Hg and the MMRs were similar between V2 and S2. A negative correlation is found between the logarithm of the particle-water partition coefficient of Hg and the MMR, indicating that the reduced MMRs in V1 are caused by the limited concentrations of dissolved Hg. Organic matter appears to play a key role in the control of MMR via the control of Hg solubility. Ó 2011 Elsevier Ltd. All rights reserved.

Keywords: mercury pore water sediment pollution vertical profiles Venice Lagoon

1. Introduction Dredging and the disposal of dredged materials are important issues concerning coastal area management. Surface sediment is dredged from estuaries and coastal areas to maintain navigation channels and often to remove contaminated materials (Alden and Young, 1982; Levinton et al., 2006). Major concerns have arisen over where to dispose of this dredged material and the ecological impacts of such disposals (Levinton et al., 2006; Burchell et al., 2007). In recent years, dredged materials have been relocated for environmentally beneficial purposes, such as the rejuvenation of intertidal habitats (Burchell et al., 2007). However, this type of relocation has been practiced only on a small scale due to a lack of understanding regarding the ecological impacts that follow the reuse of dredged sediment. Understanding the biogeochemical * Corresponding author. School of Environmental Science and Engineering, Gwangju Institute of Science and Technology, Gwangju 500-712, Korea. E-mail address: [email protected] (S. Han). 0272-7714/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.ecss.2011.03.004

processes involving these contaminants is critical in the overall evaluation of the ecological impacts resulting from the reuse of dredged sediment. To understand the biogeochemical processes of contaminants involved in the reuse of dredged channel sediment in the Venice Lagoon, Italy, the SIOSED (Scripps Institution of Oceanography SEDiment research) program was conducted from March 2005 to November 2007 (Deheyn and Shaffer, 2007). Sediments were dredged from a navigation channel and transplanted onto two shallow sites. At the relocated and surrounding (reference) sites, a multidisciplinary study was carried out, including studies on the geochemistry of trace metals, microbial community, fauna and flora content, and sedimentary ecotoxicology. This type of monitoring program was essential because the sediment in the Venice Lagoon is contaminated with various metals and organic pollutants, and consequently, most of the sediment in the lagoon has been evaluated as potentially hazardous (Frignani et al., 1997; MAV-CVN, 2004). Mercury is one of the most serious pollutants in the Venice Lagoon (Bloom et al., 2004; Han et al., 2007a), posing a potential

8

S. Han et al. / Estuarine, Coastal and Shelf Science 93 (2011) 7e13

threat to public health and marine ecosystems. Most Hg contamination in the Venice Lagoon originates from past occurrences between the 1950s and 1980s, especially from the chlor-alkali discharge located in the petrochemical zone of Porto Marghera (Bloom et al., 2004). Thus, Hg concentrations in the subsurface sediment are often higher than those in the surface sediment (Bloom et al., 2004; Han et al., 2007a). Considering the mass balance calculations using various physical volumes and flows, it has been demonstrated that Hg flux to Venice Lagoon water is dominated mainly by the resuspension of contaminated sediment (Bloom et al., 2004). The current pollution input from rivers, industry, and precipitation has been estimated to be less than 25% of the total Hg flux, highlighting the importance of the resuspension flux (Bloom et al., 2004). We speculated that the reuse of dredged sediment from the navigation channel may increase the concentration of the more toxic form of Hg, monomethylmercury (MMHg), compared to those found in the reference sediment by several reasons. First, the dredged sediment from the navigation channel may contain higher concentrations of MMHg than in the reference sediment because of common characteristics of channel sediments, such as enhanced organic concentrations and microbial activities. Secondly, the solubility of Hg in the dredge sediment would be higher than that in the reference sediment, perhaps due to the oxidation of iron sulfide and consequent release of dissolved Hg (Hammerschmidt and Fitzgerald, 2004; Rothenberg et al., 2008) when Hg-contaminated subsurface sediment is exposed to the surface and air during the dredging activity. This process would lead to higher Hg methylation rates (MMRs), if the dissolved Hg is a limiting factor for the net MMR, as evidenced in several estuarine and coastal sediments (Hammerschmidt and Fitzgerald, 2004; Wolfenden et al., 2005; Hammerschmidt et al., 2008). Finally, a reduction in the dissolved sulfide concentration in pore waters of the dredge sediments, due to the potential destruction of anoxic conditions during the dredging activities, may produce favorable conditions for Hgmethylating organisms to uptake inorganic Hg via increasing neutral Hg-sulfide species (Benoit et al., 1999; Drott et al., 2007; Han et al., 2007b, 2008). Based on these assumptions, experimental banks were built with freshly dredged sediments from a navigational channel in the Venice Lagoon, and Hg speciation and pore water geochemistry were monitored over a period of 18 months. In the present study, we report the Hg and MMHg concentrations, and Hg methylation rates in bank and reference sediments along with dissolved Hg, Fe, ammonium, sulfide, and sulfate concentrations in pore waters.

level even at low tide. The width and length of the banks were 30 m and 10 m, respectively (Deheyn and Shaffer, 2007). The surface sediment in SS1 was more sandy and contained less organic carbon than that in S2, and the dredged sediment from SS0 contained more organic carbon than the reference sediments (Table 1). 2.2. Sample collection Sampling of the sediment commenced shortly after the bank construction. Surface (2.5 cm) sediment was collected using short push cores in November 2005, and February, June, and July 2006, and February 2007. Long piston cores or vibracores (1.5 m or 30 cm) were collected in December 2005, and in May, September, and November 2006. The collected cores were extruded and sectioned (0e2.5, 2.5e5, 5e7.5, 7.5e10, 10e15, and 15e20-cm intervals) within 24 h in a N2-filled glove box in the laboratory of Thetis SpA (Venice, Italy), thus preventing the oxidation of sulfide and dissolved ferrous iron. After the extrusion and sectioning of cores in the glove box, pore waters were extracted by centrifuging under nitrogen conditions at approximately 5000 rpm. After filtering the pore water samples using 0.45-mm pore size polyethersulfone syringe filters under anaerobic conditions, approximately 10e20 cm3 of the filtered pore water sample was acidified for the analysis of dissolved Hg and Fe, and the remainder was used for measurements of dissolved sulfate, sulfide, and ammonium. The remaining sediment slices were stored frozen for analyses of sedimentary Hg and MMHg. Separately, approximately 20 g of the sediment slices were sealed in amber glass vials under N2 saturated conditions and transported to the SIO laboratory in a portable electric cooler (4
C) for Hg methylation experiments. Hg methylation experiments were carried out within 1 week after sampling.

2. Material and methods 2.1. Study area The study area consisted of three sites (Fig. 1): SS0, SS1, and S2. Freshly dredged sediment from approximately the top 1 m layer of site SS0, a previously dredged channel, was transplanted into SS1 and S2 (1.4 m water depth) to create subtidal dredge sediment banks, bank V1 in site SS1 and bank V2 in site S2, between October 25, 2005, and November 16, 2005. The sediment was dredged from SS0 using an excavating crane equipped with a clamshell grab on a barge; the sediment (w220 m3) was re-excavated from the barge and redeposited into the delimited areas that had been designated for making the banks. Initially, the relocated sediment was contained by wood pilings to protect against immediate erosion. The wood pilings were removed in June 2006 after the bank sediment had compacted and stabilized. The heights of both banks (V1 and V2) were reduced from 1 m to 70 cm following the natural compaction and stabilization of the sediment, which was below sea

Fig. 1. Locations of sediment banks (V1 bank in SS1 and V2 bank in S2) and the dredged channel site (SS0) in the Venice lagoon, Italy.

S. Han et al. / Estuarine, Coastal and Shelf Science 93 (2011) 7e13

9

Table 1 Total Hg, MMHg, fraction of MMHg in total Hg, MMR, pore water Hg (dHg), log Kd for Hg, and TOC determined in the top 2.5 cm sediment of banks (V1 and V2) and reference sites of banks (SS1 and S2). Sampling seasons are November and December, 2005, February, May, June, July, September and November 2006, and February 2007. Data are mean  SD of n samples. Hg (ng g1 dw) V1 SS1 V2 S2 n a

748 329 785 491 10

   

246 83 124 56

MMHg (ng g1 dw) 0.37 0.66 0.74 0.86 10

   

0.19 0.55 0.93 0.64

MMHg/Hg (%) 0.052 0.19 0.094 0.17 10

   

0.026 0.15 0.11 0.12

MMR (% hr1) 0.080 0.17 0.13 0.18 7

   

dHg (pM)

0.037 0.085 0.10 0.12

19 63 33 30 9

   

7.3 69 19 26

Log Kda of Hg 5.3 4.7 5.2 5.0 9

   

0.19 0.42 0.28 0.29

TOC (%) 1.1 0.46 1.3 0.78 3

   

0.15 0.10 0.13 0.20

Kd ¼ Cs (mole/Kg)/Cw (mole/dm3), Cs ¼ concentration in particles, Cw ¼ concentration in water.

2.3. Total mercury and monomethylmercury For the analysis of total Hg in the sediments, approximately 1 g of sediment was digested overnight in a TeflonÒ bottle at room temperature with 8 cm3 of 12 N HCl and 2 cm3 of 14 N HNO3, followed by dilution with 500 cm3 of Milli-Q water on the following day (Method 1631). Aliquots of sediment digests (0.1e0.3 cm3) were used to quantify the total Hg by aqueous phase reduction with stannous chloride solution, trapping onto a gold-coated quartz column, thermal desorption, and detection by cold vapor atomic fluorescence spectrometry (CVAFS; Method 1631). Acidified (0.06 N HCl) pore water samples were treated with ultraviolet (UV) irradiation prior to aqueous phase reduction and detection by CVAFS (Gill and Bruland, 1990; Choe et al., 2004). Sediment water content was determined separately to transform the concentrations of Hg in g-wet sediment to g-dry sediment. The analytical precision of the total Hg measurements calculated from the recovery of certified reference material (PACS-2, National Research Council of Canada) averaged 98  13% (mean  SD, n ¼ 26) and that from the recovery of the matrix spike (100% of the sample Hg added) averaged 100  18% (mean  SD, n ¼ 13). MMHg in sediments (0.5e1 g) was extracted into the organic phase following a reaction with 5 cm3 of acidic KBr solution, 1 cm3 of 1 mol dm3 CuSO4 solution, and 10 cm3 of CH2Cl2. An aliquot (2e5 cm3) of CH2Cl2 was back-extracted to the aqueous phase by purging out CH2Cl2 with highepurity nitrogen gas as described by Choe et al. (2004). The extracts were analyzed for MMHg through the use of aqueous phase ethylation, trapping on a Tenax column, gas chromatography separation, thermal decomposition, and detection by CVAFS (EPA Method 1630; Liang et al., 1994). The method detection limit, estimated as three times the standard deviation of the method blank, was 0.01 ng g1. The analytical precision calculated from the recovery of certified reference material (DORM-2, National Research Council of Canada) averaged 95  10% (mean  SD, n ¼ 30) and that from the recovery of matrix spike (100% of sample MMHg added) was 109  7% (mean  SD, n ¼ 8). 2.4. Mercury methylation rate We employed the tracer method using stable Hg isotopes for the measurement of net MMRs (Hammerschmidt and Fitzgerald, 2004). A cold vapor generation system was interfaced to an inductively coupled plasma-mass spectrometer to allow greater sensitivity. In our method, a 200Hg working solution diluted from 200 Hg(NO3)2 with filtered overlying water was added to sediment at concentrations of 2 ng 200Hg g1 of wet sediment under a N2-saturated atmosphere. The sediments were incubated under anoxic conditions for 4 h at appropriate temperatures (field temperature  2
C), after which the sediment samples were frozen at 80
C (Hammerschmidt and Fitzgerald, 2004). The concentrations of MM200Hg in incubated sediment samples were detected using an inductively coupled plasma-mass spectrometer, after the

separation of MMHg from the sediment samples using an organic extraction method: MMHg in lyophilized sediments (w2 g) was extracted into the organic phase following a reaction with 5 cm3 of acidic KBr solution, 1 cm3 of 1 mol dm3 CuSO4 solution, and 10 cm3 of CH2Cl2. After organic extraction, MMHg in CH2Cl2 was backextracted into the aqueous phase using the purging method described by Choe et al. (2004). Because the current analytical method for MMR does not further separate Hg species after the extraction, unlike the GC method previously described, so the correction is necessary. Freeze-dried sediments spiked just prior to extraction were used to determine the amount of inorganic 200Hg extracted with MM200Hg during the organic extraction process. The methylation of added 200Hg was evaluated as the excess concentration of 200Hg versus 198Hg in the sample extracts, as described in Eqs. (1) and (2) (Hintelmann et al., 1995):

MM200 Hgtracer ¼


X

MM198 Hg 

X

MM200 Hg  Rnatural

½ðRtracer  Rnatural Þ  A200 

. ð1Þ

MMRð%=hÞ ¼ MM200 Hgtracer  100=200 Hgadded  hours of incubation

(2)

where MM200Hgtracer is the concentration of MMHg produced from tracer (200Hg) injection, SMM198Hg is the total concentration of MM198Hg (the sum of MM198Hg originally present and MM198Hg produced from tracer injection), SMM200Hg is the total concentration of MM200Hg (the sum of MM200Hg originally present and MM200Hg produced from tracer injection), and A200 is the abundance of 200Hg in the tracer solution. The Rtracer and Rnatural are the ratio of 198Hg/200Hg in tracer solution and natural ratio of 198 Hg/200Hg, respectively. The detection limit of MM200Hg produced is a function of the ambient MMHg concentration (w0.5 ng MMHg g1 dry sediment), the natural abundance of 200Hg, and the precision of the measurement of 198Hg/200Hg. In this study, the detection limit averaged 1.3 pg g1 on a dry weight basis with 1.1% CV (coefficient of variation) in the measurement of 198Hg/200Hg. 2.5. Pore water and sediment geochemistry Dissolved sulfate was measured using a nephelometric method based on the precipitation of barium sulfate by the addition of excess BaCl2 crystals to a sample aliquot (Gieskes et al., 1991). Dissolved sulfide was determined as soon as possible using the method published by Strickland and Parsons (1972), or it was preserved as zinc sulfide through the precipitation of sulfide by addition of zinc acetate to 3e5 cm3 sub-samples. Such analysis involves the formation of a dye (Lauth’s violet) from p-phenylenediamine. The determination of ammonium was based on the method used in Solorzano (1969), originally developed to detect very low NH4þ concentrations in seawater. This method is based on

10

S. Han et al. / Estuarine, Coastal and Shelf Science 93 (2011) 7e13

the diazotization of phenol and the subsequent oxidation of the diazo compound by sodium hypochlorite to yield a blue color. Iron was also measured using a colorimetric method that makes use of an orange complex with orthophenanthroline after any ferric iron has been reduced by the hydroxylamine-HCl and the solution has been buffered by sodium citrate (Gieskes et al., 1991). For the measurement of total organic carbon (TOC) in sediment, approximately 2 g of sediment was lyophilized overnight in acidcleaned glass bottles. Approximately 0.2 g of dry sediment was weighed into silver capsules placed in a microtiter plate, and the accurate weight of each sample was recorded. A small amount of dilute HCl solution was added repeatedly until the release of CO2 was complete. Next, the sediment samples were dried at 60
C, and carefully crimped silver capsules were shipped to the UC Davis Stable Isotope Facility (Davis, CA). At the facility, the carbon mass in dried sediments was measured using an elemental analyzer (PDZ Europa ANCA-GSL).

3. Results 3.1. Pore water geochemistry The pore water geochemistry for ammonium, iron, sulfide, and sulfate in the upper 20 cm of the sediments is presented in Fig. 2. Sediments of Site SSO served as the source of sediment deposited on top the sediments of Sites SS1 and S2 with an average thickness of 70  10 cm. Cores from Site SS0 indicated that the upper 70e100 cm of the sediments at this site are, on average, characterized by high concentrations of ammonium and sulfide, as well as low concentrations of sulfate (Fig. 3) as compared to those found from SS1 and S2. These signals are distinct, because Site SSO is located in a dredged channel just north of Site SS1. The cores taken from the banks V1 and V2 are typically characterized by steep increases in ammonium between 0 cm and 20 cm (Fig. 2). In both V1 and V2, the distributions of dissolved

ammonium (mM) 0

0

5

10

15

20

0

1

2

3

5

10

15

20

0

1

2

3

V2

SS1

V1 depth (cm)

5 0

4

4

5

S2

5 10 Nov 05 Dec 05 Jul 06 Nov 06

15 20

iron (µM)

depth (cm)

0

0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50

5 10 Nov 05 Dec 05 Jul 06 Nov 06

15 20

sulfide (mM)

depth (cm)

0

0

3

6

9

0

1

2

3

0

3

0

6

9

20

30 0

1

2

10

20

3

5 10 Nov 05 Dec 05

15

Jul 06 Nov 06

20

sulfate (mM)

depth (cm)

0

0

10

20

30 0

10

20

30 0

10

30

5 10 15 20

Nov 05 Dec 05 Jul 06 Nov 06

Fig. 2. Sediment pore water profiles of ammonium, iron, sulfide, and sulfate in the banks (V1 and V2) and reference sites (SS1 and S2). Site labels on ammonium apply to all profiles and y-axis labels on V1 apply to all sites.

S. Han et al. / Estuarine, Coastal and Shelf Science 93 (2011) 7e13

ammonium (mM) 0

0

5

10

15

20

250

sulfate (mM) 10

20

sulfide (mM) 30 0

2

4

6

8

d e p t h (c m )

30 60 90 120

Fig. 3. Sediment pore water profiles of ammonium, sulfate, and sulfide in Site SS0 collected in August 2005, before the dredging activity.

ammonium, shortly after completion of the banks, showed high concentrations at the surface (4.6 mM in V1 and 2.6 mM in V2 in November 2005). This is the result of the very recent deposition of the sediment mixture of SSO, which was not entirely homogenous in composition. Below 70 cm, ammonium levels returned toward the expected low values (<0.2 mM) for the upper parts of Sites SS1 and S2 (Supporting Information, S1), upon which the banks were constructed. Sulfate concentration depth profiles generally reflected those of ammonium (Fig. 2). Data for dissolved sulfide in V1 and V2 indicate relatively shallow maxima between 10 and 15 cm, part of which is inherited from the deposition of SSO sediments, but with a potential production through in situ sulfate reduction processes. Sulfides are very reactive components in the pore fluids with possible removal as metal sulfides, the main cause of the iron disappearance below 5 cm of V1 and V2 Cores.

3.2. Mercury speciation in surface sediment The mean total Hg concentrations in the surface 2.5-cm sediments collected from November 2005 to February 2007 are shown in Table 1. Considering the Hg concentrations found in the surface sediment from the Adriatic side of the Lido Inlet (120 ng g1; Bloom et al., 2004), the range of total Hg found in the bank and reference sites is indicative of moderate Hg contamination. The total Hg concentrations in the bank and reference sites are in agreement with literature data determined from the surface sediment of the Venice Lagoon (Bloom et al., 2004; Han et al., 2007a) and the range of total Hg found commonly in urbanized estuaries (e.g., Chesapeake Bay, USA, Mason and Lawrence, 1999; Seine River estuary, France, Mikac et al., 1999). An enhanced total Hg concentration was noted in the surface of the banks as compared to the surface of the surrounding reference sites (t-test, p < 0.05), which agreed with the large amount of Hg in source (SS0) sediments (450e990 ng g1 at the depth of 0e100 cm). The MMHg concentration range in the surface 2.5 cm is in agreement with previous literature values determined from the surface sediment of the Venice Lagoon (Bloom et al., 2004; Han et al., 2007a) and typical MMHg concentrations found in urbanized estuarine sediments (e.g., Patuxent River estuary, USA, Benoit et al., 1998; San Francisco Bay, USA, Choe et al., 2004; Bay of Biscay, France, Stoichev et al., 2004). When comparing the bank and reference sediments, there were no statistically significant differences between V1 and SS1, and between V2 and S2 (t-test, p > 0.05 for each), which was notably different from the total Hg results. The values of %MMHg/Hg in the bank and reference sediments corresponded to the lower range of those in other estuarine and coastal sediments (0.1e0.75%, Fitzgerald et al., 2007). When comparing the bank and reference sediments, the %MMHg/Hg was significantly lower in V1 than SS1 (t-test, p < 0.05) and similar between V2 and S2 (t-test, p > 0.05). The close correlation between

11

%MMHg/Hg and MMR has been shown in surface lake and estuarine sediment when MMRs were measured by short-term incubation (4e48 h) of Hg isotopes, suggesting the dominance of the methylation process over demethylation or net transport processes (Hammerschmidt and Fitzgerald, 2006; Drott et al., 2008; Hammerschmidt et al., 2008). The net MMR range in the upper 2.5 cm of sediment are comparable to those values found in coastal sediments of Long Island Sound, USA (1.4e6.3% day1, Hammerschmidt and Fitzgerald, 2004), and lower than those of freshwater reservoir sediments (2e17% day1; Gray and Hines, 2009) and continental shelf sediment (2e15% day1; Hammerschmidt and Fitzgerald, 2006). Data for the net MMR revealed a pattern similar to that observed for the fraction of MMHg in total Hg, lower in V1 than SS1 but similar between V2 and S2. 4. Discussion 4.1. In situ ammonium production and sulfate reduction in the bank sediment From the data presented in Fig. 2, it appears that the near surface sulfate and ammonium anomalies did not persist for much more than a few months after the construction of the banks. Thereafter, a diffusive exchange process with the overlying waters had been well established already in a short time. However, the continued existence of the concentration maxima in ammonium and minima in sulfate at about 25 cm (Supporting Information, S1) implies that the diffusive exchange with the overlying water is not the only process affecting the concentration profiles. Dissolved chloride concentrations (Supporting Information, S2) showed considerable variability in the near surface sediments in the depth range of 0e20 cm, especially in the reference sites, SS1 and S2 and to a lesser extent in the bank sites, V1 and V2. Particularly in the near surface 10 cm, the chloride concentration changes are largest, presumably as a result of enhanced exchange resulting from tidal pumping and/or bioturbation processes. Though clearly the gradients are not always of a diffusive nature, one can approximate the exchange between overlying waters in terms of a simple mixing model using a turbulent diffusion analogy. Whereas the sediment molecular diffusion coefficient for chloride is probably about 1  105 cm2 s1 (Li and Gregory, 1974; Krom and Berner, 1980), we can estimate the magnitude of the “effective” mixing coefficient by a calculation of the diffusive path length Z ¼ (2Deff t)0.5 over a time period (t) of 1 month. With Deff of 5  105, one calculates a depth (z) of 16 cm and with Deff of (10  2)  105, one obtains a depth range (z) of 23  3 cm. In V1 and V2, 16 cm is appropriate, whereas for SS1 a more appropriate depth estimate >30 cm and for S2 roughly 25 cm. This greater depth at the reference sites is probably related to larger porosities in these sediments than in V1 and V2, which represents the greater consolidation of the SSO sediments (and hence lowers porosities in V1 and V2). Dissolved ammonium has a similar molecular diffusion coefficient as dissolved chloride (Li and Gregory, 1974) and accordingly we estimate the “effective” exchange coefficient (Deff) of NH4þ to be 5  105 cm2 s1 in V1 and V2. If the average concentration gradient of ammonium is roughly 4 mM per 10 cm, it is possible to calculate an exit flux from the sediment by means of the formula: J ¼ eDeff dC/dz, which yields a flux of 15 mmol m2 d1. This flux is probably a maximum flux, but the magnitude of this flux is similar to that reported by Scholten et al. (2000) in the vicinity of Porto Marghera in the Venice Lagoon. With a standing stock of roughly 1500 mmol m2 in the upper 50 cm of the pore water column, this flux would diminish the pore fluid’s ammonium concentrations in

12

S. Han et al. / Estuarine, Coastal and Shelf Science 93 (2011) 7e13

Log Kd of Hg

7 6

2 r = 0.42, p < 0.0001

5 4 3 0.0

0.5

1.0 TOC (%)

1.5

2.0

Fig. 4. The particle-water partition coefficient (Kd) of Hg as a function of organic carbon content at the depths of 0e2.5, 2.5e5, 5e7.5, 7.5e10, 10e15, and 15e20 cm of all sampling sites collected in June 2005, August 2005, and May 2006. Top 2.5 cm sediments are colored black.

about three months. Clearly, this is not the case, and we suggest that substantial production of ammonium resulting from organic matter decomposition via sulfate reduction must continue in the upper 50e70 cm of bank Sites V1 and V2. Probably, as a result of the deposition of the SSO sediments, containing higher levels of organic carbon, continued consumption of organic carbon through sulfate reduction must be important. Thus both sulfide and ammonium production, combined with diffusion into the overlying waters and the underling reference sediments (SS1 and S2) has reached a potential steady state. 4.2. Importance of mercury solubility in the production of monomethylmercury The results shown in Table 1 suggest that the lower MMR found in the bank V1 compared to the reference site SS1 is associated with the solubility of Hg (log Kd). Indeed, a weak to moderate correlation has been reported between the log Kd of Hg and the net MMR of inorganic Hg in several estuarine and coastal sediments (Hammerschmidt and Fitzgerald, 2004, 2006; Hammerschmidt et al., 2008). The existence of a correlation between the log Kd of Hg and MMR suggests that the partitioning of total Hg between the dissolved and particulate phases exerts a significant control over MMHg production over the long term. A direct comparison between the dissolved Hg and the bulk sediment MMHg further supports this idea: the lower levels of sediment MMHg in V1 than in SS1 are related to the lower levels of dissolved Hg in V1 than in SS1 (Table 1). Organic matter appears to play a key role in the control of Hg solubility: a positive linear relationship was found between the TOC and log Kd, particularly for the surface sediment (Fig. 4). In addition to organic matter, solid FeS may play a part in the control of Hg solubility. The diffusive flux calculations described in section 4.1 predict an abundance of solid FeS, both inherited from SS0 and newly produced in situ, in bank sediments, which would promote a decrease in Hg solubility. The apparent steady state of ammonium concentrations in V1 and V2 supports this suggestion: the diffusive efflux of ammonium was well balanced by the in situ production of ammonium, which is mainly processed via microbial sulfate reduction. Sulfide produced by sulfate reduction would cause FeS precipitation, with consequent removal of dissolved Fe from the bank sediment pore waters. 5. Summary In summary, the MMR and the %MMHg/Hg in the top 2.5 cm layers of sediments were lower in the dredged sediment bank V1 when compared to the undisturbed reference site SS1. This was

mainly due to the higher content of organic matter in V1 than in SS1, which caused a decrease in the dissolved Hg concentration of V1. The organic matter contents were more similar between V2 and S2, resulting in comparable levels of MMR and %MMHg/Hg. The present study suggests that relocation of the dredged sediment from the navigation channel causes decreases in MMHg concentrations, as compared to MMHg levels found in the reference sediments, when the TOC levels are significantly higher in the dredged sediment than in the surrounding sediment.

Acknowledgments This research could not have been conducted without the generous help of the research members at Thetis SpA (Venice, Italy): Andrea Berton, Matteo Conchetto, Emiliano Molin, Chiara Castellani, and Fabrizio Perin. We give special thanks to the members of the SIOSED program: Lisa Levin, Douglas Bartlett, Farooq Azam, Osmund Holm-Hansen, Hany Elwany, and Tony Rathburn. This material was produced in the framework of the SIOSED project (to DDD), and supported by Magistrato alle Acque di Venezia, Italy (Venice Water Authority) through the Consorzio Venezia Nuova and Thetis S.p.A. Any opinions, findings, and conclusions or recommendations expressed in this paper are those of the authors and do not necessarily reflect the views of the Magistrato alle Acque di Venezia (Venice Water Authority), Consorzio Venezia Nuova, or Thetis S.p.A.

Appendix. Supplementary material Supplementary data associated with this article can be found in the online version, at doi:10.1016/j.ecss.2011.03.004.

References Alden, R.W., Young, R.J., 1982. Open ocean disposal of materials dredged from a highly industrialized estuary-an evaluation of potential lethal effects. Archives of Environmental Contamination and Toxicology 11, 567e576. Benoit, J.M., Gilmour, C.C., Mason, R.P., Heyes, A., 1999. Sulfide controls on mercury speciation and bioavailability to methylating bacteria in sediment pore waters. Environmental Science & Technology 33, 951e957. Benoit, J.M., Gilmour, C.C., Mason, R.P., Riedel, G.S., Riedel, G.F., 1998. Behavior of mercury in the Patuxent River estuary. Biogeochemistry 40, 249e265. Bloom, N.S., Moretto, L.M., Scopece, P., Ugo, P., 2004. Seasonal cycling of mercury and monomethyl mercury in the Venice Lagoon (Italy). Marine Chemistry 91, 85e99. Burchell, M.R., Skaggs, R.W., Lee, C.R., Broome, S., Chescheir, G.M., Osborne, J., 2007. Substrate organic matter to improve nitrate removal in surface-flow constructed wetlands. Journal of Environmental Quality 36, 194e207. Choe, K.Y., Gill, G.A., Lehman, R.D., Han, S., Heim, W.A., Coale, K.H., 2004. Sedimentwater exchange of total mercury and monomethyl mercury in the San Francisco Bay-Delta. Limnology and Oceanography 49, 1512e1527. Deheyn, D.D., Shaffer, L.R., 2007. Saving Venice: engineering and ecology in the Venice lagoon. Technology in Society 29, 205e213. Drott, A., Lambertsson, L., Bjorn, E., Skyllberg, U., 2007. Importance of dissolved neutral mercury sulfides for methyl mercury production in contaminated sediments. Environmental Science & Technology 41, 2270e2276. Drott, A., Lambertsson, L., Bjorn, E., Skyllberg, U., 2008. Do potential methylation rates reflect accumulated methyl mercury in contaminated sediments? Environmental Science & Technology 42, 153e158. Fitzgerald, W.F., Lamborg, C.H., Hammerschmidt, C.R., 2007. Marine biogeochemical cycling of mercury. Chemical Reviews 107, 641e662. Frignani, M., Bellucci, L.G., Langone, L., Muntau, H., 1997. Metal fluxes to the sediments of the northern Venice Lagoon. Marine Chemistry 58, 275e292. Gieskes, J.M., Gamo, T., Brumsack, H., 1991. Chemical methods for interstitial water analysis aboard JOIDES Resolution. Technical Note 15. Ocean Drilling Program, College Station, TX, USA. http://www-odp.tamu.edu/publications/tnotes/tn15/ f_chem1.htm. Gill, G.A., Bruland, K.W., 1990. Mercury speciation in surface fresh-water systems in California and other areas. Environmental Science & Technology 24, 1392e1400. Gray, J.E., Hines, M.E., 2009. Biogeochemical mercury methylation influenced by reservoir eutrophication, Salmon Falls Creek Reservoir, Idaho, USA. Chemical Geology 258, 157e167.

S. Han et al. / Estuarine, Coastal and Shelf Science 93 (2011) 7e13 Hammerschmidt, C.R., Fitzgerald, W.F., 2004. Geochemical controls on the production and distribution of methylmercury in near-shore marine sediments. Environmental Science & Technology 38, 1487e1495. Hammerschmidt, C.R., Fitzgerald, W.F., 2006. Methylmercury cycling in sediments on the continental shelf of southern New England. Geochimica et Cosmochimica Acta 70, 918e930. Hammerschmidt, C.R., Fitzgerald, W.F., Balcom, P.H., Visscher, P.T., 2008. Organic matter and sulfide inhibit methylmercury production in sediments of New York/New Jersey Harbor. Marine Chemistry 109, 165e182. Han, S., Lehman, R.D., Choe, K.Y., Gill, G.A., 2007b. Chemical and physical speciation of mercury in Offatts Bayou: a seasonally anoxic bayou in Galveston Bay. Limnology and Oceanography 52, 1380e1392. Han, S., Obraztsova, A., Pretto, P., Choe, K.Y., Gieskes, J., Deheyn, D.D., 2007a. Biogeochemical factors affecting mercury methylation in sediments of the Venice Lagoon, Italy. Environmental Toxicology and Chemistry 26, 655e663. Han, S., Obraztsova, A., Pretto, P., Deheyn, D.D., Gieskes, J., Tebo, B.M., 2008. Sulfide and iron control on mercury speciation in anoxic estuarine sediment slurries. Marine Chemistry 111, 214e220. Hintelmann, H., Evans, R.D., Villeneuve, J.Y., 1995. Measurement of mercury methylation in sediments by using enriched stable mercury isotopes combined with methylmercury determination by gas chromatography-inductively coupled plasma mass spectrometry. J Anal Atom Spectrum 10, 619e624. Krom, M.D., Berner, R.A., 1980. The diffusion-coefficients of sulfate, ammonium, and phosphate ions in anoxic marine-sediments. Limnology and Oceanography 25, 327e337. Levinton, J.S., Pochron, S.T., Kane, M.W., 2006. Superfund dredging restoration results in widespread regional reduction in cadmium in blue crabs. Environmental Science & Technology 40, 7597e7601. Li, Y.H., Gregory, S., 1974. Diffusion of ions in sea-water and deep-sea sediments. Geochimica et Cosmochimica Acta 38, 703e714. Liang, L., Horvat, M., Bloom, N.S., 1994. An improved speciation method for mercury by GC/CVAFS after aqueous phase ethylation and room temperature precollection. Talanta 41, 371e379.

13

Magistrato alle Acque di Venezia-Consorzio Venezia Nuova, 2004. Evoluzione temporal dell’inquinamento antropico dei sediment Technical Report A.2.1. Venice, Italy. Mason, R.P., Lawrence, A.L., 1999. Concentration, distribution, and bioavailability of mercury and methylmercury in sediments of Baltimore Harbor and Chesapeake Bay, Maryland. USA. Environmental Toxicology and Chemistry 18, 2438e2447. Method 1630, 2001. Methyl Mercury in Water by Distillation, Aqueous Ethylation, Purge and Trap, and Cold Vapor Atomic Fluorescence Spectrometry. U.S. Environmental Protection Agency, Washington, D.C. Method 1631, 2001. Mercury in Water by Oxidation, Purge and Trap, and Cold Vapor Atomic Fluorescence Spectrometry. U.S. Environmental Protection Agency, Washington, D.C. Revision E. Mikac, N., Niessen, S., Ouddane, B., Wartel, M., 1999. Speciation of mercury in sediments of the Seine estuary (France). Applied Organometallic Chemistry 13, 715e725. Rothenberg, S.E., Ambrose, R.F., Jay, J.A., 2008. Mercury cycling in surface water, pore water and sediments of Mugu Lagoon, CA, USA. Environmental Pollution 154, 32e45. Scholten, M.C.T., Jak, R.G., Pavone, B., Sfriso, A., Philippart, C.J.M., de Heij, H., 2000. Bentic eutrophication studies (BEST). In: Lasserre, E.P., Marzollo, A. (Eds.), The Venice Lagoon Ecosystem-Inputs and Interaction between Land and Sea. The Parthenon Publishing Corporation, pp. 273e288. Solorzano, L., 1969. Determination of ammonia in natural waters by phenol-hypochlorite method. Limnology and Oceanography 14, 799e801. Stoichev, T., Amouroux, D., Wasserman, J.C., Point, D., De Diego, A., Bareille, G., 2004. Dynamics of mercury species in surface sediments of a macrotidal estuarinecoastal system (Adour River, Bay of Biscay). Estuarine Coastal and Shelf Science 59, 511e521. Strickland, J.D.H., Parsons, T.R., 1972. A Practical Handbook of Seawater Analysis. Fisheries Research Board of Canada, Ottawa. Wolfenden, S., Charnock, J.M., Hilton, J., Livens, F.R., Vaughan, D.J., 2005. Sulfide species as a sink for mercury in lake sediments. Environmental Science & Technology 39, 6644e6648.