Biogeochemistry of Mercury in Three European Estuaries
(Gironde, Scheldt and Rhine)
Tseng, C.M.1,2*; Amouroux, D.2; Donard, O.F.X.2
1Department of Marine
Sciences, University of Connecticut, Avery Point, Groton, CT 06340. (Email:
ctseng@uconnvm.uconn.edu)
2Laboratoire de
Chimie Bio-Inorganique et Environnement, CNRS EP 132, Centre Hélioparc,
Université de Pau et de l’Adour, Pau, 64000, France.
Abstract
The distribution and partitioning of mercury species were
measured in three European estuaries, the Gironde, the Rhine and the Scheldt
during 1996-1998. All measurements of Hg species (e.g., Hg°, Hg2+,
MeHg+) in aqueous and particulate samples were achieved by
cryofocusing / QFAAS (or ICP-MS) after hydride generation or ethylation
derivatisation. The results show that European macrotidal estuarine
environments are important locations for the transformation and transfer of
natural and anthropogenic mercury inputs. Spatial and temporal patterns of
mercury species in the estuaries indicate that the biogeochemistry of mercury
is highly dynamic. The main factors (e.g., water/particle residence time, water
temperature, organic matter, phytoplankton uptake etc.) influencing the
biogeochemistry of mercury species in estuarine environments were examined with
respect to hydrodynamic, chemical and biological forces. Furthermore, the
biogeochemical pathways of MMHg cycling coupled to the interactive processes of
hydrodynamics and biogeochemistry were also proposed. Additionally, the
speciation and partitioning of mercury in the fluid core of a highly turbid
macrotidal Gironde estuary were also determined. The results show that the
speciation of mercury in the fluid mud zone is highly dynamic due to
sedimentation and resuspension cycle. In addition, partitioning of Hg(II) and
MMHg is significantly influenced by the redox cycling of Fe/Mn and of organic
matter.
1. Introduction
Understanding biogeochemical cycling of Hg in estuarine and coastal environments is of prime importance for mankind due to the toxicological impact of methyl-Hg (Craig, 1986). The environmental behavior and biomagnification extent of Hg in natural waters are driven by chemical and biologically-mediated reactions (Fitzgerald and Mason, 1997). Monomethylmercury (MMHg) can be bacterially synthesized in estuarine environments and further incorporated into food webs. The formation of volatile species such as elemental mercury (Hgo) and dimethylmercury (DMHg) (most in the form of Hgo) enhances the relative mobility and recycling of Hg into the atmosphere through gas exchange. However, the information concerning the processes affecting the speciation and partitioning of Hg in estuarine systems such as tidal estuaries is quite limited.
Three European estuaries (Gironde [France], Scheldt [Belgium/Netherlands], and Rhine [Netherlands]), being subjected to macro-tidal variations, were chosen for seasonal field investigation. Their tidal regimes lead to an increased residence time of fresh water and particles during estuarine mixing and the generation of a maximum turbidity zone (MTZ). A characteristic oxygen-depleted zone is mostly associated with the MTZ, within which various anaerobic processes may be stimulated. Such phenomena, especially observed in the Gironde and Scheldt, play an important role in the biogeochemical cycling and partitioning of elements. The aim of this study was primarily to collect environmental data on the seasonal variations of the concentration and distribution of dissolved and particulate Hg species in these three tidal estuaries. The processes affecting the biogeochemistry of Hg during estuarine mixing were examined through the ancillary physico-biogeochemical parameters. The final objective was to investigate the influence of riverine inputs and anthropogenic impact on Hg distribution in estuaries and the biogeochemical pathways for the formation and occurrence of MMHg in these different estuarine systems.
2 Material and Methods
Surface water samples (2-3m depth) were collected in the three investigated sites and a fluid mud core in the Gironde estuary (see in Fig. 1), during BIOGEST cruises from December 1996 to October 1998. A total of nine cruises (3 cruises for each estuary), covering various seasonal and hydrographic conditions, were performed. About 15 to 20 samples along the salinity gradient, from the freshwater upstream of the tidal influence to coastal seawater, were collected during each cruise over four or five consecutive days.
Following sampling, the filtration was undertaken on board once waters were sampled. A Teflon pneumatic filtration system was used to filter the water samples under He gas pressure (2 bars). All operations were then performed within a closed system to avoid potential ship-board contamination. The pretreatments for Hg speciation anlysis in filtered water samples and filters holding the particulate matter followed the procedures described in detail elsewhere (Tseng et al., 1997, 1998, 2000 in press).
2.2. Sample analysis
Mercury species determined in aqueous and particulate samples are identified as follows: (1) dissolved inorganic mercury (Hg(II)D) and methylmercury (MMHgD) (2) particulate inorganic mercury (Hg(II)P) and methylmercury (MMHgP). All measurements of Hg species were achieved by a cryogenic trapping (CT) technique hyphenated to quartz furnace atomic absorption spectrometry (QFAAS) or inductively coupled plasma-mass spectrometry (ICP-MS), after hydride generation (HG) or ethylation (Eth) derivatization. All analytical methods are described in detail elsewhere (Tseng et al., 1997, 1998, 1999, 2000 in press) in terms of performance, interference and validation.
3. Results and discussion
3.1. Hg in the fluid mud system of the Gironde estuary
Hg speciation and partition in a fluid
mud profile, which exists in an oxygen gradient, were investigated in the
highly turbid Gironde estuary (Fig. 2) Total Hg concentration, ranging from 5
to 189 nM, increased directly proportional with SPM concentration (4-174 g l-1)
to a maximum. Particulate Hg averaged 99% of the total Hg. Particulate Hg and
MMHg exhibited a similar trend: the maximum (HgP 1.1-1.5 nmol g-1;
MMHgP 2.6-5.8 pmol g-1) was observed within the upper
layer above the depth of 7 m and the minimum (HgP 0.9-1.0 nmol g-1; MMHgP ~0.6 pmol g-1) at the bottom
layers of the fluid mud. High levels of dissolved Hg species (Hg(II)D,
HgoD, MMHgD, DMHgD) within the
oxic/anoxic interface were found, suggesting intense microbial activity.
In-situ production may account for this increased methylated and elemental Hg
concentration (HgoD 0.3-1.1 pM; MMHgD 0.5-2.4
pM; DMHgD 0.2-1.5 pM) whereas the dissolution from particulate
Hg(II) for higher Hg(II)D (52-68 pM) is through the
biologically-mediated reactions. In the anoxic fluid mud layer, increasing
Hg(II)D (from 30 to 84 pM) and MMHgD (from the values
below the detection limit to 4.6 pM) occurred, corresponding to the decreasing
layers of Hg(II)P and MMHgP in which significant
desorption takes place. Low or absent DMHgD and HgoD
occurred in sections containing less oxygen. The distribution coefficients (log
Kd) for Hg(II) averaged 4.5±0.2. The minimum Kd
at subsurface and bed fluid mud coincided with the maximum in dissolved Mn and
Fe. Log Kd for MMHg (averaged 3.3±0.9)
showed a maximum (4.3-4.7) at the depth 6.4 to 6.8 m and a sharp drop to a
minimum (~ 2.2), corresponding to the anoxic conditions in which mineralization
of POC occurred. Further results show that the partitioning of Hg and MMHg was
clearly influenced by the cycling of Fe/Mn redox and organic matter.
The potential effects of these redox oscillations for Hg species in terms of the sedimentation and resuspension cycle are presented in Fig. 3. This conceptual model briefly illustrates the potential pathway of Hg cycling in the fluid mud system. As the fluid mud settles, Hg bound to particles will be transported from the surface water to the deeper sediment. Sediment particles, once deposited, are rapidly subjected to degradation/mineralization under microbially-mediated reducing conditions due to environmental chemical changes. Hg can also be easily liberated from particulate to dissolved forms through the dissolution of Fe/Mn oxides or carbonates and the mineralization of POM. At this stage, methylation of Hg may occur at the oxic/anoxic boundary in sediment or in the water column. Re-arrangement may occur in the middle layer of fluid mud through the formation of new particle. Intense desorption takes place in bottom sediments, corresponding to denitrification or Mn reduction. Eventually, the fluid mud settling simultaneously enhances Hg remobilization from particulate to dissolved phase and methyl-Hg formation. This is supported by the observation of dissolved Hg species (Hg(II), MMHg) enrichment in fluid mud compared to estuarine surface waters.
In
summary, this
study demonstrates that the Gironde fluid mud system acts as a
biochemically-mediated fluidized reactor for remobilizing and transforming Hg
species. The outcome of the fluid mud setting and erosion can
significantly contribute to the input of Hg-rich solutes, especially MMHg, in
surface waters. Hg supply from fluid mud system may, therefore, have a
significant effect on the concentrations and distributions of Hg species in
estuaries with high turbidity maximums.
3.2 Hg in three estuaries
The summary of temporal average results of dissolved,
particulate and total Hg species in three estuaries is shown Fig. 4. They
presented significant difference among estuarine systems and appeared also to
be seasonally dependent. The mean concentration of Hg(II)P in each
estuary follows the decreasing pattern of the following seasons from spring to
summer to winter for the both Gironde and Scheldt. The reverse was observed for
the Rhine estuary. In contrast, an increase of HgP from spring to
winter was found in the Rhine. The averaged MMHgP in the Gironde and
Scheldt, showed a trend similar to that of Hg(II)P. The
concentrations declined from spring to summer with the lowest values recorded
in the winter. Little variation was observed in the 3 sampling seasons in the
Rhine in terms of concentration and range. The global HgD were
slightly elevated during the warm seasons. The concentrations of HgD
decreased seasonally for three estuaries. Overall, lower levels of reactive HgD
were measured in the Sheldt, relative to the two other estuaries. The MMHgD concentrations were on
average higher in the Gironde (626±575
fM) than in the Scheldt (421±1155 fM) and in the Rhine (<25
fM).
The higher production of MMHg was significantly identified in estuaries with greater turbidity and longer water residence time. In the aspect of seasonal variation, higher particulate mercury (Hg(II), MMHg) concentrations were observed during low river flow in the three estuaries studied. In addition to warmer weather, higher water temperature enhances the biological activities which promote the synthesis of MMHg through bacteria.
Table 1 shows that the highest annual riverine inputs of inorganic mercury calculated occurred among estuaries in the Rhine (7400 mol yr-1), whereas the highest MMHg flux to the estuary occurred in the Gironde (115 mol yr-1). Higher global average inventories of Hg(II) and MMHg were estimated for the Gironde (5100, 165 mole) than for the Scheldt (1650, 13 mole) and the Rhine (50, 0.2 mole). The highest percentage of total mercury inventory in MMHg among three estuaries was about 3.1% in the Gironde. The result shows that MMHg levels in estuaries are significantly related to total Hg concentration stock and residence time of particle/water. Overall average residence time of mercury within estuaries, corresponding to the residence time of particles, is on the order of a year for the Gironde and the Scheldt and of days for the Rhine. Watershed yields calculated to assess the contamination levels for three drainage basins show that all the three studied estuaries are, to some extent, affected by anthropogenic activities. The yields of the Scheldt (100 mmol km-2 y-1), the Rhine (40 mmol km-2 y-1) and the Gironde (30 mmol km-2 y-1) for Hg(II) are indicative of their urban relationships. A higher yield of 1.5 mmol km-2 y-1 in MMHg was observed for the Gironde estuary relative to the Scheldt (0.7 mmol km-2 y-1) and Rhine (0.2 mmol km-2 y-1).
Table 1 Relation with Hg contamination and inventory
in estuaries (Gironde, Scheldt, Rhine)
Fig. 5 illustrates the principle pathway of MMHg cycling in the
macrotidal estuarine environments. The major sedimentological processes (i.e.
sedimentation and resuspension), dominated by hydrodynamic conditions, control
the cycling and transport of particles. The particles from the river or/and
from autochthonous plankton settle during transport to the bottom through
gravitational settling processes.
Resuspension caused by residual landward flow or/and macro-fauna
perturbation results in the release of Hg (i.e. interstitial water and solid
mercury) from sediment into the water column. Overall, hydrodynamic processes
affect the MMHg cycling between two different compartments (i.e. the water
column and sediment) of the estuary. The field observation of sedimental MMHg
along the estuary shows the maximum was also located in the lower zones of the
both Gironde and Scheldt estuaries (Tseng et al., unpublished data). After
taking into account sources from direct formation or from tidal flats, sediment
resuspension appears to be one of the most important sources of MMHg in water
column.
4. Reference
Craig, P.J. (1986),
Organometallic Compounds in Environment, Principles and Reactions. (P.J. Craig,
Editor), Longman, Essex, pp. 1-64.
Fitzgerald,
W.F., Mason, R.P. (1997), Metal Ions in Biological Systems. (A. Sigel, H.
Sigel, Editors), Marcel Dekker, Inc., New York, 34, pp. 53-111.
Tseng, C. M.,
De Diego, A., Martin, F., Donard, O.F.X. (1997), J. Anal. At. Spectrom. 12,
629-635.
Tseng, C. M.,
De Diego, A., Pinaly, H., Amouroux, D., Donard, O.F.X. (1998), J. Anal. At.
Spectrom. 13, 755-764.
Tseng, C. M.,
De Diego, A., Wasserman, J.C., Amouroux, D., Donard, O.F.X. (1999), Chemosph.
39, 1119-1136.
Tseng, C. M., Amouroux, D.,
Brindle, I.D., Donard, O.F.X. (2000), J. Environ. Mon. In press.
Figure 1 Study area: (a) the Gironde and the sampling site of fluid
mud ( ); (b) the Scheldt and (c) the
Rhine.
Figure 2 Concentration and distribution of
dissolved and particulate Hg species in the fluid mud profile during the June
1997 sampling in the Gironde estuary.
