Geochemistry of trace metals in a fresh water sediment : Field results and diagenetic modeling

TL;DR: Pore water and sediment analyses indicate a shift in trace metal speciation from oxide-bound to sulfide-bound over the upper 20 cm of the sediment, and sensitivity analyses show that increased bioturbation and sulfate availability, expected upon restoration of estuarine conditions in the lake, should increase the sulfide bound fractions of Zn and Ni in the sediments.

About: This article is published in Science of The Total Environment.The article was published on 2007-08-01 and is currently open access. It has received 93 citations till now. The article focuses on the topics: Trace metal & Bioirrigation.

Summary

1. Introduction

• In freshwater sediments, sulfide mineral phases may also immobilize trace metals, despite the lower sulfate concentrations relative to marine systems (Huerta-Diaz et al., 1998; Motelica-Heino et al., 2003).
• Others have applied multi-component RTMs to assess metal sulfide oxidation (Carbonaro et al., 2005; Di Toro et al., 1996) and the controls on arsenic mobility in sediments (Smith and Jaffe, 1998).
• A better understanding of trace metal behavior in sediments is needed to guide management efforts to improve water quality and ecosystem health of the lake.

2. Study site

• In 1970, the Haringvliet estuary was converted to a freshwater lake by building a dam at the outlet to the North Sea (Fig. 1).
• A partial restoration of estuarine conditions is proposed for Haringvliet Lake, beginning in 2008.
• Changes in management of the dam will allow water from the adjacent North Sea to enter the lake at high tide.
• Trace metal contamination has adversely affected benthic communities in sediments of the Rhine–Meuse Delta where low species diversity has been correlated with sediment toxicity including elevated trace metal Netherlands with a box denoting the location of the detail section.
• Concentrations (Reinhold-Dudok van Heel and den Besten, 1999).

3.1. Sample collection

• Sediment and pore water samples were collected in September 2002, and April–May 2003.
• The sampling periods are referred to as late-summer, and spring, respectively.
• Sediment was collected using a cylindrical box corer, with a 31 cm inner diameter, deployed from RV Navicula.
• Subcores were taken with polycarbonate tubes (10 cm i.d.).
• Sub-cores for pore water and solid phase analysis were taken from a single box core and immediately sectioned in a N2 purged glove box on board the ship.

3.2. Pore water analyses

• Sediment sub-samples for pore water collection were placed in polyethylene centrifuge tubes in a N2 purged glove box during core sectioning.
• Sulfide was measured colorimeterically (Cline, 1969) using filtered pore water fixed with NaOH (10 μl 1 M NaOH per ml).
• The pore water concentration was then derived from the estimated diffusive flux through the gel following the established procedure (e.g. Zhang et al., 1995; Naylor et al., 2004).
• The probes were inserted into sediment cores and incubated for 24 h in the temperature controlled shipboard laboratory.
• Following incubation, the agarose gel was removed from each individual compartment and eluted in 1 ml 1MHNO3.

3.3. Solid phase analyses

• Sediment water content and density were determined from the weight loss upon freeze drying, allowing for the determination of sediment porosity.
• Total carbon, total sulfur, and organic carbon (Corg; following carbonate removal with 1 M HCl) were determined on freeze-dried sediment using an elemental analyzer (LECO SC-1440H).
• Analysis ofmetals in all extractants was carried out with ICP-MS unless otherwise noted.
• It is important to note that AVS includes a range of sulfide containing compounds (Rickard and Morse, 2005).
• The reactive pool extraction in the HuertaDiaz and Morse (1990) method (1 M HCl) is less specific, as it also mobilizes reactive Fe(II) phases (Kostka and Luther, 1994).

3.4. Modeling

• Reaction-transport model calculationswere carried out with the Biogeochemical Reaction Network Simulator (BRNS; Aguilera et al., 2005; Jourabchi et al., 2005).
• The discussion of reaction and transport processes in this paper is limited to those that directly involve the trace metals.
• The model describes 1-D sediment profiles at steady-state.
• The ability of thermodynamic modeling to predict the speciation of dissolved metals is limited by the use of pure, end-member solid phases and, the limited knowledge of metal–sulfide stability constants.

4.1. Porewater

• Pore water DOC displayed a gradual increase with depth in late-summer, while a subsurface maximum was observed in spring (Fig. 2).
• Pore water analyses based on the DGT method indicated the presences of free sulfide in the upper 10 cm of sediment.
• The pore water concentrations of Zn, Pb, and Cd were similar for the two sampling times.
• Pore water profiles of Co and Ni resembled those of Mn, for both sampling periods.

4.2. SPM metal content

• Trace metals in the water column of the Haringvliet Lake are mainly associated with the SPM.
• Concentrations of solid-phase Zn, Ni, Pb, and Cd in the upper 2 cm of sediment exhibit the same order of abundance and the same magnitudes as in the lake SPM (Fig. 4).
• Concentrations of Fe and Mn in the SPM varied independently from one another (Fig. 5), as previously observed at other sites in the Rhine–Meuse Delta (Paalman and van der Weijden, 1992).
• The SPM trace metal concentrations varied substantially in the period 2000–2004, with coefficients of variation ranging from 17% for Zn to 55% for Cd (see error bars on Fig. 4).
• The trace metal SPM concentrations displayed negative correlations with SPMorganicmatter content; indicating that organicmatter produced in the lake during algae blooms had a lower trace metal content than the terrestrially derived SPM.

4.3. Sediment solid phase

• The sediment is highly porous, fined grained and organic rich (Table 1).
• The total sediment profiles of Corg, Fe, Mn, Zn, Pb, Co, and Cd displayed little variation with depth, particularly in the upper 10 cm of sediment (Fig. 6).
• As expected, the AVS-SEMconcentrationswere lower than the respective total concentrations (Fig. 6).
• The concentrations of CDB extractable Fe, Ni, and Co declined more sharply with depth than the ascorbate extractable concentrations.

5.1. Pore water profiles

• The build up of alkalinity in the pore waters reflects Corg mineralization (Fig. 2).
• The authors previous work has shown that sulfate reduction is an important mineralization pathway (Canavan et al., 2006), which explains the rapid depletion of pore water SO4 2− and the presence of measurable free sulfide.
• The pore water profiles of Zn, Pb, and Cd show a near-surface enrichment in spring (Fig. 2).
• For Zn and Pb, the reductive extraction results imply that the near-surface pore water enrichments can, in part, be explained by reductive dissolution of reactive Fe and Mn oxide phases close to the sediment–water interface.
• The concentration ratios of Mn to Co and Ni in the pore waters (approximately 2000 and 700, respectively) are much higher than those measured in the reductive extractions (Mn:Co=100–450 and Mn:Ni=50–155).

5.2. Sediment solid phase

• The correlations of sediment Corg, Al, and Fe concentrations with grain size b63 μm (Table 1) indicate a close association of OM, metal oxides and clay minerals (Tessier et al., 1996).
• The concentrations of target values for of the Dutch Soil Protection Although pore water profiles suggest diagenetic remobilization may occur in the sediment (Fig. 2), those processes to not result in a redistribution of the total sediment concentration profiles, with the exceptions of S and possibly Ni (Fig. 6).
• The decreasing concentrations with increasing depth of the ascorbate and CDB extractable Fe pools are consistent with reductive dissolution of Fe(III) in the sediment (Fig. 7).
• The progressive decrease with depth of the ascorbate and CDB extractable concentrations of Mn, Zn, Ni, and Co also indicate release from reducible mineral phases (Fig. 7).

5.3. Diagenetic modeling

• The second fractio to both oxide pools 3 Oxide formation Mn2+ and Fe2+ can oxidatively precipita trace metal with the same oxideQtrace m 4 Bioirrigation ZnS concentrations rresponds with that given in Fig. 10 f Fe-oxides (Zn), Mn-oxides (Ni), or FeS2 (Ni) by a ratio derived from eductive dissolution.
• In a simulation run without ZnS oxidation the flux of dissolved Zn2+ across the SWI changed from an efflux of 24 nmol cm−2 yr−1 to an influx of 9 nmol cm−2 yr−1 into the sediment.
• Ni is estimated at 833 through model fitting of the pore water Ni2+ profile, also known as The ratio of FeS.
• The simulated changes to sediment processes resulting from estuarine restoration are shown to have a greater effect on the solid phase speciation than on metal efflux.

6. Conclusions

• The Haringvliet Lake sediment exhibits elevated concentrations of trace metals (Cd, Co, Ni, Pb, and Zn) derived from riverine suspended particles.
• Results of extractions show declining concentrations of reducible phases and an increase in sulfide species with depth.
• Pore waters are supersaturated with respect to Zn, Pb, Co, and Cd monosulfides, while Ni and Co are found to be associated with pyrite.
• These results illustrate a transition from oxide-bound to sulfide-bound trace metals with depth in the sediment.
• Total metal sediment profiles suggest that little metal release from the sediment is occurring with the possible exception of Ni. Diagenetic model simulations predict a greater mobility of Ni than Zn, as Ni does not form stable metal-sulfides, and is more slowly removed by oxidative precipitation at the sediment surface.

Figures

Fig. 4. The mean concentrations of Zn (◆), Pb (×), Ni (▼), and Cd (■) in suspended particulate matter (SPM) from monthly measurements between 2000–2004 (source: Rijkswaterstaat, www.waterbase.nl) versus total concentrations in the upper 2 cm of sediment (average of 3 samples). The error bars represent one standard deviation of the SPM values. A solid line represents the 1:1 relation; note the log scales.

Fig. 3. Pore water ion activity products (IAP) with respect to end-member metal monosulfides for Fe (a), Co (b), and Ni (c). IAP is defined as {Me2+} {HS−}/{H+} and values are calculated using visual MINTEQ with pore water data from late-summer (□) and spring (△). Vertical lines represent correspondingmetal sulfide solubilities (LogKso) for the reactionMeS(s)+H +=Me2++HS−. Values for FeS (s) determined by Davison et al. (1999; −3.0, solid line), Benning et al. (2000;−3.9, dotted line), andWolthers (Wolthers et al., 2005;−2.1, dash-dot line). The LogKso values for CoS (s) andNiS (s) are obtained from Dyrssen and Kremling (1990; CoS=−7.4; NiS=−5.6). Solutions are super-saturated when the Log IAP value is greater than the Log Kso.

Table 2 Calculated degree of pyritization (DOP), Degree of Sulfidization (DOS), and degree of trace metal pyritization (DTMP) a from sediment collected in late-summer

Table 5 The net rates of ZnS precipitation, Ni adsorption to Fe-sulfides, and release to the overlying water, increased bioturbation intensity (Db0) and deeper biotur

Fig. 1. The sampling location in Haringvliet Lake. The inset map shows the Rhine–Meuse river complex flows into the lake from the east, and the lake

Table 1 Depth weighted average concentrations of organic carbon (Corg), Ca, Al, Fe, with grain size

Fig. 5. Concentrations of Fe (●) and Mn (□) in suspended particulate matter (SPM) over an 8-month period in 2001.

Table 4 Description and values of model parameters that relate to processes affecting trace metals and the depth distributions of the bioirrigation coefficient, α. and bioturbation coefficient, Db

Fig. 10. Schematic representations of the model reaction and transport processes for Zn (a) and Ni (b). The bracketed numbers refer to the description of processes given in Table 3. Process rates are integrated over the upper 20 cm of sediment. Rates and fluxes are given on the arrows in units of nmol cm−2 yr−1; discrepancies in mass balance are a result of rounding.

Fig. 9. (a) Depth distribution of Ni from the pyrite fraction of the Huerta-Diaz and Morse (1990) extraction (○). The solid line represent the modeled concentration of pyrite-associated Ni. The dashed line is the modeled concentration of FeS associated Ni, which is not constrained by extraction data. (b) Pore water concentrations of Ni from late-summer (□) and spring (△). Modeled concentration of Ni2+ with FeS–Ni ratio (see text) set to 588 (heavy line), 833 (thin line), and 1100 (dashed line).

Fig. 6. Depth distributions of sediment concentrations of Fe, S, Corg, Zn, Pb, Cd, Mn, Ni, and, Co, in late-summer. Total concentrations (●) and concentrations from the acid volatile sulfide-simultaneously extractable metals (AVS-SEM; 6 M HCl) (△) are shown.

Fig. 7. Sediment distributions of ascorbate (○) and CDB (×) extractable Fe summer. Model distributions of ΣFe(OH)3 (a), Fe(OH)3–Zn (b), ΣMnO2 (c

Fig. 2. Porewater profiles of DOC, alkalinity, pH, Fe, SO4 2−, sulfide, Zn, Pb, Cd, Mn Ni, and Co for the late-summer (□) and spring (△) sampling times. The sediment water interface (SWI) is indicated as a broken horizontal line. Mean values of field blanks for Zn, Pb, Cd, Mn, Co, and Ni are given at the SWI (●). Model fits for alkalinity, pH, Fe2+, SO4 2−, sulfide, Zn2+, Mn2+, and Ni2+ are plotted as continuous lines in the plots. Pore water sulfide profiles measured with the AgI DGT probe (see Section 3.2), are also shown as continuous lines. The sulfide results from late-summer, spring, and the model are labeled L-S, spring, and RTM respectively. For Mn, the concentrations measured with constrained DET (see Section 3.2) are presented for the late-summer (■) and spring (▲).

Frequently asked questions

Q1. What have the authors contributed in "Geochemistry of trace metals in a fresh water sediment: field results and diagenetic modeling" ?

Canavan et al. this paper measured trace metal concentrations in pore water and sediment of a coastal fresh water lake ( Haringvliet Lake, The Netherlands ).