Deep-Sea Research II 47 (2000) 733}760 The fate of carbon in continental shelf sediments of eastern Canada: a case study A. Mucci!,*, B. Sundby", M. Gehlen!, T. Arakaki!, S. Zhong#, N. Silverberg$,% !Department of Earth and Planetary Sciences, McGill University, 3450 Universite& , Montre& al, Que& ., Canada H3A 2A7 "INRS-Oce& anologie, 310 des Ursulines, Rimouski, Que& ., Canada G5L 3A1 #De& partement de Ge& ologie et Ge& nie Ge& ologique, Universite& Laval, Pavillon Andre& Pouliot, Ste-Foy, Que& ., Canada G1K 7P4 $Maurice Lamontagne Institute, Fisheries and Oceans Canada, P.O. Box 1000, Mont-Joli, Que& ., Canada G5H 3Z4 %Centro Interdisciplinario de Ciencias Marinas, LaPaz, Mexico Received 29 April 1997; received in revised form 23 July 1998; accepted 23 July 1999 Abstract This paper discusses the e!ects of organic carbon oxidation on the dissolution and precipitation of calcium carbonate (aragonite and calcite) in "ne-grained continental shelf and slope sediments, using data obtained during the Canadian Joint Global Ocean Flux Study program in the Gulf of St. Lawrence and on the Scotia shelf. The oxygen-penetration depth in these sediments is on the order of 10}15 mm, indicating that organic carbon is mineralized aerobically within this interval. Below this depth, organic matter degradation proceeds mostly through anoxic mineralization processes. The organic carbon content of these sediments decreases smoothly with depth. At all sites, the bottom water is supersaturated with respect to calcite. However, in most cores, immediately below the sediment}water interface, the acidity produced by the aerobic degradation of organic matter is su$cient to overcome the supersaturation of the overlying waters and induce CaCO dissolution. This is most strongly re#ected by an increase 3 in the porewater calcium concentration near the sediment}water interface. Deeper in the cores, the saturation state of the porewaters increases at depth as a result of alkalinity generation by sulfate reduction and CaCO of precipitates. Unlike organic carbon, the inorganic carbon 3 content of the sediments therefore varies little or even increases with depth because what is lost through dissolution near the sediment}water interface is replaced at depth by precipitation. The carbon precipitated as CaCO in the sulfate-reduction zone originates in part from the organic 3 * Corresponding author. Fax: #1-514-398-4690. E-mail address: alm@eps.mcgill.ca (A. Mucci) 0967-0645/00/$ - see front matter ( 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 9 6 7 - 0 6 4 5 ( 9 9 ) 0 0 1 2 4 - 1 734 A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 carbon, resulting in the preservation of a fraction of the original organic carbon as inorganic carbon. The precipitation of CaCO in the anoxic zone creates an additional source of CO to 3 2 the porewaters. The geochemical signi"cance of this source is discussed. ( 1999 Elsevier Science Ltd. All rights reserved. 1. Introduction The oceanic carbon cycle comprises two principal pathways by which carbon is drawn down from the surface ocean: the reduction of inorganic carbon to organic carbon through photosynthesis, and the precipitation of inorganic carbon as calcium carbonate by pelagic and reef-building organisms. Much of the organic carbon is rapidly converted back to CO in the water column as a result of bacterial oxidation, 2 and only a small proportion of it "nds its way to the sea #oor. A variable proportion of the calcium carbonate reaches the sea #oor depending on the saturation state of the overlying waters. Of the total amount of organic and inorganic biogenic carbon that reaches the sea #oor, only a small fraction is buried and preserved. It is only this carbon, which is actually removed from the ocean, that can contribute to the long-term attenuation of perturbations of the global carbon cycle. The organic and inorganic carbon pathways interact through the pH-dependent equilibrium reactions of the carbonate system: (1) CO #H O H HCO~#H` H CO2~#2H`. 3 3 2 2 For example, if the consumption of CO during photosynthesis is intense, the pH 2 increases, as does the saturation state of the waters which, in turn, may lead to the precipitation of CaCO . Conversely, the production of CO during the oxidation of 3 2 organic carbon can lower the pH and lead to the dissolution of CaCO . The latter 3 reaction can be especially important in sediment porewaters where solute transport is slow. Inorganic carbon is added to these sediments by the sedimentation of particulate matter containing aragonite and calcite, two common biogenic CaCO minerals of 3 di!erent solubilities. Carbon is also added in the form of organic matter. Although the latter is a complex mixture of organic compounds, its composition is most often approximated by the following stoichiometry (CH O) (NH ) H PO (Red"eld 2 106 3 16 3 4 et al., 1963). Sedimentary organic matter can be oxidized by a suite of electron acceptors that react in sequence according to their free energy yield (Froelich et al., 1979). The microbially mediated oxidation by O and nitrate, which takes place in the 2 uppermost layer of the sediment, produces acidity and may lead to the dissolution of CaCO (Froelich et al., 1979; Emerson and Bender, 1981; Archer et al., 1989; Cai et al., 3 1995; Jahnke et al., 1997). Conversely, the oxidation of organic matter by Mn and Fe oxides as well as sulfate, which takes place in the suboxic and anaerobic subsurface layers of the sediment, produces alkalinity and may induce the precipitation of CaCO (Sholkovitz, 1973; Gaillard et al., 1989; Boudreau et al., 1992). By considering 3 how the microbial degradation of organic carbon constrains the chemical equilibria of A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 735 the carbonate system, we can examine the transformations of organic and inorganic carbon during burial and determine the amount of carbon that survives these transformations and enters the geological record. This record provides important information for assessing how marine systems responded to variations in biospheric conditions in the past and to predict how the oceans will respond to future changes. In this paper we present a summary of our current conceptual understanding of the coupling between organic matter degradation and carbonate geochemistry in "negrained organic-rich sediments such as those most frequently encountered on continental margins. This summary is presented using data obtained during the Canadian Joint Global Ocean Flux Study program (CJGOFS). The emphasis is on the fate of inorganic carbon. A quantitative evaluation of organic carbon mineralization reactions by various electron acceptors and the establishment of a carbon budget, using measured (i.e., incubation experiments) and estimated (i.e., porewater concentration gradients) #uxes at the sediment}water interface as well as sediment trap data, are presented in accompanying papers (Silverberg et al., 2000; Romero et al., 2000). We consider primarily the two most quantitatively important organic matter oxidation reactions, namely the microbially mediated oxidations using, respectively, O and 2 sulfate as terminal electron acceptors, and their in#uence on the carbonate chemistry of the porewaters and sediments. 2. Methods Undisturbed sediment cores were collected using a 0.12 m2 Ocean Instruments Mark II box corer in May 1993 (M-93), December 1993 (D-93) and June 1994 (J-94) at three stations in the Gulf of St. Lawrence, one in the Emerald Basin on the Scotia Shelf, and one on the upper continental slope o! the shelf break (Fig. 1). The exact coordinates and selected characteristics of the sampling sites are presented in Table 1. At all these sites, the sediments consist of clay- and silt-sized particles. The sampling sites underlie water with 150}250 lM dissolved oxygen, the lowest values are found in the inner part of the Gulf of St. Lawrence (Savenko! et al., 1996). The oxygenpenetration depth into the sediments, measured with oxygen micro-electrodes, varies from 10 to 15 mm, and the oxygen uptake is between 2 and 5 mmol cm~2 d~1 (Silverberg et al., 2000). Detailed descriptions of the sites and the rational behind their selection are given in Buckley (1991) and Silverberg et al. (2000). The cores were subsampled at close intervals (see Table 2a}e) under nitrogen atmosphere (Edenborn et al., 1986). As each sampling interval was sequentially exposed, the pH was measured by inserting a calibrated combination electrode directly into the sediment. Solid samples were transferred to pre-weighed plastic scintillation vials. These samples were freeze-dried and the water content used for estimating the porosity. The dried samples were then homogenized by grinding in an agate mortar and analyzed for inorganic carbon. Porewaters were extracted within a few hours of core recovery at the in situ temperature using Reeburgh-type squeezers (Reeburgh, 1967) modi"ed to "lter the water through a 0.4 lm Millipore "lter as it passed directly into a 50-cc syringe. 736 A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 Fig. 1. Sampling sites in the Gulf of St. Lawrence and the Scotia Shelf: (1) Anticosti Gyre, 360 m; (2) Jacques Cartier Passage, 245 m; (3) Cabot Strait, 531 m; (B) Emerald Basin, 230 m; (S) Scotia Slope, 830 m. A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 737 Table 1 Location and selected characteristics of sites sampled during CJGOFS Phase I Solution Month/ Year Location B S 1 2 3 B 1 2 3 B S May 1993 May 1993 Dec 1993 Dec 1993 Dec 1993 Dec 1993 Jun 1994 Jun 1994 Jun 1994 Jun 1994 Jun 1994 43350@N 42388@N 49330@N 49340'N 47350@N 43350@N 49330@N 49340@N 47350@N 43350@N 42388@N 62348@W 61375@W 66300@W 62300@W 60305@W 62349@W 66300@W 62300@W 60305@W 62349@W 61375@W Depth (m) ¹ (3C) S C (%) 03' C 232 816 331 262 494 230 360 245 531 230 830 9.40 6.10 5.07 4.78 5.36 9.86 5.18 5.12 5.32 10.24 5.00 34.81 34.83 34.26 33.98 34.61 34.68 34.43 34.39 34.71 35.09 34.91 2.13}2.44 1.23}1.60 1.24}1.83 2.21}2.54 2.38}2.93 2.03}2.44 1.29}1.62 2.10}2.28 2.07}2.43 2.25}2.54 1.06}1.34 0.73}0.87 0.93}1.15 0.43}0.53 0.49}0.64 0.89}1.15 0.73}0.81 0.44}0.57 0.78}0.82 0.80}1.04 0.73}0.87 1.06}1.26 */03' (%) Artifacts due to decompression of the core upon recovery and resulting from CaCO 3 precipitation were minimal since all cores were retrieved from depths of less than 830 m. In addition, given the presence of relatively high concentrations of reaction inhibitors in the porewaters (e.g., DOC and SRP), CaCO precipitation resulting from 3 decompression should be negligible (Mucci, 1986). Nevertheless, decompression and degassing during subsampling may explain some of the erratic porewater measurements obtained at the two deepest sites, Stations 3 and S (Tables 1 and 2c, 2d). The porewater samples were partitioned among a number of plastic and glass vials, and treated according to the type of analyses to be performed. The samples were stored untreated for chlorinity titrations, acidi"ed with a 1% equivalent volume of concentrated HCl for calcium and sulfate determinations, preserved with a Zn-acetate solution for dissolved sul"de analyses, and poisoned with HgCl crystals and stored in indi2 vidual vials without headspace gas for total alkalinity (A ), total dissolved inorganic 5 carbon (&CO ), and dissolved organic carbon (DOC) measurements. Soluble reactive 2 phosphate (SRP) and ammonium ion (&NH ) concentrations were determined on3 board using, respectively, a small-volume #ow-injection molybdenum-blue colorimetric method (Ruzika and Hansen, 1980) and a conductivity method (Hall and Aller, 1992). Dissolved sul"de concentrations were measured using the #ow-injection method of Sakamot-Arnold et al. (1986). The procedures were calibrated using standard solutions. For unknown reasons, the sul"de analysis samples did not maintain their integrity and results were inconsistent. Otherwise, the precision of the various analytical procedures described above is estimated to be better than 5%. The pH electrode was calibrated using three NBS (now NIST)-traceable bu!ers and a synthetic seawater TRIS bu!er solution (Hansson, 1973; Millero, 1986). pH measurements were corrected to the in situ temperature and pressure using the equation found in Millero (1979). Total alkalinity, calcium concentration, and chlorinity were determined by potentiometric titrations using, respectively, standardized dilute HCl (Gieskes and Rogers, 1973), EGTA (Lebel and Poisson, Depth (cm) 738 Table 2 / pH *4 (TRIS) [Ca2`] (mmol kg~1 SW) SO 4 (mmol kg~1 SW) &CO 2 (mmol kg~1 SW) A 5 (meq kg~1) SRPO 4 (lmol l~1) &NH 3 (lmol l~1) C ORG (wt% dw) C INORG (wt% dw) FeS AVS 2 (lmol g~1 DW) (a) Selected geochemical characteristics of the sediments and porewaters from a box core taken at Station 1 (Anticosti Gyre) in June 1994. * 0.90 0.90 0.89 0.88 0.87 0.86 0.85 0.83 0.83 0.83 0.82 0.83 0.83 0.83 0.82 0.83 7.73 7.75 7.88 7.90 7.92 7.89 7.90 7.90 7.91 7.83 7.83 7.85 7.83 7.80 7.77 7.84 7.80 10.15 10.21 10.25 10.34 10.38 10.35 10.36 10.28 10.27 10.26 10.30 10.28 10.16 10.11 9.99 10.00 10.09 27.9 27.6 27.8 27.6 27.4 27.2 27.7 27.7 27.3 27.5 * 27.4 27.1 26.9 26.4 25.9 * 2.279 2.413 2.555 2.536 2.660 2.723 2.703 2.744 2.820 2.916 2.987 3.107 3.380 3.691 4.010 4.335 4.561 2.329 2.530 2.567 2.590 2.613 2.667 2.711 2.806 2.891 3.061 3.190 3.371 3.649 3.938 4.291 4.599 4.855 6.35 8.24 8.15 8.96 10.9 13.3 14.2 12.6 12.4 15.4 19.4 18.7 25.3 33.3 37.7 35.7 40.6 0.00 8.66 2.16 13.5 7.24 12.9 18.2 19.6 34.7 45.1 52.6 70.9 82.7 113 158 189 209 * 1.62 1.61 1.67 1.54 1.56 1.57 1.51 1.49 1.35 1.38 1.35 1.37 1.33 1.34 1.29 1.31 * 0.44 0.52 0.45 0.52 0.52 0.51 0.52 0.53 0.56 0.54 0.56 0.56 0.56 0.56 0.58 0.57 * 0.93 1.14 1.02 1.10 1.65 1.29 1.39 1.65 1.86 3.72 9.66 15.7 20.0 24.3 24.0 24.4 * 1.52 0.48 0.54 0.64 0.65 0.62 0.54 0.83 0.99 1.45 1.81 1.56 3.52 2.58 1.57 3.05 (b) Selected geochemical characteristics of the sediments and porewaters from a box core taken at Station 2 (Jacques Cartier Passage) in June 1994. O.L.W. 0.0}0.5 0.5}1.0 1}2 2}3 3}4 4}5 5}7 7}9 9}11 * 0.86 0.85 0.85 0.84 0.84 0.82 0.79 0.80 0.80 7.71 7.56 7.55 7.59 7.59 7.58 7.63 7.64 7.63 7.65 10.12 10.12 10.11 10.10 10.12 10.10 10.04 10.09 10.07 9.98 27.7 27.7 27.6 27.2 27.6 27.4 * 27.0 26.8 27.0 2.291 2.349 2.382 2.548 2.575 2.581 2.653 2.768 2.816 2.853 2.337 2.360 2.382 2.561 2.580 2.572 2.684 2.792 2.900 3.030 4.01 12.9 12.7 10.8 9.32 11.9 11.1 11.2 12.5 12.5 0.00 4.86 14.3 16.6 21.7 27.0 33.7 41.0 45.8 56.8 * 2.28 2.19 2.23 2.20 2.18 2.19 2.22 * 2.19 * 0.80 0.78 0.78 0.81 0.81 0.80 0.80 0.82 0.81 * 5.63 6.37 9.18 17.8 13.9 25.4 31.9 33.4 36.5 * 1.54 2.30 1.49 2.91 4.03 3.15 3.98 4.70 4.63 A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 O.L.W. 0.0}0.5 0.5}1.0 1}2 2}3 3}4 4}5 5}7 7}9 9}11 11}13 13}15 17}19 21}23 25}27 29}31 33}35 11}13 13}15 17}19 21}23 25}27 29}31 33}35 0.79 0.79 0.79 0.79 0.79 0.79 0.79 7.63 7.65 7.64 7.65 7.66 7.71 7.84 10.00 9.96 10.03 9.87 10.01 9.95 9.98 27.1 * * * 26.5 26.5 * 3.067 3.275 3.435 3.390 3.560 3.691 3.853 3.275 3.459 3.694 3.565 3.821 4.014 4.174 15.4 15.5 20.4 21.1 26.8 26.4 28.5 73.8 96.5 131 154 186 207 258 * 2.19 2.12 2.10 * * * 0.80 0.79 0.80 0.79 0.78 0.78 0.78 5.70 6.62 2.91 1.94 3.06 2.50 3.23 * 2.89 2.46 2.32 2.52 2.55 3.01 4.07 6.64 9.56 15.2 21.3 28.7 36.8 44.4 43.1 49.2 * 1.14 2.53 1.98 4.03 3.86 4.06 2.96 3.48 3.50 2.13 3.08 3.52 2.37 4.71 4.70 6.67 * 2.94 3.22 4.21 5.96 11.2 11.5 * 1.42 1.35 1.60 1.59 1.60 1.55 (c) Selected geochemical characteristics of the sediments and porewaters from a box core taken at Station 3 (Cabot Strait) in June 1994. O.L.W. 0.0}0.5 0.5}1.0 1}2 2}3 3}4 4}5 5}7 7}9 9}11 11}13 13}15 17}19 21}23 25}27 29}31 33}35 * 0.91 0.90 0.89 0.89 0.89 0.88 0.87 0.87 0.87 0.86 0.86 0.85 0.85 0.85 0.85 0.85 7.75 7.68 7.71 7.86 7.87 7.92 7.95 7.92 7.91 7.90 7.91 7.91 7.91 7.91 7.93 7.94 7.96 10.17 10.35 10.31 10.26 10.24 10.24 10.14 10.26 10.31 10.25 10.18 10.16 10.04 9.97 9.97 9.81 9.76 27.9 * 27.8 27.9 27.6 27.8 27.2 26.9 26.8 26.9 26.8 26.2 25.3 25.2 24.9 24.6 24.1 2.336 2.640 2.834 3.128 3.884 3.393 3.568 3.790 4.384 4.395 4.764 5.096 5.976 5.689 6.243 6.647 7.284 2.346 2.680 2.897 3.358 4.018 3.597 3.759 4.098 4.715 4.669 4.969 5.542 6.348 6.107 6.746 7.196 7.864 2.43 8.11 11.0 13.4 16.2 16.6 19.7 33.7 38.1 41.8 40.7 46.5 51.4 58.5 61.6 57.4 60.0 0.00 19.4 36.8 59.9 91.4 97.6 111 128 192 178 181 233 298 331 390 462 500 * 2.43 2.32 2.36 2.31 2.33 2.33 2.25 2.32 2.26 2.25 2.22 2.20 2.14 2.14 2.07 2.11 * 0.80 0.82 0.81 0.86 0.84 0.86 0.87 0.86 0.89 0.91 0.94 0.98 1.00 1.01 1.03 1.04 A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 53.1 45.3 44.5 50.6 60.9 63.9 71.2 (d) Selected geochemical characteristics of the sediments and porewaters from a box core taken at Station B (Emerald Basin) in June 1994. * 0.89 0.87 0.85 0.84 0.83 0.82 7.71 7.59 7.54 7.44 7.43 7.40 7.41 10.31 10.29 10.28 10.37 10.32 10.31 10.32 28.1 28.2 28.0 28.0 28.1 28.2 28.0 2.297 2.322 2.345 2.499 2.486 2.598 2.675 2.333 2.298 2.332 2.483 2.543 2.661 2.740 2.07 5.21 5.21 6.30 9.51 10.2 9.91 0.00 9.65 15.9 17.3 19.0 23.4 31.6 * 2.44 2.44 * 2.42 2.37 2.31 * 0.74 0.73 0.73 0.75 0.76 0.76 739 O.L.W. 0.0}0.5 0.5}1.0 1}2 2}3 3}4 4}5 740 Table 2 (continued) / pH *4 (TRIS) [Ca2`] (mmol kg~1 SW) SO 4 (mmol kg~1 SW) &CO 2 (mmol kg~1 SW) A 5 (meq kg~1) SRPO 4 (lmol l~1) 5}7 7}9 9}11 11}13 13}15 17}19 21}23 25}27 29}31 33}35 0.81 0.81 0.80 0.80 0.80 0.79 0.79 0.79 0.77 0.76 7.41 7.41 7.42 7.45 7.44 7.45 7.45 7.45 7.51 7.48 10.42 10.37 10.35 10.41 10.38 10.41 10.38 10.29 10.33 10.38 27.9 28.0 27.8 27.9 27.9 27.7 27.8 28.0 27.9 27.6 2.795 2.826 2.831 2.866 3.084 3.156 3.089 3.095 3.051 3.274 2.848 2.863 2.882 2.998 3.172 3.312 3.266 3.172 3.217 3.395 10.7 10.2 10.2 11.7 11.5 20.4 25.1 18.8 21.4 26.9 &NH 3 (lmol l~1) 29.2 29.9 38.7 37.4 45.7 48.8 54.8 49.6 47.9 60.5 C ORG (wt% dw) C INORG (wt% dw) FeS AVS 2 (lmol g~1 DW) 2.34 * 2.34 * * 2.35 * * 2.25 * 0.78 0.78 0.79 0.79 0.79 0.87 0.79 0.78 0.78 0.79 13.1 13.3 16.1 23.9 20.1 24.2 33.0 44.3 53.4 53.7 1.64 1.90 1.51 1.63 1.69 1.57 1.26 1.52 1.55 1.55 * 1.82 1.43 1.71 1.18 1.00 1.32 1.42 1.63 1.97 1.85 2.64 4.91 4.21 6.38 9.69 12.5 * 0.60 0.45 0.63 0.56 0.72 0.44 0.52 0.39 0.66 0.50 0.52 0.72 0.68 0.83 0.92 0.65 (e) Selected geochemical characteristics of the sediments and porewaters from a box core taken at Station S (Scotia Slope) in June 1994. O.L.W. 0.0}0.5 0.5}1.0 1}2 2}3 3}4 4}5 5}7 7}9 9}11 11}13 13}15 17}19 21}23 25}27 29}31 33}35 * 0.76 0.72 0.72 0.71 0.69 0.68 0.68 0.66 0.67 0.68 0.68 0.67 0.68 0.69 0.69 0.69 7.82 7.69 7.58 7.60 7.62 7.59 7.60 7.60 7.60 7.62 7.63 7.65 7.66 7.62 7.63 7.63 7.63 10.13 10.21 10.20 10.32 10.27 10.26 10.24 10.22 10.18 10.17 10.10 10.03 9.88 9.87 9.84 9.85 9.81 27.7 27.4 27.1 27.6 27.9 27.9 27.8 27.6 27.7 27.6 27.3 27.1 26.8 26.6 26.4 26.1 * 2.244 2.524 2.750 2.781 2.841 2.678 2.984 3.243 2.847 2.914 3.879 4.025 4.077 3.059 3.285 3.490 3.484 2.318 2.537 2.764 2.755 2.800 2.665 3.008 3.282 2.972 3.091 3.915 4.011 4.185 3.140 3.400 3.491 3.589 * 6.89 9.32 6.89 6.89 7.70 9.32 10.1 15.4 17.0 23.5 23.1 15.8 18.6 14.2 17.8 12.2 0.00 21.8 23.1 29.8 56.2 8.54 12.2 21.1 33.7 62.1 66.9 86.8 130 135 165 171 176 * 1.30 1.30 1.34 1.28 1.26 1.23 * 1.15 1.17 * 1.11 1.06 1.06 * * * * 1.18 1.13 1.08 1.06 1.09 1.09 1.08 1.12 1.20 1.24 1.26 1.24 1.24 1.20 1.17 1.20 A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 Depth (cm) A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 741 1976), and AgNO solutions. The reproducibility of these titrations was better than 3 $0.5%. Porewater salinities were calculated from the chlorinity determinations (S"1.80655 Cl; Fofono!, 1985). Sulfate was measured by ion chromatography (Dionex, 1986) after 100-fold dilution in distilled water with a reproducibility of $1%. The total dissolved inorganic carbon concentration, &CO , of the porewaters 2 was determined upon the acidi"cation of a weighed amount of porewater with 2 ml of 2N HCl and coulometric titration of the evolved CO . The porewaters were taken 2 with a syringe from the sample vial and injected directly into the gas stripper in order to avoid gas exchange with the atmosphere. The inorganic carbon content (C ) of INORG the freeze-dried sediments was determined, using the same method, on a weighed amount of solid. The precision of the &CO and C analyses is estimated at better 2 INORG than $2%. Total carbon concentrations (C ) were measured using a Carlo-Erba TOT or Perkin-Elmer CHN analyzer with a reproducibility of better than $4%. Organic carbon (C ) was calculated from the di!erence between the C and C , and ORG TOT INORG thus carries a cumulative uncertainty of about $5%. DOC concentrations were measured following acidi"cation to pH 2 and stripping of the inorganic carbon with CO -free air prior to thermal decomposition on a Pt catalyst (6803C) using a Shimazu 2 TOC-5050 system. The reproducibility of these analyses was better than 2% for concentrations exceeding 1 mg l~1. 2.1. Carbonate speciation and saturation calculations The in situ porewater carbonate ion concentration was calculated from combinations of two of the following three measured parameters: temperature-corrected punch-in pH, porewater &CO and the carbonate alkalinity, using carbonic acid 2 stoichiometric dissociation constants based on the total hydrogen ion concentration scale (Millero, 1979,1995) after appropriate pressure corrections. The carbonate alkalinity, A , was calculated from the total alkalinity, A , after subtracting the # 5 contributions of boric acid, phosphate, and ammonia (Dickson, 1981). The overdetermination of the carbonate system allowed us to verify the self-consistency of the raw data in the oxic and suboxic waters (i.e., pH, A , and &CO ). Because of the # 2 absence of reliable dissolved sul"de measurements, carbonate ion concentration calculations in the sulfate reduction zone were carried out using pH and &CO . The 2 saturation state of the porewaters () ) with respect to calcite (C) or aragonite (A) C 03 A were derived from the ratio of the [Ca2`][CO2~] concentration product to the 3 ) determined by stoichiometric solubility constants of calcite and aragonite (KH C 03 A Mucci (1983) corrected for in situ temperatures and pressures (Millero, 1983). ) "[Ca2`][CO2~]/KH . C 03 A 3 C 03 A (2) 3. Results Particle #uxes and composition were obtained from the analysis of material collected in free-#oating and moored sediment traps set at 150 m depth for periods of 742 A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 days to 6 months (Romero et al., 2000). A limited data set is available from the moored traps because, despite navigational warnings and the installation of protective buoys, many of the traps were lost to "shing activity. Nevertheless, analysis of the material recovered from these traps at Sta. 1 and B indicate that total carbon #uxes vary seasonally with maximum values observed in the early summer: Sta. 1, four-month average (Dec. 1994}March 1995) of 2.9 mmol C m~2 d~1; Sta. B, "ve-month average (May 1993}Oct. 1993) of 0.57 mmol C m~2 d~1. The inorganic carbon content of the sediment trap material was measured on selected splits recovered by both drifting and moored traps. Fluxes were estimated at: Sta. 1, 0.17$0.04 mmol C m~2 d~1 (four-month average); Sta. 2, 0.44 mmol C m~2 d~1 (June-94); Sta. B, 0.02 mmol C m~2 d~1 (April-94). More detailed presentation and discussion of these data appear in Romero et al. (2000). The range of concentrations of organic and inorganic carbon in the sediment cores collected on various occasions during the CJGOFS program is presented in Table 1. Complete vertical pro"les for the cores recovered in the summer of 1994 are presented in Table 2a}e. With few exceptions, organic carbon concentrations decrease smoothly with depth. In general, the C content of the sediments decreases slightly just INORG below the sediment}water interface and then increases with depth. The concentration of C in the sediments is also spatially variable and highest at Stations 3 and S. INORG X-ray di!raction analysis of the sediments con"rmed the presence of calcite, but the aragonite component could not be resolved. Microscopic examination of the trap material, however, revealed the presence of foraminifera (calcite) and pteropods (aragonite) (N. Romero, pers. comm.). Oxygen penetration depths, as determined on-board ship by micro-electrode measurements, vary spatially and seasonally but are typically on the order of 10}15 mm. Results and a detailed discussion of these measurements are presented in Silverberg et al. (2000). The oxygen uptake rates at each station re#ect both the rain rate and the reactivity of the organic matter (Silverberg et al., 2000). Porewater pH generally decreases sharply at or immediately below the sediment}water interface and stabilizes at depth (Table 2a}e). Calcium concentrations often increase slightly below the sediment} water interface and then decrease progressively with depth, whereas the alkalinity increases smoothly throughout the cores. Sulfate concentrations decrease slightly with depth, but irrefutable evidence of sulfate reduction is provided by the steady increase in pyrite and dissolved ammonium concentrations (Table 2a}e). Dissolved ammonium and SRP concentrations increase asymptotically with depth at all stations. Their concentrations at the bottom of the cores re#ect the extent of mineralization of the organic matter buried at each station. They are highest at Cabot Strait (Sta. 3), which is consistent with the higher C content and catabolic rates (SilverORG berg et al., 2000; Boudreau et al., 1998) in these sediments. The porewater DOC concentrations were only measured on samples collected in December 1993, but pro"les were similar in all sampled cores (Fig. 2). The DOC concentrations increase sharply just below the sediment}water interface. They display minimum values a few cm below the interface and increase asymptotically with depth, reaching similar values at the bottom of the cores (19}30 mg C~1 l). A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 743 Fig. 2. Dissolved organic carbon (DOC) concentrations in porewaters recovered from the four sites visited in December 1993. For clarity, the concentration scales are shifted for each station. 4. Discussion The vertical distribution of solid-phase and dissolved sediment components described above is consistent with our current understanding of the system, as described qualitatively in the introduction. The oxidation of organic matter by, respectively, oxygen and sulfate, and the dissolution and precipitation of CaCO , which occur 3 concurrently, will now be dealt with in detail. We elected to focus our discussion on the e!ect of O and sulfate reduction because one or both are the pre-eminent 2 terminal electron acceptors at each station (Boudreau et al., 1998; Table 3). Furthermore, the in#uence of other electron acceptors on the carbonate chemistry of the porewaters is either negligible (e.g., NO~, Froelich et al., 1979) or the stoichiometry of 3 the reactions (e.g., Mn(III, IV) and Fe(III)) is poorly constrained (Boudreau and Can"eld, 1988). A more detailed justi"cation of this approach is provided below. 4.1. Aerobic oxidation of organic carbon Organic matter degradation in the presence of oxygen results in the release of metabolic CO to the porewaters. The stoichiometry of the mineralization reaction is 2 stipulated by the composition of the sedimentary organic matter (Jahnke et al., 1994). The latter, however, is poorly characterized and most commonly quali"ed using C/N ratios or estimated from changes in the composition of the porewaters (Gaillard, 1994). Nevertheless, the organic matter is most commonly assumed to have a Red"eld composition. Given this model composition, the complete microbially mediated oxidation of organic matter using oxygen as the electron acceptor can be represented by (CH O) (NH ) H PO #138 O P 2 106 3 16 3 4 2 106 HCO~#16 NO~#16 H O#HPO2~#124 H`. 4 3 2 3 (3a) 744 A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 Table 3 Percent of carbon oxidized by various oxidants at three of the "ve CJGOFS sites sampled in June 1994! Station O 2 NO~ 3 Mn(III,IV) Fe(III) SO2~ 4 3 B S 15.0 73.7 53.4 3.65 14.2 20.3 15.9 2.39 0.0 1.63 0.65 0.0 52.7 8.84 25.9 !From Boudreau et al. (1998). The reaction is written using the dominant protolytic compounds present in the range of pH normally encountered in seawater and marine porewaters (i.e., 6.8(pH(8.2). This reaction leads to a decrease in the pH as a result of the dissociation of the carbonic, nitric, and phosphoric acids and to a concomitant lowering of the saturation state with respect to carbonate minerals. Note that sulfur has been neglected from the organic matter composition. It may be present in greater proportion than phosphorus (i.e., S : P"1.7 : 1, Red"eld et al., 1963). Its inclusion would require the addition of the following reaction to (3a): (3b) (H S) #3.4 O P1.7 SO2~#3.4 H`. 4 2 1.7 2 The net change in porewater alkalinity from reaction (3a or 3a#3b) is negative (*A "*[HCO~]#*[HPO42~]!*[H`](0). The decrease in saturation state 5 3 (or [CO2~] since [Ca2`] remains constant in the absence of CaCO dissolution or 3 3 precipitation) can also be represented as the titration of carbonate ions in the original solution by carbonic acid formed by the hydration of metabolic CO : 2 CO #H O H H CO , (4a) 2 2 2 3 (4b) H CO #CO2~ H 2 HCO~. 3 3 2 3 If the rate of remineralization is su$ciently fast, the accumulation of metabolic CO 2 may overcome the supersaturation of the overlying waters () (1; Emerson and C Bender, 1981; Archer et al., 1989; Jahnke et al., 1994) and induce the dissolution of CaCO present in the sediments according to the following reactions: 3 CO #H O H H CO , (4a) 2 2 2 3 (5a) H CO H H`#HCO~, 3 2 3 (5b) CaCO #H` H Ca2`#HCO~. 3 3 CaCO #CO #H O H Ca2`#2 HCO~. (5c) 3 3 2 2 Organic carbon oxidation via nitrate reduction Nitrate reduction also contributes to the acidi"cation of the porewaters, but results in a net increase of the alkalinity (*A "*[HCO~]#*[HPO2~]!*[H`]'0). 4 3 5 Given the limited availability of nitrate, the net e!ect is a small reduction of the A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 745 saturation state of the porewaters. This process can be represented by the following reaction: (CH O) (NH ) H PO #94.4 NO~P 3 2 106 3 16 3 4 (6) 106 HCO~#55.2 N #HPO2~#71.2 H O#13.6 H` 4 2 3 2 Given the low *H`/*NO~ of this reaction, the concomitant increase in alkalinity, 3 and the fact that nitrate #uxes at the sediment-water interface are less than 20% of the oxygen #uxes at all stations (Silverberg et al., 2000), nitrate reduction can be assumed to have a negligible in#uence on the preservation of carbonates in these sediments. 4.2. Calcium carbonate dissolution Porewater calcite saturation state pro"les for each core from the three cruises are presented in Fig. 3a}c. For all the sampled cores, the overlying water column is supersaturated with respect to calcite. In most cores, immediately below the sediment}water interface, the saturation state decreases below aragonite saturation and occasionally also below calcite saturation as a result of aerobic organic matter degradation. In some cases (Stations B/May-93, 2/June-94, S/June-94; see Fig. 4), the inorganic carbon content of the sediments decreases slightly below the sediment}water interface, presumably re#ecting calcium carbonate dissolution. However, the solid-phase composition is not very sensitive to these subtle changes. Furthermore, we cannot be certain that the supply of C to the sediments has been constant over the time INORG period represented by the sampling interval. The best indicator of CaCO dissolution 3 is the increase in the calcium concentration of the porewaters below the sediment}water interface (Table 2a}e, Fig. 4). Total and carbonate alkalinity also increase as a result of the dissolution (Eq. (5c); Table 2a}e), but this signal is masked by the di!usion of alkalinity from the sulfate reduction zone. Nevertheless, the dissolution reaction creates a calcium gradient and contributes to the alkalinity gradient between Fig. 3. Calcite saturation state pro"les in the porewaters of sediments recovered from the "ve sampling sites on three di!erent occasions () "1.5) ). C A 746 A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 Fig. 4. Calcite saturation state, porewater calcium and solid sediment CaCO concentrations for the Scotia 3 Slope station core recovered in June 1994. the interstitial and overlying waters. Consequently, there is a net #ux of both Ca2` and alkalinity to the overlying waters. The magnitude of these #uxes is estimated below. 4.3. Other suboxic oxidation reactions Manganese and iron oxide/oxyhydroxide reduction consumes protons, produces alkalinity (*A "*[HCO~]#*[HPO2~]!*[H`]'0), and increases the satura4 3 5 tion state of the porewaters (Froelich et al., 1979). (CH O) (NH ) H PO #236 MnO #36 4H` H 2 106 3 16 3 4 2 236 Mn2`#106 HCO~#8 N #HPO2~#636 H O 4 2 3 2 (7) (CH O) (NH ) H PO #424 Fe(OH) #756 H` H 2 106 3 16 3 4 3 424 Fe2`#106 HCO~#16 NH`#HPO2~#1060 H O. 4 2 4 3 (8) Thus, these reactions can have a strong in#uence on the carbonate chemistry of the porewaters. The role of these alternate electron acceptors on the oxidation of organic matter in shelf and slope sediments, however, is highly variable (Jorgensen, 1982; Bender and Heggie, 1984; Henrichs and Reeburgh, 1987; Can"eld, 1989,1993; Can"eld et al., 1993a,b; Wang and Van Cappellen, 1996). Their role is made even more di$cult to assess given that they can also be involved in the oxidation of reduced metabolites A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 747 (e.g., Aller et al., 1986; Burdige and Nealson, 1986; Luther et al., 1997). Boudreau et al. (1998) used the CANDI model (Boudreau, 1996) to evaluate the percent of carbon oxidized by various oxidants at three of the CJGOFS sites (Stations 3, B and S; Table 3). With the possible exception of Station 3, the role of iron and manganese oxides on organic matter oxidation is negligible. Unfortunately, a similar treatment of Stations 1 and 2 was not possible because of an incomplete dataset. Consequently, as a "rst approximation, we have neglected these reactions in our subsequent treatment of the data. In addition, we do not account explicitly for the oxidation of reduced byproducts such as Fe(II), Mn(II), NH`, and HS~, which di!use to the oxic layer from 4 the anoxic sediments below. These reactions generate protons and contribute to the acidi"cation of porewaters at or near the oxic}suboxic boundary. The stoichiometry of these reactions is not well constrained (i.e., unknown mineralogy of the metal oxides). As indicated below, however, these reactions are considered as contributing to the net oxygen uptake rates. 4.4. Oxidation of organic carbon via sulfate reduction Although we have not measured the rates of sulfate reduction directly in this study, evidence for signi"cant sulfate reduction is provided by the accumulation of dissolved ammonia and authigenic pyrite (Table 2a}e). Detectable ('0.2 lM) and increasing dissolved sul"de concentrations were measured using micro-electrodes (Luther et al., 1997) between 3 and 5 cm depth at Stations 3 and B. Given that the porewater sulfate gradient at Station B is the smallest of the "ve stations investigated, it is probably safe to assume that active sulfate reduction at Stations 1, 2, and S is also initiated within this depth range. Boudreau et al. (1998) estimated the contribution of sulfate reduction to organic matter degradation at three of the "ve stations presented in this study (Table 3). We found a very good linear correlation (r"0.999) between the fraction of organic carbon oxidized by sulfate reduction and the porewater sulfate concentration gradient between the sediment}water interface and a depth of 31 cm (our last porewater sampling depth at most stations) at these stations. Since the sulfate concentration gradient at the two stations for which model estimates are not available fall within the range of the three others, we used the linear correlation to interpolate the contribution of sulfate reduction to organic carbon mineralization at these two stations. The percentages of carbon oxidized by sulfate at Stations 1 and 2 obtained by this procedure are, respectively, 33 and 23%. Similar estimates (respectively, 37 and 27%) are obtained when the correlation is made using our estimated integrated sulfate reduction rates (Table 4). The reaction describing the microbially mediated organic matter oxidation by sulfate reduction can be represented by: (CH O) (NH ) H PO #53 SO2~P 2 106 3 16 3 4 4 106 HCO~#16 NH`#53 HS~#HPO2~#39 H`. 4 4 3 (9) 748 A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 Table 4 Summary of #ux calculations and estimated reaction rates for sites sampled during CJGOFS Phase I Station F(O )! 2 meas. F(&CO )! 2 dC/dX F(Ac)! dC/dX B/May-93 S/May-93 1/Dec-93 2/Dec-93 3/Dec-93 B/Dec-93 1/June-94 2/June-94 3/June-94 B/June-94 S/June-94 ND ND !1.54 !1.81 !1.66 !1.78 !2.78 !4.45 !4.06 !4.05 !3.28 3.73 1.68 2.92 0.73 2.50 4.18 3.03 0.96 5.79 0.57 2.43 3.17 1.26 2.60 0.35 1.54 3.13 3.10 0.40 5.51 2.12 ISRR" 0.05 0.14 0.26 CCDR# *A t CCPR$ *F(&CO )/*[Ca2`] 2 1.57 0.56 0.08/0.08 0.08/0.07 0.23/ 0.19 0.11 0.32 0.08 0.54 1.40 1.50 0.11 2.60 0.10 1.02 0.17/ 0.15/ 0.10/0.12 0.24/0.23 ) CvOW 2.00 2.02 1.24 1.25 1.86 1.83 1.40 1.35 1.45 1.65 1.57 ND-Not determined. !Fluxes are in mmol/m2/day or meq/m2/day. "Integrated Sulfate Reduction Rate (mmol/m2/day) estimated from the [NH`] gradient and the 4 stoichiometry (*&CO /*NH`) of the pore waters. Not estimated when the gradient was below analytical 2 4 uncertainty. #CaCO dissolution rate (mmol/m2/day) estimated as 0.5 times the di!erence between the alkalinity #ux at 3 the sediment}water interface and the #ux at the depth of oxygen penetration. $CaCO precipitation rate (mmol/m2/day) estimated from [F(&CO ) !F(&CO ) ] or the max3 2 (SRR) 2 (.%!4) imum [Ca2`] gradient in the sulfate reduction zone. Note that the latter were only estimated for gradients larger than our analytical uncertainties. Sulfate reduction results in a small decrease in pH but a net production of alkalinity (*A "*[HCO~]#*[HS~]#*[HPO2~]!*[H`]'0). The decrease in pH 4 3 5 would normally promote the dissolution of CaCO according to Eq. (5b), but the 3 buildup of alkalinity increases the saturation state of the porewaters and will, ultimately, lead to the precipitation of CaCO . The production of acid will only a!ect 3 CaCO preservation when very little sulfate reduction occurs (Morse and Mackenzie, 3 1990; Stoessell, 1992). Furthermore, the precipitation of iron sul"de minerals, such as pyrite, during sulfate reduction consumes H` and bu!ers the solution pH (BenYaakov, 1973). The latter can be represented by the transformation of goethite to an iron monosul"de or to pyrite according to 8 Fe(OH) #9 HS~#7 H`P8 FeS#SO2~#20 H O, 4 2 3 8 Fe(OH) #15 HS~#SO2~#17 H`P8 FeS #28 H O. 4 2 2 3 (9a) (9b) 4.5. Porewater saturation state and calcium carbonate precipitation At all sites visited in this study, the saturation state of the porewaters increases at depth as a result of alkalinity production in the sulfate reduction zone (Fig. 3a}c). The A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 749 saturation state either increases to high values, well beyond aragonite saturation (i.e., ) '1.5 or ) '1), or is poised at or near calcite or aragonite saturation C A () "1; ) "1.5) ) by the rapid precipitation of an authigenic CaCO or the C 03 A C A 3 recrystallization of aragonite to calcite. The supersaturated state of the porewaters is determined by a steady state between the precipitation rate of CaCO and the 3 alkalinity production rate from sulfate reduction. Besides the saturation state, the rate of CaCO precipitation in the supersaturated 3 sul"dic sediments depends on the availability of reactive growth surfaces (the calcium carbonate content) and on the abundance of reaction inhibitors such as phosphate and dissolved organic carbon. In the absence of su$cient growth surfaces and/or in the presence of high concentrations of precipitation inhibitors (e.g., SRP and DOC), the saturation state of the porewaters may reach very high values (Berner et al., 1970; Gaillard et al., 1989; Boudreau et al., 1992). For example, at the Jacques Cartier Passage station (2/D-93 & J-94), the high saturation state of the porewaters is explained by a dearth of suitable growth surfaces (C "0.80$0.01% dry INORG weight), whereas at Cabot Strait (3/J-94) the supersaturation is maintained, despite noticeable precipitation, by a high concentration of SRP. Conversely, if the growth rate is fast, the saturation state of the porewaters will be determined by the solubility of the authigenic carbonate. Although the composition of the solid phase is not as sensitive as the porewaters to the dissolution and precipitation of CaCO because of the larger size of the reservoir, 3 these processes are nevertheless re#ected, at most stations (e.g., B/M-93, 1/D-93&J-94, 3/D-93&J-94, B/D-93&J-94, S/J-94), by variations in the solid C content of the INORG sedimentary column. Below an occasionally discernible decrease just below the sediment}water interface, the C content generally increases signi"cantly with INORG depth (e.g., Fig. 4). In some cases, such as Sta. 3 (3/D-93 & J-94), the C content of INORG the sediments at the bottom of the core exceeds its value at the sediment}water interface. Similar observations were reported by Louchouarn et al. (1997) for cores collected almost exactly at Stations 2 and 3. If it can be assumed that the accumulation of CaCO occurred under steady-state conditions over the sampled interval, 3 these data can be interpreted as a net conversion of C to C and storage in the ORG INORG form of CaCO in anoxic sediments. This phenomenon has been documented by 3 Gaillard et al. (1989) in the anoxic sediments of Villefranche Bay (Mediterranean Sea, France) and in studies of carbonate concretions (e.g., Irwin et al., 1977), which show C as a source of carbon. As was the case for the CaCO dissolution in the oxic ORG 3 layer, the precipitation of a carbonate mineral is most strongly re#ected by the depletion of the porewater calcium ion concentrations in the sulfate reduction zone (Table 2a}e, Fig. 4). 4.6. Total dissolved inorganic carbon and alkalinity yuxes at the sediment}water interface As a result of metabolic CO production and CaCO dissolution at and just below 2 3 the sediment}water interface, &CO and alkalinity gradients are established between 2 the interstitial and overlying waters. These gradients sustain a net #ux to the water 750 A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 column. The &CO #uxes were determined from the sum of the #uxes of the 2 individual carbonic acid species: F(&CO )"F(H COH)#F(HCO~)#F(CO2~), 3 3 3 2 2 F(Ac)"F(HCO~)#2F(CO2~), 3 3 across the sediment}water interface according to Fick's "rst law: Flux"!/D4 dC , dx (10a) (10b) (11) where / is the porosity, dC/dx is the concentration gradient across the sediment}water interface, and D is the tortuosity and temperature corrected di!usion 4 coe$cient (sediment di!usion coe$cient). Molecular di!usion coe$cients were derived from self di!usion coe$cients (Hayduk and Laudie, 1974; Li and Gregory, 1974) adjusted to in situ salinity and temperature (Li and Gregory, 1974). Sediment di!usion coe$cients were corrected for tortuosity according to: D 0 D" , 4 1#n(1!/) (12) using n"3 for clay}silt sediments (Iversen and Jorgensen, 1993).The concentration gradient for each carbonate species was calculated from the composition (i.e., pH, A , 5 &CO ) of the bottom waters and porewaters collected from the "rst interval of 2 sampled sediments (i.e., 0}0.5 cm, thus *x"0.25 cm). Other species, such as B(OH)~, 4 NO~ and HPO2~, also contribute to the total alkalinity #ux, but in most cases their 3 4 #uxes across the sediment}water interface are relatively small compared to the carbonate alkalinity #ux. Consequently, in the following discussion, we assumed that the carbonate alkalinity #ux can be represented by the total alkalinity #ux. The #uxes were also estimated using the integrated form of Fick's "rst law as expressed by Reimers et al. (1986): AC D C D B xi d d Flux"+ /D4 CO2 ! /D4 CO2 . (13) dx i dx i~1 x x 0 In this case, the summation was carried out over the whole length of our sampled sediment cores and gradients were calculated between each sampling interval. Eqs. (11) and (13) give nearly identical results; those calculated from Eq. (11) are presented in Table 3 for stations visited during the three cruises. These calculated #uxes are minimum values since biologically-enhanced transport processes are neglected. A comparison of the O uptake rates measured on-board following incubation of the 2 sediment cores and uptake rates estimated from concentrations gradients obtained with micro-electrode pro"ling indicate that molecular di!usion dominates near the sediment}water interface (Silverberg et al., 2000). Decreased macrofaunal activity resulting from decompression and manipulation of the cores (Glud et al., 1994), however, may also explain this observation. &CO and A #uxes estimated from porewater concentration gradients or measured 2 5 from the incubation experiments (Table 4) can be compared with model estimates A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 751 based on the stoichiometry of the organic matter degradation reaction Eq. (3a), and the oxygen uptake rate (Table 3, F(&CO )"!1.58 F(O ) and F(A )"!1.5 F(O ); 2 2 t 2 Emerson et al., 1982; Berelson et al., 1990). This model assumes that the contribution of other electron acceptors (i.e., Fe/Mn oxides, SO2~) to the mineralization of organic 4 matter is re#ected in the O uptake rate through the oxidation of reduced species 2 di!using up to the oxygen penetration zone. With the exception of Station 2, the model estimates (not shown) are within a factor of 2 of the porewater gradient-derived or incubation #uxes. For Station 2, model estimates are about one order of magnitude larger. We calculated, using a nominal sedimentation rate of 0.01 cm/yr (0.08 cm yr~1 at Station 1 based on 210Pb pro"le, Silverberg et al., 2000, Muslow et al., 1998; %9 0.008}0.01 08 cm yr~1 at Station 2 based on palynostratigraphy, de Vernal et al., 1993; 0.008}0.015 cm yr~1 at Station 3 based on 14C AMS dating of the Holocene sediments and palynostratigraphy, de Vernal et al., 1996; Louchouarn et al., 1997), that with the exception of Station 2, only 4}6% of reduced sul"de is preserved as pyrite or AVS in these sediments. In other words, on average, more than 95% of reduced sul"de may be re-oxidized at these stations. In contrast, at Station 2, our calculations indicate that 22% of the reduced sul"de is preserved as pyrite or AVS. 4.7. Estimates of the calcium carbonate dissolution rates In the oxic sediment surface layer, we can assume that alkalinity is only generated in the sediments through the dissolution of CaCO by metabolic CO according to 3 2 Eq. (5c). There is a small decrease in alkalinity due to the oxic degradation of organic matter, but it represents less than 7% of the alkalinity production after equilibrium is re-established with respect to calcite saturation (see Eq. (3b); Froelich et al., 1979, Eq. (10a)). Provided that the steady-state assumption is valid, we should be able to estimate the dissolution rate of CaCO from the alkalinity #ux. Given the 3 stoichiometry of the CaCO precipitation reaction, the dissolution rate would be 3 equal to 0.5]F(A ). We can easily estimate the alkalinity #ux from concentration 5 gradients at the sediment}water interface (see above), but it is not trivial to separate the relative contributions of individual reactions to the total #ux, in particular, the contributions of suboxic and anoxic reactions which occur at depth. We attempted to circumvent this problem by subtracting the alkalinity #ux across the redox boundary (i.e., depth of oxygen penetration, Silverberg et al., 2000) from the #ux at the sediment}water interface and then estimate the CaCO dissolution rate 3 (CCDR) as one-half of the corrected #ux (Table 4). The corrected alkalinity #ux estimates probably carry large uncertainties because gradients are generally small and the #ux values are strongly dependent on where we assign the position of the redox boundary. Similarly, Berelson et al. (1996) estimated the net CaCO dissolution rate 3 in sediments of six California basins by subtracting the net sulfate reduction (SR) from the carbonate alkalinity #ux, F(A ), at the sediment}water interface: # CCDR"0.5[F(A )!2 Net SR]. (14) # Berelson et al. (1996) assumed the net sulfate reduction rate equal to the burial #ux of sulfur or estimated from sulfate or ammonium porewater gradients. The net sulfate 752 A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 Fig. 5. Estimated calcite dissolution rates as a function of the oxygen uptake rates for all stations. Uncertainties on oxygen uptake rates are probably on the order of 10%. Given the method of estimation, the dissolution rates may carry much larger uncertainties but are shown as 20%. The two data points for Station 2 are dismissed from the correlation since porewater data indicate that suboxic reactions (i.e., Fe and Mn oxides reduction) may not be negligible. It is entirely reasonable that the correlation is not linear nor that it does not cross the origin since the supersaturation state of the overlying waters is di!erent at each station and must be overcome before dissolution can occur. reduction rates in the sediments of the CJGOFS stations presented in this paper were estimated from the porewater ammonium gradients since sulfate gradients are weak. The net CaCO dissolution rates and metabolic CO #uxes at the sediment}water 3 2 interface derived using the method of Berelson et al., 1996; not shown) give almost exactly the same results (*CCDR/CCDR"1.5}10%) as obtained above from the di!erential alkalinity #ux calculations. Although the estimated dissolution rates carry a large uncertainty (possibly as much as 50%), with the exception of data from Station 2, there appears to be a positive correlation between CaCO dissolution rates and O uptake rates (Fig. 5). 3 2 The correlation need not be linear because the saturation state of the overlying waters is not the same at all stations (see Table 4). Dissolution of calcium carbonate in the sediments will only be initiated when the supersaturation of the overlying waters, in di!usive contact with the porewaters, is overcome by metabolic CO production. In 2 other words, the higher the supersaturation state of the overlying waters, the greater the rate of O uptake required to render the porewaters undersaturated and induce 2 calcium carbonate dissolution. We discounted the data from Station 2 on the basis that the highest dissolved iron concentrations were measured at this site and the role of suboxic reactions on the porewater chemistry was neglected from our treatment (see sections which address suboxic oxidation reactions above). Until a general diagenetic model such as CANDI (Boudreau, 1996) is applied to these data and the contribution of suboxic and by-product oxidation reactions (e.g., alkalinity production by Mn and Fe oxyhydroxides: Froelich et al., 1979; oxidation of reduced species: Aller and Rude, 1988) can be A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 753 accounted for, we will not be able to derive reasonably accurate estimates of the CaCO dissolution rates from the porewater composition of these 3 sediments. 4.8. Estimates of the calcium carbonate precipitation rates In principle, mass balance calculations should allow us to determine the amount of C converted to and stored as C . Unfortunately, the depth variations of ORG INORG the C content and most of the [Ca2`] porewater gradients are too small to INORG estimate accurately the amount and rate of CaCO precipitation at each station and 3 relate them to the prevailing diagenetic conditions. As an alternative, we have estimated the amount of CaCO precipitated over the sampled interval from the 3 di!erence between the amount of dissolved inorganic carbon (&CO ) generated by 2 sulfate reduction and the #ux of &CO out of the sulfate reduction zone under 2 steady-state conditions. The integrated sulfate reduction rate was estimated from the ammonium gradient (NH`) by applying the stoichiometry as it appears in Eq. (6), as 4 well as the stoichiometry of the particulate organic matter derived from changes in porewater composition (i.e., *&CO /*NH`) and from direct elemental analysis of the 4 2 solid sediments (i.e., C /N) at each station. The CaCO precipitation rate (CCPR) ORG 3 was also estimated from the maximum [Ca2`] gradient at depth for those stations where the gradient was larger than our analytical uncertainty. Results of these calculations are presented in Table 4 and, as a function of integrated sulfate reduction rates (ISRR), in Fig. 6a}c. The observed trend is in good agreement with our understanding of the system: the rate of CaCO precipitation is directly proportional 3 to the alkalinity production rate through sulfate reduction. The signi"cance of authigenic CaCO precipitation in the sulfate reduction zone on 3 the carbon cycle is di$cult to assess. Unlike the in#uence of reef growth through geological time, the global signi"cance of this phenomenon has not yet been addressed. As indicated above, sulfate reduction in anoxic sediments may initially promote CaCO dissolution, but the buildup of carbonate alkalinity will eventually 3 lead to an increase in the saturation state of the porewaters and subsequent CaCO 3 precipitation. The precipitation results in the storage of C as C , but it also ORG INORG contributes to an increase in the PCO of the porewaters and to a stronger H COH 3 2 2 (H CO #but mostly CO (aq)) gradient at depth (reverse of Eq. (5c)). Given the low 2 3 2 temperature, elevated PCO , and bu!ering capacity of the sediment porewaters, the 2 ratio of CO released to CaCO precipitated is nearly equal to 0.9 (Frankignoulle 2 3 et al., 1994). Under the resulting gradient, H COH migrates up towards the oxic 3 2 sediment layer where it may be consumed through the dissolution of CaCO . Vertical 3 pro"les of the PCO in the sediment porewaters at three Gulf of St. Lawrence stations 2 visited in June 1994 (1,2,3/J-94; Fig. 7) demonstrate that: (a) PCO increases in the 2 oxic layer because of catabolic processes, (b) PCO decreases immediately below this 2 zone because of CaCO dissolution under the locally-induced undersaturation, and 3 (c) PCO increases again in the sulfate reduction zone because of C mineralization 2 ORG and CaCO precipitation. The H COH generated by sulfate reduction and CaCO 3 3 3 2 precipitation, however, may not escape to the overlying waters if enough CaCO is 3 754 A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 Fig. 6. Estimates of calcium carbonate precipitation rates as a function of the integrated sulfate reduction rates (ISRR) for various stoichiometries (C : N) of the sedimentary organic matter. (a) Red"eld composition (6.63), (b) *&CO /*NH` of the porewaters corrected for di!erential di!usion (average of "ve stations from 2 4 the May 1993 and June 1994 cruises; C/N"9.7), (c) elemental analysis of the particulate organic matter (average of surface sediments from the "ve stations &15; K. Juniper, pers. comm., Louchouarn et al., 1997). available and a local undersaturation exists in the oxic zone. This scenario may work as a closed loop because H COH formation from CaCO precipitation will not occur 3 3 2 unless enough mineral surfaces are available for precipitation in the sulfate reduction zone and/or the saturation state becomes high enough to nucleate calcite. Fig. 8 A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 755 Fig. 7. CO partial pressure in equilibrium with the porewaters recovered from Stations 1, 2 and 3 in June 2 1994. illustrates the fate of the CO produced by CaCO precipitation in the sulfate 2 3 reduction zone within the porewater CO pool. 2 5. Summary With the exception of high-energy environments, sediments on the continental margins are generally rich in organic matter. This material is supplied both by the high primary productivity associated with these regions and from terrestrial sources entering through coastal areas. Upon deposition and burial, this organic matter fuels a sequence of microbially mediated oxidation reactions that strongly a!ect the chemical composition of the porewaters. The nature and extent of these reactions dictate the fate of calcium carbonate (CaCO ) entering the sediments in the form of 3 exoskeletons of pelagic micro-organisms and shell remains of benthic fauna. We have examined the interactions between organic carbon and inorganic carbon in "ne-grained continental shelf and slope sediments using data obtained during the CJGOFS program on the eastern Canadian continental margins. In these sediments, the oxygen penetration depth is on the order of 10}15 mm, indicating that organic carbon is mineralized aerobically to a depth of 1}2 cm from the sediment}water interface and mineralized anaerobically below this depth. The porewater sulfate concentration does not decrease signi"cantly with depth but evidence of sulfate reduction is provided by the steady increase in pyrite and dissolved ammonium. Porewater pH generally decreases sharply at or immediately below the sediment}water interface and stabilizes at depth. At all sites visited in this study, the bottom water column is supersaturated with respect to calcite. However, in most cores, immediately below the sediment}water interface, the saturation state decreases below aragonite and occasionally below calcite saturation as a result of the acidity produced by the aerobic degradation of organic matter. The dissolution of CaCO is most strongly re#ected by an increase in 3 756 A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 Fig. 8. Schematic representation of the various porewater CO pools, #uxes and reactions in continental 2 margin sediments. porewater calcium concentrations near the sediment}water interface. Total and carbonate alkalinity also increase as a result of the dissolution and contribute to the net #ux of alkalinity and &CO to the overlying waters. Conversely, in the anoxic 2 portion of the sediments, the saturation state of the porewaters increases at depth in response to the production of alkalinity by sulfate reduction. The saturation state of the porewaters increases to high values, well beyond aragonite saturation, or is poised at or near calcite or aragonite saturation. Under these conditions, and in the presence of suitable growth surface, CaCO precipitates from the porewaters. The precipitation 3 is most clearly re#ected by a decrease of the porewater calcium concentration at depth. The importance and relative rates of these processes determine the amount of carbon preserved. Whereas only a fraction of the organic matter that reaches the sediment}water interface on the Canadian continental margins is buried and preserved, the inorganic carbon content of the sediments varies little with depth because what is lost through dissolution near the sediment}water interface is replaced at depth by precipitation upon the conversion of C to C in the sulfate reduction zone. ORG INORG In fact, in organic-rich sediments, the CaCO content of the sediments within the 3 sulfate reduction zone may exceed its concentration at the sediment}water interface. A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 757 The precipitation of CaCO in the sulfate reduction zone creates an additional source 3 of CO to the porewaters but is an insigni"cant source to the overlying waters since 2 most of this CO will be neutralized by CaCO as it migrates up through the oxic, 2 3 CaCO -undersaturated zone. 3 Acknowledgements This study was carried out under the auspices of the Canadian Joint Global Ocean Flux Study and was supported "nancially by the Natural Sciences and Engineering Research Council of Canada. The authors wish to thank Constance Guignard for her technical support in the laboratory. Thanks are also due to the captains and crews of the CSS Alfred Needler, CSS Parizeau, and CSS Hudson for their help at sea. The manuscript greatly bene"ted from the critical comments of the three anonymous journal reviewers, their e!orts were appreciated. References Aller, R.C., Mackin, J.E., Cox, R.T., 1986. Diagenesis of Fe and S in Amazon inner shelf muds: apparent dominance of Fe reduction and implications for the diagenesis of ironstones. Continental Shelf Research 6, 263}289. Aller, R.C., Rude, P.D., 1988. Complete oxidation of solid phase sul"des by manganese and bacteria in anoxic marine sediments. Geochimica et Cosmochimica Acta. 52, 751}765. Archer, D., Emerson, S., Reimers, C., 1989. Dissolution of calcite in deep-sea sediments: pH and O microelectrode results. Geochimica et Cosmochimica Acta. 53, 2831}2845. 2 Bender, M.L., Heggie, D.T., 1984. Fate of organic carbon reaching the deep sea #oor: a status report. Geochimica et Cosmochimica Acta. 48, 977}986. Ben-Yaakov, S., 1973. pH bu!ering of porewater of recent anoxic marine sediments. Limnology and Oceanography 18, 86}94. Berelson, W.M., Hammond, D.E., Cutter, G.A., 1990. In situ measurements of calcium carbonate dissolution rates in deep-sea sediments. Geochimica et Cosmochimica Acta. 54, 3013}3020. Berelson, W.M., McManus, J., Coale, K.H., Johnson, K.S., Kilgore, T., Burdige, D., Pilskaln, C., 1996. Biogenic matter diagenesis on the sea #oor: a comparison between two continental margin transects. Journal of Marine Research 54, 731}762. Berner, R.A., Scott, M.R., Thomlinson, C., 1970. Carbonate alkalinity in the porewaters of anoxic marine sediments. Limnology and Oceanography 15, 544}549. Boudreau, B.P., 1996. A method-of-lines code for carbon and nutrient diagenesis in aquatic sediments. Computers Geoscience 22, 479}496. Boudreau, B.P., Can"eld, D., 1988. A provisional diagenetic model for pH in anoxic porewaters: application to the FOAM site. Journal of Marine Research 46, 429}455. Boudreau, B.P., Can"eld, D., Mucci, A., 1992. Early diagenesis in a marine sapropel, Mangrove Lake, Bermuda. Limnology and Oceanography 37, 1738}1753. Boudreau, B.P., Mucci, A., Sundby, B., Luther, G.W., Silverberg, N., 1998. Comparative diagenesis at three sites on the Canadian continental margin. Journal of Marine Research 56, 1259}1284. Buckley, D.A., 1991. Deposition and diagenetic alteration of sediment in Emerald Basin, the Scotian Shelf. Continental Shelf Research 11, 1099}1122. Burdige, D.J., Nealson, K.H., 1986. Chemical and microbiological studies of sul"de-mediated manganese reduction. Geomicrobiology Journal 4, 361}387. 758 A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 Cai, W.-J., Reimers, C.E., Shaw, T., 1995. Microelectrode studies of organic carbon degradation and calcite dissolution at a California Continental rise site. Geochimica et Cosmochimica Acta 59, 497}511. Can"eld, D.E., 1989. Sulfate reduction and oxic respiration in marine sediments: implications for organic carbon preservation in euxinic sediments. Deep-Sea Research 36, 121}138. Can"eld, D.E., 1993. Organic matter oxidation in marine sediments. In: Wollast, R., Mackenzie, F.T., Chou, L. (Eds.), Interactions of C, N, P and S Biogeochemical Cycles and Global Change, NATO ASI Series. Springer, Berlin, pp. 333}363. Can"eld, D.E., Thamdrup, B., Hansen, J.W., 1993a. The anaerobic degradation of organic matter in Danish coastal sediments: iron reduction, manganese reduction, and sulfate reduction. Geochimica et Cosmochimica Acta. 57, 3867}3883. Can"eld, D.E., Jorgensen, B.B., Fossing, H., Glud, R., Gundersen, J., Ramsing, N.B., Thamdrup, B., Hansen, J.W., Nielsen, L.P., Hall, P.O.J., 1993b. Pathways of organic carbon oxidation in three continental margin sediments. Marine Geology 113, 27}40. de Vernal, A., Guiot, J., Turon, J.-L., 1993. Late and postglacial paleoenvironments of the Gulf of St. Lawrence: marine and terrestrial palynological evidence. GeH ographic Physique et Quaternaire 47, 167}180. de Vernal, A., Hillaire-Marcel, C., Bilodeau, G., 1996. Reduced meltwater out#ow from the Laurentide ice margin during the Younger Dryas. Nature 381, 774}777. Dickson, A.G., 1981. An exact de"nition of total alkalinity and a procedure for the estimation of alkalinity and total inorganic carbon from titration data. Deep-Sea Research 6, 609}623. Dionex, 1986. Method for the determination of trace sulfate in brine. Application Note 53, Dionex Corp. Sunnyvale, CA. Edenborn, H.M., Mucci, A., Belzile, N., Lebel, J., Silverberg, N., Sundby, B., 1986. A glove box for the "ne-scale subsampling of sediment box cores. Sedimentology 33, 147}150. Emerson, S.R., Bender, M., 1981. Carbon #uxes at the sediment-water interface of the deep-sea: calcium carbonate preservation. Journal of Marine Research 39, 139}162. Emerson, S.R., Grundmanis, V., Graham, D., 1982. Carbonate chemsitry in marine porewaters: MANOP sites C and S. Earth and Planetary Science Letters 61, 220}232. Fofono!, N.P., 1985. Physical properties of seawater: A new salinity scale and equation of state for seawater. Journal of Geophysical Research 90, 3332}3342. Frankignoulle, M., Canon, C., Gattuso, J.-P., 1994. Marine calci"cation as a source of carbon dioxide: positive feedback of increasing atmospheric CO . Limnology Oceanography 39, 458}462. 2 Froelich, P.N., Klinkhammer, G.P., Bender, M.L., Luedtke, N.A., Heath, G.R., Cullen, C., Dauphin, P., Hammond, D., Hartmenn, B., Maynard, V., 1979. Early oxidation of organic matter in pelagic sediments of the eastern equatorial Atlantic: suboxic diagenesis. Geochimica et Cosmochimica Acta 43, 1075}1090. Gaillard, J.-F., 1994. Early diagenetic modelling: a critical need for process studies, kinetics rates, and numerical methods. In: Trends in Chemical Geology. Research Trends, Council of Scienti"c Information, Trivandrum, India, pp. 239}252. Gaillard, J.-F., Pauwels, H., Michard, G., 1989. Chemical diagenesis in coastal marine sediments. Oceanologia Acta 12, 175}187. Gieskes, J.M., Rogers, W.C., 1973. Alkalinity determination in interstitial waters of marine sediments. Journal of Sedimentary Petrology 34, 272}277. Glud, R.N., Gundersen, J.K., Jorgensen, B.B., Revsbech, N.P., Schulz, H.D., 1994. Di!usive and total oxygen uptake of deep-sea sediments in the eastern South Atlantic Ocean: in situ and laboratory measurements. Deep Sea Research 41, 1767}1788. Hall, P.O.J., Aller, R.C., 1992. Rapid, small-volume #ow injection analysis of SCO and NH` in marine and 4 2 freshwaters. Limnology and Oceanography 37, 11119}11313. Hansson, I., 1973. A new set of pH-scales and standard bu!ers for seawater. Deep-Sea Research 20, 479}491. Hayduk, W., Laudie, H., 1974. Prediction of di!usion coe$cients for nonelectrolytes in dilute aqueous solutions. American Institute of Chemical Engineering Journal 20, 611}615. A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 759 Henrichs, S.M., Reeburgh, W.S., 1987. Anaerobic mineralization of organic matter: rates and the role of anaerobic processes in the oceanic carbon economy. Geomicrobial Journal 5, 191}237. Irwin, H., Curtis, C., Coleman, M., 1977. Isotopic evidence for source of diagenetic carbonates formed during burial of organic rich sediments. Nature 269, 209}213. Iversen, N., Jorgensen, B.B., 1993. Di!usion coe$cients of sulfate and methane in marine sediments: In#uence of porosity. Geochimica et Cosmochimica Acta 57, 571}578. Jahnke, R.A., Craven, D.B., Gaillard, J.-F., 1994. The in#uence of organic matter diagenesis on CaCO 3 dissolution at the deep-sea #oor. Geochimica et Cosmochimica Acta 58, 2799}2809. Jahnke, R.A., Craven, D.B., McCorkle, D.C., Reimers, C.E., 1997. CaCO dissolution in California 3 continental margin sediments: the in#uence of organic matter mineralization. Geochimica et Cosmochimica Acta 61, 3587}3604. Jorgensen, B.B., 1982. Mineralization of organic matter in the sea bed * the role of sulfate reduction. Nature 296, 643}645. Lebel, J., Poisson, A., 1976. Potentiometric determination of calcium and magnesium in seawater. Marine Chemisrty 4, 321}332. Li, Y.-H., Gregory, S., 1974. Di!usion of ions in sea water and in deep-sea sediments. Geochimica et Cosmochimica Acta. 38, 703}714. Louchouarn, P., Lucotte, M., Duchemin, E., de Vernal, A., 1997. Early diagenetic processes in recent sediments of the Gulf of St. Lawrence: phosphorus, carbon and iron burial rates. Marine Geology 139, 181}200. Luther III, G.W., Sundby, B., Lewis, B.L., Brendel, P.J., Silverberg, N., 1997. Interactions of manganese with the nitrogen cycle: alternative pathways for dinitrogen formation. Geochimica et Cosmochimica Acta 61, 4043}4052. Millero, F.J., 1979. The thermodynamics of the carbonate system in seawater. Geochimica et Cosmochimica Acta 43, 1651}1661. Millero, F.J., 1983. In#uence of pressure on chemical processes in the sea. In: Riley, J.P., Chester, R., (Eds.), Chemical Oceanography, Vol. 8, 2nd Edition. Academic Press, New York, pp. 1}88 (Chapter 43). Millero, F.J., 1986. The pH of estuarine waters. Limnology and Oceanography 31, 839}847. Millero, F.J., 1995. Thermodynamics of the carbon dioxide system in the oceans. Geochimica et Cosmochimica Acta. 59, 661}677. Morse, J.W., Mackenzie, F.T., 1990. Geochemistry of Sedimentary Carbonates. Elsevier, New York, 707 pp. Mucci, A., 1983. The solubility of calcite and aragonite in seawater at various salinities, temperatures and one atmosphere total pressure. American Journal of Science 283, 780}799. Mucci, A., 1986. Growth kinetics and composition of magnesian calcite overgrowths precipitated from seawater: quantitative in#uence orthophosphate ions. Geochimica et Cosmochimica Acta 50, 2255}2265. Mulsow, S., Boudreau, B.P., Smith, J.N., 1998. Bioturbation and porosity gradient. Limnology and Oceanography 43, 1}9. Red"eld, A.C., Ketchum, B.H., Richards, F.A., 1963. The in#uence of organisms on the composition of seawater. In: Hill, N.M. (Ed.), The Sea, Vol. 2. Wiley-Interscience, New York, pp. 26}77. Reeburgh, W.S., 1967. An improved interstitial water sampler. Limnology and Oceanography 12, 163}165. Reimers, C.E., Fischer, K.M., Merewether, R., Smith Jr., K.L., Jahnke, A., 1986. Oxygen micropro"les measured in situ in deep ocean sediments. Nature 320, 741}744. Romero, N., Silverberg, N., Roy, S., Lovejoy, C., 2000. Sediment trap observations from the Gulf of St. Lawrence and the continental margin of eastern Canada. Deep-Sea Research II 47, 545}583. Ruzika, J., Hansen, E.H., 1980. Flow Injection Analysis. Wiley Interscience, New York. Sakamot-Arnold, C.M., Johnson, K.S., Beehler, C.L., 1986. Determination of hydrogen sul"de in seawater using #ow injection analysis and #ow analysis. Limnology and Oceanography 31, 894}900. Savenko!, C., VeH zina, A.F., Packard, T.T., Silverberg, N., Therriault, J.-C., Chen, W., BeH rubeH , C., Mucci, A., Klein, B., MespleH , F., Tremblay, J.-E., Legendre, L., Wesson, J., Ingram, R.G., 1996. Oxygen and nutrient variations in the deep layer of the Gulf of St. Lawrence and their implications for the carbon cycle. Canadian Fisheries and Aquatic Science 53, 2451}2465. 760 A. Mucci et al. / Deep-Sea Research II 47 (2000) 733}760 Sholkovitz, E., 1973. Interstitial water chemistry of the Santa Barbara Basin sediments. Geochimica et Cosmochimica Acta 37, 2043}2073. Silverberg, N., Sundby, B., Mucci, A., Zhong, S., Arakaki, T., Hall, P., LandeH n, A., Tengberg, A., 2000. Remineralization of organic carbon in eastern Canadian continental margin sediments. Deep-Sea Research II 47, 699}731. Stoessell, R.K., 1992. E!ects of sulfate reduction on CaCO dissolution and precipitation in mixing-zone 3 #uids. Journal of Sedimentary Petrology 62, 873}880. Wang, Y., Van Cappellen, P., 1996. A multicomponent reactive transport model of early diagenesis: application to redox cycling in coastal marine sediments. Geochimica et Cosmochimica Acta 60, 2993}3014.