Abstract. The complexity of organic matter (OM) degradation mechanisms represents a significant challenge for developing biogeochemical models to quantify the role of aquatic sediments in the climate system. The common representation of OM by carbohydrates formulated as CH2O in models comes with the assumption that its degradation by fermentation produces equimolar amounts of methane (CH4) and dissolved inorganic carbon (DIC). To test the validity of this assumption, we modeled using reaction-transport equations vertical profiles of the concentration and isotopic composition (δ13C) of CH4 and DIC in the top 25 cm of the sediment column from two lake basins, one whose hypolimnion is perennially oxygenated and one with seasonal anoxia. Our results reveal that methanogenesis only occurs via hydrogenotrophy in both basins. Furthermore, we calculate, from CH4 and DIC production rates associated with methanogenesis, that the fermenting OM has an average carbon oxidation state (COS) below −0.9. Modeling solute porewater profiles reported in the literature for four other seasonally anoxic lake basins also yields negative COS values. Collectively, the mean (±SD) COS value of −1.4 ± 0.3 for all the seasonally anoxic sites is much lower than the value of zero expected from carbohydrates fermentation. We conclude that carbohydrates do not adequately represent the fermenting OM and that the COS should be included in the formulation of OM fermentation in models applied to lake sediments. This study highlights the need to better characterize the labile OM undergoing mineralization to interpret present-day greenhouse gases cycling and predict its alteration under environmental changes.

Vertical profiles of the concentration and isotopic composition (delta C-13) of methane (CH4) and dissolved inorganic carbon (DIC), as well as of ancillary parameters, were obtained in the top 25 cm of a sediment column in a seasonally anoxic basin from an oligotrophic boreal lake. Modeling the profiles of CH4 and DIC concentrations and those of their delta C-13 signatures with reaction-transport equations allowed us to determine the organic matter (OM) degradation rates according to various reactions and to constrain the in situ isotopic fractionation factors and diffusivity coefficients of CH4 and DIC. This exercise reveals inter alia that (i) CH4 production occurs below a depth of 5 cm, with the highest production rate between 5 and 7.5 cm depth, (ii) all CH4 is produced through hydrogenotrophy, and (iii) methanogenesis yields a production rate of CH4 about three times greater than that of DIC. This latter observation indicates either that fermentation of OM is not the exclusive source of H-2 sustaining hydrogenotrophy, or that the commonly assumed model molecule CH2O does not adequately represent the fermenting OM, since its fermentation yields identical rates of CH4 and DIC production. The porewater profiles of Fe and SO42- suggest that some H-2 may be produced during the reoxidation of reduced sulfur by Fe(III), but the rate of H-2 production via this process, if active, would be insignificant in comparison to that required to sustain the estimated rate of hydrogenotrophy. We deduce that the imbalance between CH4 and DIC production rates is rather due to the fermentation of organic substrates that are more reduced than CH2O, i.e., having a negative average carbon oxidation state (COS). From the constraints on reaction rates and on fermentation pathways imposed by the delta C-13 data, we infer that the organic substrate fermenting between depths of 5 and 7.5 cm should have a COS of -1.87. We thus submit that CH4 is produced in the sediments of the seasonally anoxic basin of our boreal lake through hydrogenotrophy coupled to the fermentation of reduced organic substrates that can be represented by a mixture of fatty acids (e.g. C16H32O2; COS of -1.75) and fatty alcohols (e.g., C(16)H(3)4O; COS of -2.00). This study emphasizes the importance of characterizing the sedimentary OM undergoing mineralization in order to improve diagenetic model predictions of CH4 cycling in boreal lakes and of its significance in climate change. (C) 2018 Elsevier Ltd. All rights reserved.

Sediment porewater was analyzed at several sampling dates in two adjacent basins of an oligotrophic boreal lake, one basin perennially oxygenated (Basin A) and the other occasionally anoxic (Basin B). Depth concentration profiles of methane (CH4), dissolved inorganic carbon (DIC), and electron acceptors were modeled with a one-dimensional transport-reaction equation to constrain the depth intervals (zones) where solutes are produced/consumed in the top 10 cm of the sediment column, and to obtain the net reaction rates in each zone. This multicomponent geochemical modeling reveals that CH4 was produced below 4-7 cm depth at lower rates in Basin A (250-800 fmol cm(-2) s(-1)) than in Basin B (1900-6500 fmol cm(-2) s(-1)) and that methanogenesis accounted for 30-64% and 84-100% of the sediment organic matter (OM) mineralization in Basins A and B, respectively. We show that methanogenesis did not always yield equimolar amount of CH4 and DIC, as would be expected from the fermentation of the model molecule CH2O. While similar to 50% of the CH4 produced in Basin A is oxidized in the sediment column, this proportion decreases to similar to 20% in Basin B. Dioxygen is by far the main electron acceptor for CH4 and OM oxidations in both basins. Methanotrophy in the sediment, however, is not limited to the similar to 4-mm thick surface layer in which O-2 diffuses from bottom water but occurs down to 4-7 cm depth where O-2 is transported through bioirrigation. Thermodynamic calculations suggest that, in addition to O-2, Fe oxyhydroxides, and sulfate may serve as oxidants for methanotrophy in that zone. We predict that Basin B sediments release more CH4 than DIC whereas Basin A sediments mainly export DIC. This study highlights that small changes in hypolimnetic O-2 levels may significantly alter the magnitude of OM mineralization pathways and the fate of CH4 in boreal lake sediments.

Sediment cores were collected at the deepest site of twelve headwater lakes from the Province of Quebec, Canada that receive contaminants only from atmospheric deposition, either directly to the lake surface or indirectly from the watershed. Several of the lakes are located within relatively short distance (<40 km) and others at more than 200 km from potential sources of contamination. The sediments were dated and analyzed for In and other elements including Fe, Mn, Al and organic C. Fe-rich authigenic material was collected on Teflon sheets inserted vertically into the sediments at the only study site whose hypolimnion remains perennially oxic. Porewater samples collected at the coring site of four of the lakes were also analyzed for In and other solutes including sulfide, sulfate, Fe, Mn, inorganic and organic C and major ions. The porewater In profiles display concentration gradients at or below the sediment-water interface. Modeling these profiles with a one-dimensional transport-reaction equation assuming steady state allows definition of depth intervals (zones) where In is either released to or removed from porewater and quantification of net In reactions rates in each zone.The position of the In consumption zones, the shape of the vertical profiles of dissolved In, sulfide and iron, as well as thermodynamic calculations of saturation states collectively suggest that In(OH)(3(s)) and In2S3(s) do not precipitate in the sediments and that adsorption of In onto sedimentary FeS(s) does not occur. However, similarities in the In and Fe porewater profiles, and the presence of In in the authigenic Fe-rich solids, reveal that part of the In becomes associated with authigenic Fe oxyhydroxides in the perennially oxic lake and is coupled to the Fe redox cycling. Comparison of the In/C-org and In/Fe molar ratios in the authigenic Fe-rich material and in surface sediments (0-0.5 cm) of this lake suggests that most non-lithogenic In was bound to humic substances. From the magnitude of the net In reaction rates, we infer that the post-depositional redistribution of this element is quantitatively not important and that the In sedimentary record represents accurately In deposition at the sampling sites. Reconstructed chronologies of the anthropogenic In deposition and comparison of In inventories among lakes point to non-ferrous metal smelters as a past significant source of atmospheric In contamination and to a significant reduction of industrial In emissions into the North American atmosphere in recent decades. (C) 2014 Elsevier Ltd. All rights reserved.

Sediment cores from eight headwater lakes located in Southern Québec, Eastern Canada, were analyzed for Pb, stable Pb isotopes, and the radioelements 210Pb, 137Cs, 241Am and 226Ra. The depth profiles of stable Pb isotope ratios show, for the post-19th century period, the influence of several isotopically distinct anthropogenic lead sources, mainly including emissions from two Canadian smelters and from leaded gasoline combustion in Canada and in the United States. A most interesting feature of the profiles, however, is the presence of sharp stable Pb isotope ratio peaks near the depth horizon, where excess 210Pb becomes undetectable. Using a binary mixing model and assuming that natural Pb concentrations and isotopic compositions from the catchment are given by the pre-industrial sediments at the bottom of the cores, we find that a significant part of the anthropogenic Pb supplied to the sediments at this horizon originated from smelting activities in the Upper Mississippi Valley. We assess that the Pb isotope ratio peaks, also observed in the laminated sediments of the Pettaquamscutt Estuary, Rhode Island, USA, are an accurate chronostratigraphic marker for the validation of mid-19th century 210Pb-derived dates. Given that the study lakes are located up to 2000km from the Mississippi Valley, we conclude that this isotopic Pb signal provides a widely distributed time-marker that is key to validate 210Pb chronologies in environmental archives from Eastern North America.

Triplicate porewater depth-profiles of pH and concentrations of total Hg (Hg-T), methylmercury (MeHg), Fe, Mn, sulfate, total sulfide, total zero-valent sulfur, organic C and major ions were determined at two sampling dates in a perennially oxygenated basin and a seasonally anoxic basin from Lake Tantare, a Canadian Shield lake. The vertical distribution of Hg-T, MeHg, acid volatile sulfide, total S, Fe, Mn, Al and organic C were also determined in dated sediment cores from the same lake basins and from the deepest site of two other lakes, one also located in the Canadian Shield and the other in the Northeastern part of the Appalachian Mountains. Application of a one-dimensional transport-reaction equation to the dissolved Hg-T and MeHg profiles constrains the depth intervals (zones) where these species are produced or consumed in the sedimentary column and yields estimates of net reaction rates of Hg-T or MeHg in each of the zones as well as their fluxes at the sediment-water interface.Dissolved Hg-T and MeHg diffused from the overlying water into the sediments, except for MeHg at one of the sampling dates in the perennially oxygenated basin. About 97% and 50% of the MeHg flux to the sediments is presently deposited with settling particles in the perennially oxygenated and seasonally anoxic basins, respectively. Removal of porewater Hg-T and MeHg occurred at all dates and sampling sites. Comparison of the consumption zones of porewater Hg-T and MeHg with the profiles of ancillary parameters, coupled with thermodynamic calculations, suggest that pure Hg mineral phases do not form in the sediments, that Hg-T and MeHg adsorption onto authigenic Fe oxyhydroxides occurs in minor proportions, and that the association of Hg-T and MeHg to Fe sulfide phases or sulfidized organic matter is possible. Assuming that the net consumption of MeHg in the porewaters was essentially due to demethylation, an apparent first-order rate constant for MeHg demethylation of 0.04-0.8 d(-1) was estimated. Production of porewater MeHg occurred only in the perennially oxygenated basin, at sediment depths where SO4 was consumed. Assuming that the net production of porewater MeHg was essentially due to methylation, an apparent first-order rate constant for Hg methylation ranging between 0.006 d(-1) and 0.1 d(-1) was calculated. These field-derived Hg methylation and MeHg demethylation rate constant values are within the range of those derived from Hg-spiked experiments. We also show that the post-depositional redistribution of total Hg during the early stages of sediment diagenesis is minor and that the solid-phase Hg-T record can be used to reconstruct the evolution of the anthropogenic Hg-T deposition. (C) 2011 Elsevier Ltd. All rights reserved.

Uranium geochemistry has been investigated in three acid lakes located on the Canadian Shield and one circumneutral lake in the Appalachian Region of Eastern Canada. In all Shield lakes, dissolved U concentrations were higher in the porewater than in the overlying water. In one of them, whose hypolimnion is perennially oxic, U released to porewater at depths of Fe remobilization was removed from the porewater at depths of Fe oxyhydroxides precipitation; these similarities in the U and Fe profiles indicate that part of the U becomes associated to Fe oxyhydroxides. The dissolved U and Fe profiles in the other two Shield lakes, whose hypolimnions were anoxic when sampled, did not show any significant recycling of these elements in the vicinity of the sediment–water interface and both elements diffused from the sediment to the overlying water. In contrast, in the Appalachian Lake, dissolved U concentrations were higher in the overlying water than in porewater, strongly decreased at the vicinity of the sediment–water interface and then remained relatively constant with sediment depth. Diagenetic modeling of the porewater U profiles, assuming steady-state, reveals that authigenic U always represented ⩽3% of the total U concentration in the sediments of all lakes. This observation indicates that diagenetic reactions involving U are not quantitatively important and that most of the U was delivered to the sediments at our study sites as particulate U and not through diffusion across the sediment–water interface, as is seen in continental margin sediments. Comparison of the U:Corg and U:Fe molar ratios in diagenetic material collected across the sediment–water interface with Teflon sheets and in surface sediments (0–0.5 cm) of the lake having a perennially oxic hypolimnion suggest that solid phase U was mainly bound to organic matter originating from the watershed; a strong statistical correlation between sediment non-lithogenic U and Corg in the Appalachian Lake supports this contention. Thermodynamic calculations of saturation states suggest that dissolved U was not removed from porewater through precipitation of UO2(s), U3O7(s) and U3O8(s) as previously proposed in the literature.

Samples of authigenic material, sediment overlying water and oxic surface sediment (0–0.5-cm depth) from a perennially oxygenated lacustrine basin were analysed to investigate which solid phases are important for binding a suite of trace elements (Ag, As, Ca, Cd, Cu, Hg, In, methylmercury (MeHg), Mg, Mo, Pb, Sb and Zn). The authigenic material, which was collected with inert Teflon sheets deployed for several years across the sediment–water interface, contained mainly poorly crystallized Fe oxyhydroxides and natural organic matter, presumably humic substances derived from the watershed. Manganese oxyhydroxides were not present in the collected authigenic material due to the slightly acidic condition (pH=5.6) of the lake that prevents the formation and recycling of these compounds. Conditional equilibrium constants for the adsorption of cationic (KFe–M) and anionic (KFe–A) trace elements onto the authigenic Fe oxyhydroxides were estimated from their concentrations in the authigenic material and in bottom water samples. These field-derived values of KFe–M and KFe–A were compared with those predicted by the surface complexation model, using laboratory-derived intrinsic adsorption constants and the water composition at the study site. Equilibrium constants (KPOM–M) were also calculated for the adsorption of the cationic trace elements onto the humic substances contained in the diagenetic material. The field-derived values of KPOM–M were compared to those predicted by the speciation code WHAM 6 for the complexation of the trace elements by dissolved humic substances in the lake. Combining the results of the present study with those on the distributions of trace elements in the porewater and solid-phase sediments reported in previous studies at the same site, it was determined whether the trace elements bind preferentially to Fe oxyhydroxides or natural organic matter in oxic sediments. The main inferences are that the anionic trace elements As, Mo and Sb, as well as the cationic metal Pb are preferentially bound to the authigenic Fe oxyhydroxides whereas the other trace elements, and especially Hg and MeHg, are preferentially bound to the humic substances.

A one-dimensional reactive transport model describing the coupled biogeochemical cycling of As, C, O, Fe, and S was used to interpret an extensive geochemical sediment (As, Fe, S, (210)Pb, (137)Cs, C(org)) and pore water (As, Fe, SO(4)(2-), SigmaS(-II) and pH) data set collected in the perennially oxygenated basin of an oligotrophic lake. Historical variations in atmospheric deposition of As and SO(4)(2-) were explicitly included as upper boundary conditions in the model calculations. The results show that the depth profile of sediment-bound As reflects both the past changes in As deposition and the diagenetic redistribution of As among the Fe(III) oxyhydroxide and Fe(II) sulfide pools. The model-predicted benthic release of dissolved As to the water column peaks 26 years after the maximum anthropogenic As input to the lake, which occurred around 1950. Two major environmental forcings of the benthic recycling of As are the organic matter degradation in the sediment and the atmospheric sulfate deposition to the lake. More oxidizing conditions associated with lower organic matter degradation rates yield a greater abundance of Fe(III) oxyhydroxides in the topmost sediment, which act as a barrier to pore water As. Variations in sulfate availability have more complex effects on benthic As remobilization, since sulfide produced by sulfate reduction may enhance both the uptake of dissolved As through the precipitation of Fe(II) sulfides and the release of dissolved As through the reductive dissolution of Fe(III) oxyhydroxides.

Profiles of porewater pH and dissolved As, Fe, Mn, sulfate, total sulfide (Sigma S-11), total zero-valent sulfur (Sigma S-0), organic carbon and major ion concentrations, as well as those of solid As, acid-volatile sulfide (AVS), total S. Fe, Mn, Al, organic C, Pb-210 and Cs-137 were determined in the sediment of four lakes spanning a range of redox and geochemical conditions. An inverse modeling approach, based on a one-dimensional transport-reaction equation assuming steady-state, was applied to the porewater As profiles and used to constrain the net rates of reactions involving As (R-net(As)). The model defines depth intervals where As is either released to (positive R-net(As)) or removed from (negative R-net(As)) the porewaters.At two of the sites, whose bottom water were oxygenated at sampling time, a production zone (R-net(As) = 12 x 10(-18) mol cm(-3) s(-1)-71 X 10(-18) mol cm(-3) s(-1)) is inferred a few cm below the sediment-water interface, coincident with sharp porewater As and Fe peaks that indicate an intense coupled recycling of As and Fe. This process is confirmed by solid As and Fe maxima just below the sediment surface. In these two lakes a zone of As consumption (R-net(As) = -5 x 10(-18) mol cm(-3) s(-1) to -53 x 10(-18) mol cm(-3) s(-1)), attributed to the slow adsorption of As to authigenic Fe oxyhydroxides, occurs just above the production zone. A second-order rate constant k(ads)(As) of 0.12 +/- 0.03 cm(3) mol(-1) s(-1) is estimated for this adsorption reaction.Such features in the porewater and solid profiles were absent from the two other lakes that develop a seasonally anoxic hypolimnion. Thermodynamic calculations indicate that the porewaters of the four lakes, when sulfidic (i.e., Sigma S-11 >= 0.1 mu M), were undersaturated with respect to all known solid As sulfides; the calculation also predicts the presence of As-V oxythioanions in the sulfidic waters, as suggested by a recent study. In the sulfidic waters, the removal of As (R-net(As) = -1 X 10(-18) mol cm(-3) s(-1) to -23 x 10(-18) mol cm(-3) s(-1)) consistently occurred when saturation, with respect to FeS(,), was reached and when Asv oxythioanionswere predicted to be significant components of total dissolved As. This finding has potential implications for As transport in other anoxic waters and should be tested in a wider variety of natural environments. Crown copyright (C) 2009 Published by Elsevier Ltd. All rights reserved.

Summary1. Stable isotopes of carbon are useful for differentiating between freshwater food chains based on planktonic algae or benthic algae, but are reported to be of limited use for identifying food chains based on sedimentary detritus. Because data from marine systems suggest that stable isotopes of sulphur (δ34S values) have potential in this regard, we tested their utility in freshwater lakes.2. We found that sulphate in the water column of four boreal lakes was enriched in 34S compared to the sulphur in bulk sediments from these lakes. Furthermore, within a given lake, insects known to feed on sediment (directly or via predation) had δ34S values similar to those of sediment, whereas planktonic and benthic invertebrates known to feed on suspended particles had δ34S values similar to those of sulphate in the water column.3. Using the stable S isotope values of invertebrates that obtain their S from either the sediment or the water column as end members in a two‐source mixing model, we show that two fish species obtain their food from both planktonic and sedimentary sources. Furthermore, model results suggest that, as expected, the more benthic‐feeding fish species obtains more of its S from the sediment compartment than does the species that feeds in the water‐column.4. Our results suggest that measurements of stable sulphur isotopes provide a means of distinguishing between members of food chains that are based in the water column from those based on sedimentary detritus. As such, they would be a useful complement to stable C isotopes that are used to distinguish between food chains based on planktonic or benthic algae.

We measured the aqueous solubility of rhombic sulfur and used this information to incorporate, in speciation codes, the thermodynamic constants reported in the literature for the formation of polysulfide complexes. Using the values of pH and total concentrations of dissolved zerovalent sulfur, sulfide, humic substances, trace metals (Ag, Cd, Cu, Hg, methylmercury, Pb, Zn) and major ions measured in anoxic porewaters of nine oligotrophic Canadian lakes as input to these speciation codes, we show that the porewaters of these lakes are undersaturated or close to saturation with respect to rhombic sulfur and that sulfide and polysulfide ligands play a dominant role in controlling metal speciation in freshwater anoxic environments when they are present at micromolar levels. The study also highlights the need for further research on the formation constants of metal complexes with sulfide and polysulfides.