A multicomponent reactive transport model of early diagenesis: Application to redox cycling in coastal marine sediments

Abstract

STEADYSED1 is a multicomponent reactive transport code for steady-state early diagenesis which fully incorporates the reaction couplings among the elements, C, O, N, S, Fe, and Mn. The model is tested against extensive datasets collected by Canfield et al. (1993a,b) at three coastal marine sites that exhibit high rates of combined iron and manganese (hydr)oxide reduction. It is shown that the model provides a consistent explanation of the entire body of multicomponent multisite observations. The measured concentration profiles of twenty-eight individual porewater and solid sediment species are satisfactorily reproduced. Furthermore, the model predicts the observed distributions of sulfate reduction rates, as well as diagnostic features of the porewater pH profiles. The coupled nature of the reactive species imposes stringent constraints when fitting model-calculated distributions to the data, because a single set of reaction and transport parameters must account for the profiles of all the species at each site. The parameters are further separated into site-specific (e.g., deposition fluxes, bottom water composition) and reaction-specific parameters (e.g., rate coefficients, apparent equilibrium constants, limiting substrate concentrations). By minimizing the variations of the reaction-specific parameters from one site to another, the fitting strategy emphasizes the retrieval from field data of reaction parameters that are mechanistically meaningful. The model reproduces the significant vertical overlap between organic carbon oxidation pathways observed in the sediments. The overlap is explained by the combination of high rates of organic carbon oxidation plus intense bioturbation and irrigation at the sites studied. The model-derived contributions of the various respiratory processes compare favorably with the incubation results of Canfield et al. (1993a,b). Aerobic respiration accounts for less than 32% of total organic carbon oxidation at the three sites. From 22 to 46% of the O 2 uptake by the sediments is directly coupled to organic carbon oxidation, while the remainder is used for the oxidation of secondary reduced species, in particular dissolved, adsorbed and solid Fe(II) and Mn(II) species. According to the model, the oxidation of Fe 2+ by Mn (hydr)oxides is responsible for the observed spatial separation of the porewater build-up of Mn 2+ and Fe 2+. At two of the sites, 75 and 97% of the total rate of Mn oxide reduction is due to chemical reaction with dissolved Fe 2+. In contrast, at the same sites, iron (hydr)oxides are mostly utilized by bacteria for the oxidation of organic matter (dissimilatory iron reduction). As shown also by Canfield et al. (1993a,b), dissimilatory reduction is the principal dissolution pathway for manganese oxides at the third site. A sensitivity analysis suggests that, for given deposition fluxes of reactive Fe and Mn, the competition between dissimilatory and nondissimilatory metal reduction pathways depends primarily on the total carbon oxidation rate and the intensity of porewater irrigation. The simulations also highlight the importance of adsorption-desorption of Fe(II) and Mn(II) in the redox cycling of the metals, as well as their impact on porewater alkalinity and pH. Based on the calculated rate distributions, detailed budgets of Fe, Mn, and O 2 in the sediments are presented. The model-calculated benthic exchange fluxes of solutes are dominated by irrigation. For nitrate, molecular diffusion and irrigation cause fluxes in opposite directions. As a result, there is a net transfer of nitrate from the water column to the sediments, although the interfacial porewater gradients predict diffusional fluxes out of the sediments. A significant fraction of the benthic Fe and Mn fluxes to the bottom waters may be due to desorption at the water-sediment interface.