The effect of resuspension on chemical exchanges at the sediment-water interface in the deep sea — A modelling and natural radiotracer approach

We review the processes responsible for the formation of vertical gradients in the chemical composition of suspended particles across the benthic nepheloid layer. Such gradients have usually explained by resuspension of surface sediments, but it is shown here that these gradients can only be understood as part of a dynamic exchange between the water column and the sediments. A coupled model, developed in a companion paper, is expanded to include chemical reactions above and below the sediment-water interface. Three cases are discussed: A tracer with first-order decay (Model 1), the dissolution of a major constituent (Model 2), and a particle-reactive tracer with first-order decay and production in the water column (Model 3). Using typical parameter values for a well-developed benthic nepheloid layer, the three models reproduce typical distributions of C_org (organic carbon), opal, and ²³⁴Th, respectively, on particles above and below the sediment-water interface. Sensitivity analyses illustrate how bioturbation can cause the large discrepancy observed between suspended and surface sediment C_org values (Model 1). The model also reconciles this observed discrepancy with observations that the major part of the decomposition takes place within the sediment. For opal (Model 2), the influence of resuspension on the burial rate of opal is shown to be negligible, as long as dissolution follows first-order kinetics and is not enhanced by turbulence in the suspended phase. The modelling of ²³⁴Th (Model 3) successfully links the depletion of ²³⁴Th in bottom waters with the distribution of excess ²³⁴Th in surface sediments and on resuspended particles.²³⁴Th is a powerful example of the tools supplied by the radioactive daughters of the natural U and Th decay series in studying fluxes and exchange rates of solutes and particles across the sediment-water interface. A short review is given of these tools, and it is shown how they can be used to obtain rate information required to apply and calibrate specific resuspension models.