Strong and Dynamic Benthic-Pelagic Coupling and Feedbacks in a Coastal Upwelling System (Peruvian Shelf)

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Monthly time-series data (1998–2009) of bottom water oxygen, nitrate and nitrite concentrations from the outer shelf (150 m water depth) in the oxygen minimum zone offshore Peru were coupled to a layered biogeochemical sediment model to investigate benthic-pelagic coupling over multi-annual time scales. The model includes the mineralization of four reactive pools of particulate organic carbon (POC) with lifetimes of 0.13, 1.3, 20 and 1700 yr that were constrained using empirical data. Total POC rain rates to the seafloor were derived from satellite based estimates of primary production. Solute fluxes and concentrations in sediment porewater showed highly dynamic behavior over the course of a typical year. Conversion of fixed N to N2 by denitrification varied from 0.9 mmol m-2 d-1 of N in winter to 1.6 mmol m-2 d-1 of N in summer with a long term mean N loss for the shelf of 1.3 mmol m-2 d-1 of N. Fixed N loss across the whole time-series agreed very well with a previously-published vertically-integrated sediment model for coupling the benthic and pelagic N cycle in regional and global models. Dissimilatory nitrate reduction to ammonium (DNRA) emerges as a major process in the benthic N cycle, producing on average 2.0 mmol m-2 d-1 of ammonium: more than twice the rate of ammonification of organic nitrogen. The model predicts sulfide emissions from the sediment of up to 1 mmol m-2 d-1 when POC rain rate exceeds 20 mmol m-2 d-1, in agreement with past observations of benthic sulfide fluxes and sulfide plume distributions in the water column. This study demonstrates that sediments on the Peruvian shelf are not static repositories that are independent of changes taking place in the water column. Our results strongly suggest the shelf sediments must exert an important feedback on biogeochemical processes in the overlying waters, and should be considered in regional model studies.

This publication has 67 references indexed in Scilit: