• PDF

    Final published version, 1 MB, PDF-document


Carbon, nitrogen, oxygen and sulfide budgets are derived for the Black Sea water column from a coupled physical–biogeochemical model. The model is applied in the deep part of the sea and simulates processes over the whole water column including the anoxic layer that extends from similar, equals115 m to the bottom (similar, equals2000 m). The biogeochemical model involves a refined representation of the Black Sea foodweb from bacteria to gelatinous carnivores. It includes notably a series of biogeochemical processes typical for oxygen deficient conditions with, for instance, bacterial respiration using different types of oxidants (i.e denitrification, sulfate reduction), the lower efficiency of detritus degradation, the ANAMMOX (ANaerobic AMMonium OXidation) process and the occurrence of particular redox reactions. The model has been calibrated and validated against all available data gathered in the Black Sea TU Ocean Base and this exercise is described in Gregoire et al. (2008). In the present paper, we focus on the biogeochemical flows produced by the model and we compare model estimations with the measurements performed during the R.V. KNORR expedition conducted in the Black Sea from April to July 1988 (Murray and the Black Sea Knorr Expedition, 1991). Model estimations of hydrogen sulfide oxidation, metal sulfide precipitation, hydrogen sulfide formation in the sediments and water column, export flux to the anoxic layer and to the sediments, denitrification, primary and bacterial production are in the range of field observations. With a simulated Gross Primary Production (GPP) of 7.9 mol C m−2 year−1 and a Community Respiration (CR) of 6.3 mol C m−2 year−1, the system is net autotrophic with a Net Community Production (NCP) of 1.6 mol C m−2 year−1. This NCP corresponds to 20% of the GPP and is exported to the anoxic layer. In order to model Particulate Organic Matter (POM) fluxes to the bottom and hydrogen sulfide profiles in agreement with in situ observations, we have to consider that the degradation of POM in anoxic conditions is less efficient that in oxygenated waters as it has often been observed (see discussion in Hedges et al., 1999). The vertical POM profile produced by the model can be fitted to the classic power function describing the oceanic carbon rate (CR=Z−α) using an attenuation coefficient α of 0.36 which is the value proposed for another anoxic environment (i.e. the Mexico Margin) by Devol and Hartnett (2001). Due to the lower efficiency of detritus degradation in anoxic conditions and to the aggregation of particles that enhanced the sinking, an important part of the export to the anoxic layer (i.e. 33%, 0.52 mol C m−2 year−1) escapes remineralization in the water column and reaches the sediments. Therefore, sediments are active sites of sulfide production contributing to 26% of the total sulfide production. In the upper layer, the oxygen dynamics is mainly governed by photosynthesis and respiration processes as well as by air–sea exchanges. similar, equals71% of the oxygen produced by phytoplankton (photosynthesis+nitrate reduction) is lost through respiration, similar, equals21% by outgasing to the atmosphere, similar, equals5% through nitrification and only similar, equals2% in the oxidation of reduced components (e.g. Mn2+, Fe2+, H2S). The model estimates the amount of nitrogen lost through denitrification at 307 mmol N m−2 year−1 that can be partitioned into a loss of similar, equals55% through the use of nitrate for the oxidation of detritus in low oxygen conditions, similar, equals40% in the ANAMMOX process and the remaining similar, equals5% in the oxidation of reduced substances by nitrate. In agreement with data analysis performed on long time series collected since the 1960s (Konovalov and Murray, 2001), the sulfide and nitrogen budgets established for the anoxic layer are not balanced in response to the enhanced particle fluxes induced by eutrophication: the NH4 and H2S concentrations increase.
Original languageEnglish
Pages (from-to)2287-2301
JournalEcological Modelling
Issue number19
StatePublished - 2010

ID: 322145