Carbon Pumps in the Benguela Current upwelling system

Abstract

The biological carbon (C) pump influences the flux of carbon dioxide (CO2) between the ocean and the atmosphere by carbon assimilation during the photosynthesis of particulate organic carbon (POC) and the precipitation of calcium carbonate particles (PIC). Regions of intense biological carbon pumping are eastern boundary upwelling systems, among which the Benguela upwelling system (BUS) is the most productive but the least studied one. The main objective of this thesis is to investigate the functioning of the biological carbon pump in the northern BUS (NBUS), to assess whether the BUS is a net source or sink for carbon and to give basis for evaluating potential responses of the BUS to global change. Dissolved inorganic nutrient and carbonate chemistry data was raised and used to study the spatio-temporal variability of the remineralisation ratios of C, nitrogen (N) and phosphate (P) (C/N/P) with emphasis on the variability of N deficiency, which is a major factor that limits the production of POC. Records of annual particle fluxes along with hydrographical data were obtained by a mooring on the shelf off Walvis Bay. In conjunction with meteorological records, these data were used to investigate the influence of upwelling on the particle fluxes with a focus on the factors that control the export of POC and PIC on the shelf. Furthermore, carbonate chemistry data allow to determine the variability of the carbonate saturation state (Ω) which influences calcification, growth and survival in many marine calcifying organisms. Along with patterns of air-sea CO2 fluxes, these results were used to evaluate the C source-sink function of the BUS. The results show that the release of C, N and P in the course of organic matter decomposition follows the Redfield stoichiometry of C/N/P of 106/16/1 in the oxygenated water column. However, N loss and P input, which occur in sub- and anoxic bottom waters overlying the mud belt of the Namibian shelf, lower the N/P to <16 in upwelling intermediate water. Together, these processes result in an N deficiency (-N*), equivalent to a relative P excess (+P*), which surfaces and extends to the adjacent hemipelagic ocean. The results suggest that the magnitude of exported P* is not only controlled by the degree of local O2 deficiency on the shelf, but as well by the amount of remotely supplied N* excess (N/P >16, +N*). The findings point to a mechanism which replenishes the N* deficiency on the NBUS shelf and thereby counteracts the N limitation that hampers the N-driven CO2 drawdown of the biological carbon pump in the NBUS. The combined results of hydrographical, meteorological and particle flux data show that even comparatively weak upwelling of high SACW fractions fuels higher POC fluxes (>91 g POC m-2 yr-1) than intense upwelling of Eastern South Atlantic Central Water (ESACW). This reflects the higher nutrient inventory of SACW compared to ESACW, among other factors. The results further indicate that the POC export on the shelf is dominated by diatom productivity. The decomposition of the basically diatomaceous organic matter causes very low Ω (ΩAmin = 0.7, ΩCmin = 1.2) and high silicate concentrations in ascending sub-thermocline waters. This corrosive water reduces the PIC formation in the surface and its export (11 g PIC m-2 yr-1) on the Namibian shelf, relative to the high, diatom-dominated POC formation and export (66 g POC m-2 yr-1). This results in a high POC/PIC ratio of 6 in sinking matter, pointing to an efficient CO2 uptake by the coastal biological carbon pump. Towards the slope and the open ocean, an increasing contribution of PIC to POC export is enabled by rising Ω and depleted silicate concentrations, providing an ecological advantage for carbonate producers. As indicated by the results of P* in offshore waters, the biological pump’s efficiency in the offshore sector of the upwelling system is hampered by N limitation, thus lowering the N-driven CO2 drawdown. In terms of nutrient stoichiometry, the unused P* constitutes a leakage of potential C export that accounts for ~25 % of the total N-driven potential export fluxes of ~98 Tg C yr-1 of the NBUS. Clearly, NBUS productivity suffers from the impact of N loss or P gain incurred from suboxic conditions on the shelf. The CO2 emission patterns reveal that the NBUS is a CO2 source in the order of 13.6 Tg C yr-1, likely reflecting the N deficit in upwelling subsurface water. The southern BUS (SBUS) is a significant sink of -3.4 Tg C yr-1 and reflects the higher contribution of biologically unused, preformed nutrients in upwelling source waters of the SBUS compared to the NBUS. Adding the combined POC and PIC export sink of (-66 and -11 g C m-2 yr-1 to the CO2 emissions of the NBUS (+71 g C m-2 yr-1) indicates that the NBUS is a small net sink of -6 g C m-2 yr-1.