Phosphorus cycling in anoxic sediments
Worldwide oxygen minimum zones (OMZs) as well as coastal oxygen-deficient regions have been shown to be expanding during recent decades. When such oxygen minima impinge on the sea floor, the retention capacity of sediments for phosphate (TPO4), ferrous iron (Fe2+), as well as ammonium (NH4+) is strongly reduced, resulting in high sea-bed release rates of these key nutrients into the bottom water. Despite the significance of the benthos exerting a major positive feedback on surface-water primary productivity and in turn maintenance of oxygen (O2) deficiency, the nutrient release in OMZ and coastal O2-deficient regions has hardly been quantified. The aim of this study was to investigate the benthic nutrient turnover in two different highly O2-deficient systems: i. the intense OMZ off Peru and ii. the landlocked Gotland Basin, Baltic Sea, which suffers from anthropogenically induced eutrophication. The focus was on the phosphorus (P) cycle but associated cycles of iron (Fe) and nitrogen (N) were also included. Off the coast of Peru, benthic fluxes of TPO4 and Fe2+ were quantified in situ using benthic landers and were calculated from pore-water profiles across a latitudinal depth transect at 11°S. This transect extended from 80 m to 1000 m water depth and covered anoxic to oxic bottom-water conditions. The working area was divided into three different zones: the shelf that is subjected to periodically fluctuating bottom-water O2 conditions, the core of the OMZ where anoxia can be assumed to be permanent, and the depth range below 500 m where O2 levels increased again. TPO4 fluxes were high (maximum 292 mmol m-2 yr-1) throughout the shelf and in the core of the OMZ. In contrast, Fe2+ fluxes were high on the shallow shelf (maximum 316 mmol m-2 yr-1) but moderately low (15.4 mmol m-2 yr-1) in water depths between 250 m and 600 m due to the continuous reduction of Fe oxides and Fe hydroxides (henceforth referred to as Fe oxyhydroxides). Below 600 m, where O2 concentrations increased, Fe2+ fluxes became negligible due to the precipitation of Fe2+ in the oxic sediment surface. Ratios between organic carbon degradation and TPO4 flux indicated an excess release of P over carbon (C) when compared to Redfield stoichiometry. This was most likely caused by preferential P release during organic matter degradation, dissolution of fish debris, and/or P release from sulfide-oxidizing microbial mat communities. Fe oxyhydroxides were relevant as a P source only on the shallow shelf. The benthic fluxes are among the highest reported from similar O2-deficient continental margin systems, and highlight the efficiency of OMZ sediments returning TPO4 and Fe2+ to the bottom water. The shelf region is particularly important in this regard since O2 fluctuations likely trigger a complex biogeochemical reaction network of P, Fe and sulfur turnover resulting in transient, high TPO4 and Fe2+ release under anoxia. Sources for P release were further constrained by combining P speciation data, based on sequential extraction of sediment samples, with a mass balance and benthic modeling. P speciation revealed that authigenic calcium phosphate (Ca-P; including carbonate fluorapatite, biogenic apatite from fish remains, and calcium carbonate-bound P), was the major fraction along the transect. It accounted for 35 to 47% of the depth-averaged total extracted P on the shelf and upper slope, but for > 70% below 300 m water depth. Further extraction of fish-P showed that below 259 m water depth this fraction dominated the authigenic Ca-P pool by 60 to 69%. Organic P was present in considerable amounts (18 to 37%) only at the shelf and the upper slope, whereas detrital P and P bound to Fe oxyhydroxides was generally of minor importance at all sites. Organic matter in surface sediments was highly depleted in P relative to Redfield stoichiometry with C:P ratios of up to 516. The benthic model found preferential P mineralization in the water column or, alternatively, preferential P release during organic matter degradation in the sediment surface as possible pathways explaining such high C:P ratios. Nevertheless, both model and mass balance calculations revealed that irrespective of which pathway prevails, organic P was only of minor importance for the benthic P budget of Peruvian OMZ sediments. According to the solid phase speciation, authigenic Ca-P, with a high contribution of fish debris, is a likely candidate for the missing source of P required to close the P budget. These sediments were identified as weak sinks for P, as more than 80% of the imported P was recycled back into the water column. In the Gotland Basin, TPO4 and DIN fluxes were quantified in situ across an oxic to anoxic depth-transect using benthic landers. A CTD-water sampling rosette was deployed to record the nutrient and O2 distribution in the water column and thereby investigate the benthic-pelagic coupling because of its significance for the euthrophication state of the Baltic Proper. The study area was divided into three different zones: the oxic zone at 60 m to < 80 m water depth, the hypoxic transition zone between > 80 m and 120 m, and the deep anoxic and sulfidic basin at > 120 m. The hypoxic transition zone was characterized by fluctuating O2 levels as well as the occurrence of extended mats of sulfur bacteria. Beside the deep anoxic basin, the hypoxic transition zone was revealed as a major release site for TPO4 and NH4+ with rates of up to 0.2 mmol m-2 d-1 and 1 mmol m-2 d-1, respectively. There are clear indications that the bacterial mats converted NO3-/NO2- into NH4+ during dissimilatory nitrate reduction to ammonium (DNRA), thereby retaining reactive N in the ecosystem. The transient release and uptake of TPO4 during oscillating anoxic and oxic conditions by these bacteria, however, can only be speculated as the entire TPO4 release from the sediment could be potentially covered by preferential P release during organic matter degradation. Extrapolation of benthic fluxes to the Baltic Proper resulted in internal TPO4 and DIN loads of 109 kt yr-1 and 295 kt yr-1, respectively, which is significantly higher than external P and DIN loads. This up-scaling of fluxes revealed the importance of the hypoxic transition zone for the internal nutrient loading, which only covered 51% of the total considered area, but released as much as 70% of the total TPO4 load. Likewise, 75% of the internal NH4+ load (200 kt yr-1) was released from this particular environment; however, this NH4+ did not reach the surface mixed layer. This resulted in the supply of water with a low N:P ratio to the euphotic zone. In summertime, such low N:P ratios favor the development of N2-fixing cyanobacterial blooms which, by different feedback processes, counteract the recovery of the Baltic Proper from eutrophication.