g. Koop et al. (1990) and Mort et al. (2010). The simulation of ammonium generated from organic matter is split into pathways of nitrification of ammonium, which intensifies with increasing oxygen concentration, and the release of ammonium to the water column. The check details deep parts of the Baltic Sea, such as the Gotland Deep and the Gdańsk Deep, are on occasion characterised by anoxic sediments. Under such conditions nitrification is highly dependent on the dynamics of the redoxcline, which determines the mixing of ammonium-rich waters with oxygenated ones (Hietanen et al. 2012). In the Gulf of Riga, long-term average and minimal oxygen

concentrations rarely reach hypoxic levels and never anoxic levels (Müller-Karulis & Aigars 2011). Furthermore, organic nitrogen mineralisation in the Gulf of Riga delivers large amounts of ammonium (Henriksen & Kemp 1988, Tuominen et al. 1998, Savchuk 2002). Therefore, both ammonium and oxygen supplies should be appropriate for continuous nitrification. However, despite the suitable conditions for nitrification and the reasonable correlation between the simulated and observed ammonium fluxes (Table 1), the dynamics of observed 3Methyladenine ammonium and thus its modelling approach contains some issues that need clarification, for example, the high observed experimental values of NH4+ flux at an O2 concentration

of 2 mg l−1 (Figure 4). This oxygen concentration marks the borderline between hypoxic and oxygenated conditions, as well as the oxygen level needed to sustain most animal life (Hansson et al. 2011). According to Henriksen & Kemp (1988), the higher observed ammonium flux at oxygen concentrations of 2 mg l−1 may be related to the less efficient activity of nitrifying bacteria, which 5-FU in vivo are outcompeted by

heterotrophic bacteria at low oxygen concentrations. Moreover, McCarthy et al. (2008) indicate that the hypoxia threshold provides good conditions for dissimilatory nitrate reduction to ammonium (DNRA). The findings of these authors, as well as the 108% higher NH4+ flux at oxygen concentrations of 2 mg l−1 as compared to ammonium fluxes at lower and higher concentrations (Figure 4), lead us to the conclusion that studies of DNRA and the processes driving it in the Gulf of Riga should be undertaken and that the biogeochemical model should be expanded to include DNRA. Compared to the previously reported results of the biogeochemical model of the Gulf of Riga (Müller-Karulis & Aigars 2011), the simulation of the nitrate flux has been improved in the current study. Here, the simulated nitrate flux increases with oxygen concentration. It is formed as the sum of nitrate diffusion, which marks the nitrate inflow in sediments from the overlying water and thus Dw, and the portion of nitrified nitrate that escapes denitrification, which represents the outward flux from sediments.