The fraction of the mineralised phosphorus is adsorbed to sediment particles but the rest is instantly released

to the water column. In this study, this pathway was simplified by excluding the desorption process. The model equations and parameter values are described in detail in  and . Calibration of the new N flux model and a simplified version of P flux model presented by Müller-Karulis & Aigars (2011) against median PO43−, NOx− and NH4+ flux measurements was performed using a simulated annealing routine (SANN) in statistical analysis software R v.3.0.2. Epacadostat The average fluxes of PO43− (42–115 μmol m−2 d−1) were always directed out of the sediments. Although PO43− fluxes tended to decrease with increasing O2 concentration in the near-bottom water,

they exhibited no significant differences (ANOVA; p < 0.01) among treatments, most likely due to the substantial variability of fluxes within the treatments ERK inhibitor ( Figure 3). The simulated values of PO43− flux (Figure 3) are in good agreement with the median values of the experimental data set and show nearly constant maximum values (105–106 μmol PO43− m−2 d−1) at an O2 concentration range of 1–2 mg l−1 and a smooth decline with increasing O2 concentrations, reaching the lowest fluxes (57 μmol PO43− m−2 d−1) at oxygen concentrations in the range between 5 and 10 mg l−1. Sediment-water fluxes of NH4+ are always positive and exhibit large variability within and among O2 treatments, ranging on average from Erastin research buy 1800 μmol m−2 d−1 at an O2 concentration of 2 mg l−1 to 140 μmol m−2 d−1 at an O2 concentration of 10 mg l−1 ( Figure 4). At this latter O2 concentration the observed fluxes vary between –734 and 528 μmol NH4+ m−2 d−1 (the highest observation

is treated as an outlier) with 90 μmol NH4+ m−2 d−1 as the median value. Although there is no significant difference in NH4+ fluxes between treatments 1 and 3, the significant differences between treatments 2 and 3 (ANOVA; p < 0.01) and 3 and 4 (ANOVA; p < 0.01) clearly demonstrate increasing NH4+ fluxes when O2 concentrations are < 4 mg l−1. Larger oxygen concentrations do not result in a further decrease of NH4+ fluxes, however. The modelled NH4+ fluxes (Figure 4) show a smooth decline with increasing concentration, reaching the lowest value (2.3 μmol NH4+ m−2 d−1) at an O2 concentration of 10 mg l−1. The model fits the data well at low (1 mg l−1) and intermediate to high (≥ 4 mg l−1) O2 concentrations, but does not correspond with the high fluxes observed at an O2 concentration of 2 mg l−1, which vary between 1051 and 2467 μmol NH4 m−2 d−1. In contrast to NH4+, NOx− fluxes are mostly directed into the sediments, although, like NH4+, these fluxes exhibit a considerable variability within and among treatments, ranging on average from –390 μmol m−2 d−1 at an O2 concentration of 1 mg l−1 to 85 μmol m−2 d−1 at an O2 concentration of 10 mg l−1 ( Figure 5).