the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 18 Dec 2025
Seasonal variations and controlling factors of nitrogen fluxes at the sediment-water interface in a semi-enclosed inland sea
Abstract. Nitrogen fluxes across the sediment-water interface and nitrogen removal from sediments are essential components of the nitrogen cycle and ecosystem in semi-enclosed inland seas. However, the difficulty in observational sampling hinders the acquisition of continuous data necessary to understand their seasonal variations and underlying mechanisms. In response to this issue, we have developed a one-dimensional vertical model of the nitrogen cycle within sediments and used it to reproduce the seasonal changes of observed nitrogen fluxes in a typical semi-enclosed inland sea and investigate their controlling factors through sensitivity experiments. Model results indicate that 40 % of particulate organic nitrogen (PON) settling into sediments is returned to the bottom water as dissolved inorganic nitrogen (DIN), while 30 % is removed via N-loss flux (dinitrogen gas and nitrous oxide). Although PON flux is controlled by PON concentration in the bottom water, DIN and N-loss fluxes show temperature-driven seasonal variations, suggesting a decoupling between nitrogen return and PON input. Additionally, seasonal variations in oxygen penetration depth (OPD), ranging from 1 to 3 mm, also affect nitrogen fluxes. In nitrate-depleted sediments of semi-enclosed seas, the denitrification rate is no longer significantly higher than the anammox rate in the nitrogen removal.
- Preprint
(1811 KB) - Metadata XML
-
Supplement
(2509 KB) - BibTeX
- EndNote
Status: open (until 13 Feb 2026)
- RC1: 'Comment on egusphere-2025-6187', Andy Dale, 31 Dec 2025 reply
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 88 | 45 | 12 | 145 | 23 | 26 | 23 |
- HTML: 88
- PDF: 45
- XML: 12
- Total: 145
- Supplement: 23
- BibTeX: 26
- EndNote: 23
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Wu et al. investigate seasonal nitrogen (N) sources and sinks in a shallow, semi-enclosed temperate sea in Japan using an empirically constrained diagenetic model. I am very sympathetic to the approach taken. Considerable effort has clearly gone into generating the data used to constrain the model, and the study has the potential to provide valuable insights into N cycling in the Seto Inland Sea and coastal seas in general. At present, however, the manuscript suffers from a lack of clarity in the description of the model formulation, assumptions, and parameterization. As a result, it is difficult to properly evaluate the results and discussion. Much more care is needed to explain the model step-by-step. The issues outlined below require substantial revision in my view, followed by a second round of reviews. If a revised version is invited, I would be happy to re-evaluate the manuscript in detail.
Study area. The study area should be described before the model. Key contextual information is missing, including hydrography, sediment characteristics, primary production, and especially dissolved oxygen (O₂) dynamics. Given the central role of O₂ in the biogeochemical reaction network, available O₂ data should be presented (e.g., in Fig. 3).
Inconsistent units. Units are inconsistent throughout the text, tables, figures, and equations. For example, diffusion coefficients in Eqs. 1–2 are given in m² s⁻¹, while sedimentation rates are reported in mm yr⁻¹, without explicit unit conversions. Table 2 mixes mm, cm, and m, and concentrations are variously expressed as µmol L⁻¹, mmol m⁻³, or mass-based units. Units should be standardized throughout.
Conceptual model and state variables. The conceptual diagram in Fig. 1 is incomplete. The fluff layer is not shown, and NO₂⁻ is absent as a state variable despite appearing in the anammox rate formulation (Table 2). The Supplement suggests that NO₂⁻ is prescribed as a constant, but the rationale and implementation are unclear. If NO₂⁻ is important for anammox, it should be treated as a dynamic state variable. This may partly explain the unusually large modelled contribution of anammox to fixed N loss (Fig. 5d), which exceeds typical values reported for coastal sediments (e.g., Dalsgaard et al., 2005; doi:10.1016/j.resmic.2005.01.011). Although porewater profiles in Fig. 4 are reasonably reproduced, this may be a result of unrealistic internal N cycling if key intermediates are not modelled explicitly.
Treatment of oxygen. Several reaction terms depend on O₂, yet O₂ is not included as a state variable. Instead, an oxygen penetration depth (OPD) appears to be imposed. It is unclear whether O₂ concentrations are assumed constant above the OPD. Given the strong control of O₂ on N transformations, this approach is difficult to justify. Including O₂ as a dynamic, seasonally varying state variable would substantially improve the model. At minimum, a fixed O₂ upper boundary condition would be preferable to a static OPD. Again, without doubt this will have an important influence on the anammox rate. Seasonal variability in O2 concentrations over depth versus time would be a key plot to show in the main manuscript.
Other comments
N₂O is reported as a model output (Line 277), but the corresponding governing equations are not included in Table 1. These must be provided.
Table 1 lists a diffusive boundary layer thickness of 3 m, which is likely a typo (∼3 mm would be more realistic). This should be clarified. If a diffusive boundary layer is included, Fig. 1 and Eq. 9 should be revised to reflect flux continuity at the sediment–water interface.
It is unclear how accumulation and erosion of the fluff layer are treated, particularly in relation to the advection and diffusion terms in Eqs. 1–2. This needs explicit explanation.
Table 1. PO4 does not need to be included in the model description since it is not simulated.
Reaction equations in Table 1are not balanced with respect to H, O, or charge. Each POC degradation term includes an additional limitation factor (1/lim), even though limitation terms are already specified. The stoichiometric coefficients (x, y) could be used directly in the mass-balance equations (Eqs. 5–7), potentially removing the need for the rCN parameter. The basis for the chosen x and y values should be explained. The assumed oxidation state of organic carbon (apparently zero) should also be stated explicitly.
Anaerobic solutes are represented by a lumped oxygen demand unit (ODU), but ODU is not treated as a state variable. Consequently, O₂ and NO₃⁻ consumption during ODU oxidation is not represented. Explicit inclusion of ODU would improve internal consistency and confidence in the model output.
Why was bioirrigation not included in the model? I would assume that this would be a major solute transport term in coastal sediments, even if hypoxic (Dale et al., 2013; doi:10.5194/bg-10-629-2013). This needs careful justification.
The denominator in Eq. 4 should read 2 ln(porosity) instead of 2.02 ln(porosity).
According to Fig. 4, NH4 fluxes are out of the sediment, opposite to NO3, yet in Fig. 5b the NO3 and NH4 fluxes have the same sign.
The numerical code used for the model should be specified, and model mass-balance performance should be reported. Analytical methods are insufficiently described. Finally, both the model code and the empirical data should be made publicly available in an online repository for scrutiny by the reviewers.
Andy Dale 31.12.2025