the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 10 Jun 2024
Dissolved Mn(III) is a key redox intermediate in sediments of a seasonally euxinic coastal basin
Abstract. Manganese (Mn) is an essential micronutrient and key redox intermediate in marine systems. The role of organically complexed dissolved Mn(III) (dMn(III)-L) as an electron acceptor and donor in marine environments is still incompletely understood. Here, we use geochemical depth profiles and a reactive transport model to reconstruct the seasonality in sedimentary dMn(III)-L dynamics and benthic Mn release in a eutrophic, seasonally euxinic coastal basin (Lake Grevelingen, the Netherlands). We find that dMn(III)-L is a major component of the dissolved Mn pool throughout the year. Our model indicates that, when O2 is present in the bottom water, there are three major sources of pore water dMn(III)-L, namely reduction of Mn oxides coupled to the oxidation of Fe(II), reduction of Mn oxides coupled to organic matter degradation and oxidation of Mn(II) with O2. Removal of pore water dMn(III)-L primarily takes place through reduction by dissolved Fe(II). When bottom waters are euxinic in summer, rates of sedimentary Mn cycling decrease strongly, because of a lower supply of Mn oxides. The dMn(III)-L transformations in summer mostly involve reactions with Fe(II) and organic matter. Benthic release of Mn mainly occurs as dMn(III)-L when bottom waters are oxic, as Mn(II) upon initial bottom water euxinia and as both Mn(II) and dMn(III)-L when the euxinia becomes persistent. Our findings highlight strong interactions between the sedimentary Fe and Mn cycles. Dissolved Mn(III)-L is a relatively stable and mobile Mn species, when compared to Mn(II), and is therefore more easily transported laterally throughout the coastal zone and possibly also to open marine waters.
- Preprint
(1260 KB) - Metadata XML
-
Supplement
(1218 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2024-1706', Anonymous Referee #1, 02 Aug 2024
Review Klomp et al.
General comments
Klomp et al. investigate seasonal differences in the sedimentary Mn cycling in a eutrophic coastal basin in the Netherlands. Their study provides a comprehensive data set and combines geochemical water column, pore-water and solid-phase data with reactive transport modeling.
The authors emphasize that dissolved Mn(III)-L is an important component of the Mn cycle throughout the year. Depending on the seasonal oxygen concentration in the bottom water and the influx of Mn oxides, Mn is predominantly released to the overlying water as dissolved Mn(III)-L in winter (oxic bottom water and higher influx of Mn oxides), whereas in summer (euxinic bottom water and lower influx of Mn oxides), both dissolved Mn(II) and dMn(III)-L are released, but Mn(II) is the dominant species. In contrast to dissolved Mn(II), the relatively stable and mobile dMn(III)-L may be transported from coastal areas into the open ocean. In addition, the biogeochemical processes leading to the formation and removal of dMn(III)-L in the sediment are strongly linked to the Fe cycle. Therefore, this study is very helpful in improving our understanding of sedimentary Mn and Fe cycling, especially the coupling between the two cycles, in coastal areas.
Overall, the manuscript is well written, however some of the figures could be improved (see specific comments). I would recommend the publication of this manuscript after some minor revisions.
Specific comments
Abstract
Line 11: It is better to write “depth profiles of water, pore-water and solid-phase data” instead of just “depth profiles”. Otherwise, it is not clear which comprehensive data set has been the basis for this study.
Methods
Figure 1a: The map of Lake Grevelingen including the overview map of the Netherlands in the top right corner is too small. It is very difficult to identify the water depths, especially in the Scharendijke basin.
Results
Figure 3: In contrast to the water column profiles, the pore water profiles are relatively small. As some of the discussion relates specifically to the upper centimeters, it would be helpful if these were shown in more detail. In addition, the light grey dots and lines are hard to see.
Figure 3h-j: If I see it correctly, the sum of dMn(III)-L and Mn(II) does not correspond to the TD Mn concentrations. Is there any explanation for this?
Figure 4: Again, the light grey dots and lines are hard to see.
Line 237 + Figure 4a: As the Corg content usually refers to the total sediment weight, it is better to speak of Corg contents rather than Corg concentrations.
Line 245-311: The description of the model results already contains many possible interpretations or approaches for discussion. It would be good if these were directly linked with the discussion part.
Figure 5: At first glance, the figure is a little confusing due to many numbers/rates. One suggestion here would be to split the figure in two: (a) situation in March, (b) situation in September. In this way, the size of the arrows could be adjusted to the respective rates to better highlight the differences in the rates between March and September. In addition, the oxygen conditions in the bottom water could then be integrated for both situations.
Discussion
Line 334: Again, it would be very good if the top 0 to 10 cm were shown in more detail to see the strong counter gradients.
Line 394-395: At this point it should be described in more detail to what extent the transport of trace metals is decoupled from that of Mn when it is mainly present as dMn(III)-L. Were trace metals measured in the water column samples that could confirm the mentioned hypothesis?
Technical corrections
Line 16: Since manganese (Mn) is introduced in line 11, oxygen should be introduced in the same way: Our model indicates that, when oxygen (O2) is present in the bottom water, there are three major sources of pore water dMn(III)-L (…).
Line 43: Again, please also introduce oxygen for consistency.
Line 307-308: Here the authors are certainly referring to Figure 8, as Figure 9 does not exist.
Line 370: The sentence structure is a bit awkward. It would be better: When bottom water O2 re-establish in October, the influx of Mn and Fe oxides, the rates of sedimentary Mn cycling and the benthic flux of Mn all increase.
Line 388: Instead of saying “when it meets O2”, it is more appropriate to write “when it is exposed to O2”.
Citation: https://doi.org/10.5194/egusphere-2024-1706-RC1 -
AC1: 'Reply on RC1', Robin Klomp, 02 Oct 2024
Klomp et al. investigate seasonal differences in the sedimentary Mn cycling in a eutrophic coastal basin in the Netherlands. Their study provides a comprehensive data set and combines geochemical water column, pore-water and solid-phase data with reactive transport modeling.
The authors emphasize that dissolved Mn(III)-L is an important component of the Mn cycle throughout the year. Depending on the seasonal oxygen concentration in the bottom water and the influx of Mn oxides, Mn is predominantly released to the overlying water as dissolved Mn(III)-L in winter (oxic bottom water and higher influx of Mn oxides), whereas in summer (euxinic bottom water and lower influx of Mn oxides), both dissolved Mn(II) and dMn(III)-L are released, but Mn(II) is the dominant species. In contrast to dissolved Mn(II), the relatively stable and mobile dMn(III)-L may be transported from coastal areas into the open ocean. In addition, the biogeochemical processes leading to the formation and removal of dMn(III)-L in the sediment are strongly linked to the Fe cycle. Therefore, this study is very helpful in improving our understanding of sedimentary Mn and Fe cycling, especially the coupling between the two cycles, in coastal areas.
Overall, the manuscript is well written, however some of the figures could be improved (see specific comments). I would recommend the publication of this manuscript after some minor revisions.
Reply: We thank the reviewer for the positive words and comments and suggestions. Below, we indicate how we will address the comments in the revised manuscript.
Specific comments
Abstract
Line 11: It is better to write “depth profiles of water, pore-water and solid-phase data” instead of just “depth profiles”. Otherwise, it is not clear which comprehensive data set has been the basis for this study.
Reply: We will modify the text as suggested.
Methods
Figure 1a: The map of Lake Grevelingen including the overview map of the Netherlands in the top right corner is too small. It is very difficult to identify the water depths, especially in the Scharendijke basin.
Reply: We will modify the figure as suggested.
Results
Figure 3: In contrast to the water column profiles, the pore water profiles are relatively small. As some of the discussion relates specifically to the upper centimeters, it would be helpful if these were shown in more detail. In addition, the light grey dots and lines are hard to see.
Reply: A zoom of the upper 20 cm of the porewater profiles will be added to the supplements. We will change the color of the light grey dots and lines to a different color.
Figure 3h-j: If I see it correctly, the sum of dMn(III)-L and Mn(II) does not correspond to the TD Mn concentrations. Is there any explanation for this?
Reply: TD Mn is measured independently of dMn(III)-L and Mn(II), therefore these values can differ slightly. Furthermore, the kinetic Mn(II)/Mn(III) method that we use does not include the Mn(III) that is bound to strong ligands, which are included in the ICP-OES results for TD Mn (Oldham et al., 2015). We assume that the difference between dMn(III)-L + dMn(II) and the TD Mn measured by ICP-OES is the fraction of dMn(III)-L that is bound to strong ligands. In the supplements figures S2 and S3 it is visible that in March TD Mn and the sum of dMn(III)-L and Mn(II) overlap well, but that in September TD Mn exceeds the sum of dMn(III)-L and Mn(II) in many measurements, indicating that the strongly bound dMn(III)-L makes up for a larger part of the dMn(III)-L in September. We will revise the text to point this out.
Figure 4: Again, the light grey dots and lines are hard to see.
Reply: The light grey color in the figure will be changed to a different color.
Line 237 + Figure 4a: As the Corg content usually refers to the total sediment weight, it is better to speak of Corg contents rather than Corg concentrations.
Reply: We will modify the text as suggested.
Line 245-311: The description of the model results already contains many possible interpretations or approaches for discussion. It would be good if these were directly linked with the discussion part.
Reply: We agree that the points regarding the degassing and the seasonality in the solid phase profiles (line 246-251) could be interpreted as discussion. These lines will be removed from the results. The remainder of the section describes the model results and does not include interpretation of the model results. We will modify the text to clarify that we are presenting the outcome of the model.
Figure 5: At first glance, the figure is a little confusing due to many numbers/rates. One suggestion here would be to split the figure in two: (a) situation in March, (b) situation in September. In this way, the size of the arrows could be adjusted to the respective rates to better highlight the differences in the rates between March and September. In addition, the oxygen conditions in the bottom water could then be integrated for both situations.
Reply: We appreciate the suggestion of the reviewer but find that comparing the two seasons becomes difficult when the figure is split into two panels, because the outcome for the two seasons is then presented separate from each other. The seasonal contrast is a key aspect of this section of the discussion, therefore we prefer to keep the one panel figure. We will add the redox state of the bottom water in both seasons in the panel as suggested.
Discussion
Line 334: Again, it would be very good if the top 0 to 10 cm were shown in more detail to see the strong counter gradients.
Reply: Please, see our reply to the earlier comment on Figure 3. A zoom of the top 20 cm of the porewater profiles will be added to the supplements.
Line 394-395: At this point it should be described in more detail to what extent the transport of trace metals is decoupled from that of Mn when it is mainly present as dMn(III)-L. Were trace metals measured in the water column samples that could confirm the mentioned hypothesis?
Reply: We will expand the section of the text on the trace metals as suggested, referencing the relevant literature on this topic. Unfortunately, we do not have water column profiles of cobalt, nickel and zinc for March and September 2020 to include in this paper.
Technical corrections
Line 16: Since manganese (Mn) is introduced in line 11, oxygen should be introduced in the same way: Our model indicates that, when oxygen (O2) is present in the bottom water, there are three major sources of pore water dMn(III)-L (…).
Reply : The text will be modified as suggested.
Line 43: Again, please also introduce oxygen for consistency.
Reply : The text will be modified as suggested.
Line 307-308: Here the authors are certainly referring to Figure 8, as Figure 9 does not exist.
Reply : We will correct this in the text.
Line 370: The sentence structure is a bit awkward. It would be better: When bottom water O2 re-establish in October, the influx of Mn and Fe oxides, the rates of sedimentary Mn cycling and the benthic flux of Mn all increase.
Reply : The text will be modified as suggested.
Line 388: Instead of saying “when it meets O2”, it is more appropriate to write “when it is exposed to O2”.
Reply : The text will be modified as suggested.
Bibliography
Oldham, V. E., Owings, S. M., Jones, M. R., Tebo, B. M., & Luther, G. W.: Evidence for the presence of strong Mn(III)-binding ligands in the water column of the Chesapeake Bay. Mar. Chem., 171, 58–66. https://doi.org/10.1016/j.marchem.2015.02.008, 2015
Citation: https://doi.org/10.5194/egusphere-2024-1706-AC1
-
AC1: 'Reply on RC1', Robin Klomp, 02 Oct 2024
-
RC2: 'Comment on egusphere-2024-1706', Aubin Thibault de chanvalon, 30 Aug 2024
General comment
The manuscript written by Klomp et al. describes the manganese cycle in a coastal sediment with extremely high accumulation rates (> 10 cm yr-1) and overlaid by a seasonally anoxic saline water. The manuscript recycles part of the data published by Żygadłowska et al. (2023) with the addition of new data on manganese speciation (called Mn(II) and Mn(III)=L). The main originality of the manuscript is the use of dissolved manganese speciation measurements into a reactive transport model. Most of the kinetic parameters concerning Mn(III)=L are deduced from the model fit, which allows the author to discuss the reactivity of Mn(III)=L and its importance in manganese efflux from the sediment. Beside the clear importance of this topic, the real novelty of this approach for Mn speciation and the good quality of the dataset, it seems that in the discussion, the authors feel too confident about the model and the analytical results and avoid discussing the underlying hypothesis and limitations. Ultimately, it leads to a general overinterpretation of the data with many direct affirmative sentences not supported by detailed argumentation. Model results are taken as true while they rely on many cases of hypotheses hidden by the complexity of the model, preventing the reader to appreciate the model's limitations and thus its scope.
In particular, a) the model fit to Mn(III)=L is not properly discussed while it fits mainly to one unique Mn(III)=L measurement (March, 0-1 cm depth) and fails to fit the deeper part of the Mn(III)=L profile; b) the analytical demonstration of the true occurrence of Mn(III)=L is not detailed while caution have been published on this method since the Madison et al. (2011) paper (Kim et al., 2022) which requires a particular effort of clarity; c) it seems that most of the model output are not produced by the modelled chemical reaction but mainly result from the model input i. e. by the strong seasonality of the manganese oxides deposition rate chosen by the authors but not discussed; d) Important parts of the sedimentary Mn cycle are not discussed, neither mentioned, in particular the interaction with the nitrogen cycle and the role of adsorbed Mn2+. These reservations are detailed below.
Main reservations
- a) The model fit to Mn(III)=L and Mn(II) seems insufficient to deduce fluxes, production and constant rates with a high level of confidence. First, it seems very dangerous to base most of the paper interpretation only on one unique Mn(III)=L (March, 0-1 cm depth) measurement since contamination or analytical errors are always possible. Even is the data is validated, the sampling uncertainty on one point should obviously produce important uncertainties in the model results. For example, the authors suppose that the observed maximum is at 0.5 cm depth, while it could also occur at 0.2 cm, given the centimeters resolution of the sampling. Does such difference significantly change the model results ? Many additional sampling uncertainties described in the literature should prevent the author to stand most of their results on only one measurement (spatial heterogeneity from macro organism, erosion during sampling with a gravimetric core, loss of any fluffy layer on the top of the sediment, …). Second, the fit favors the Mn(III) maximum in the 0-1 cm depth layer, at the cost of a bad fit at depth. Could it be possible to ignore the high value at the top to favor a good fit at depth ? What would be the model result in this case ? Why do you not select these results?
- b) The identification of Mn(III)=L needs to be strengthened since skeptical points of view have been published about this method (Kim et al., 2022). The credibility of the competitive ligand exchange kinetic methods requires more information about the deconvoluting of the kinetic signal and the precise conditions of the essay. In particular, the kinetic of manganese complexation is very sensitive to the chlorite content during the measurement, as detailed in (Thibault de Chanvalon and Luther, 2019). The oxygen concentration during the essay is also critical and should be discussed (Kim et al., 2022). I recommend publishing as supplementary material some examples of the time series of Mn=porphyrin formation rate including the March 0-1 cm depth sample, together with detailed essay conditions (salinity, oxygen), the strategy developed to overcome the method known limitations and the profile of apparent rate constants obtained for Mn(II) and Mn(III)=L.
- c) Most of the model output is not produced by the modeled chemical reaction, but by the model input: it seems that the Mn-ox concentration in the settling particles varies from 9.6 µmol/g in winter to 0.4 µmol/g during euxinic condition. Such important forcing needs to be discussed in detail, along with the most important geochemical reaction constraining the system. In particular, 1 - Zygadlowska et al. 2023 measured suspended material concentration and demonstrated that the bottom Mn concentration in particles does not change so much between seasons; 2 - if important Mn oxide consumption in euxinic water is credible (e.g. Thibault De Chanvalon et al., 2023) why should it be the same for Mn carbonate ? I was expecting an increase of Mn-carbonate in this case as it occurs in the euxinic sediment and because primary production favors carbonate precipitation; 3 – there is no direct proof of H2S in September; 4 – why assuming that anoxic water is necessary euxinic while transitory period with dominance of dissolved manganese has been observed over months in similar environment (Shaw et al., 1994) ? 5 – the discussion should clearly underline that most of the seasonality is driven by settling particle composition, while it is currently suggested by the topic discussed in section 4.2 that the “sediment becomes depleted in Mn oxides” because of sediment efflux and OM oxidation. 6 - Nice oscillations for Mn-carbonate and Corg content in the sediment are reported and present phase opposition (maximum on one fit with the minimum of the other); is there any possibility to explain them because of geochemical reaction rather than because of input seasonality ? Something in the model switches the oscillation from phase opposition to in-phase oscillation below 65 cm depth, what it is ? 7 - The observed loss of 4 umol/g of Mn oxides between March and September in the top 5 cm sediment would require a sediment efflux of approximately 4 umol/g x 2 600 g/dm3 x (1-0.90) x 0.5 dm / 6 months = 87 umol/dm2/month = 290 umol/m2/d which approximately fits with the observed gradient and the calculated flux of 210 umol/m2/d in March. So, on one hand, the rapid calculation suggests biogeochemical processes strong enough to produce the observed MnO2 depletion between March and September. And on the other hand, the elaborated model requires very strong forcing in the Mn content input to fit the data. Is the rapid calculation I propose wrong ? why ? 8 - Why does the published model decrease the benthic efflux as soon as the water column becomes anoxic? I expected the opposite, the absence of oxygen should favor Mn efflux (since there is no more MnO2 precipitating in the oxygenated layer) until reactive MnO2 is consumed (which should take about 6 months). Why is it not modelled ? This counterintuitive result should be underlined and discussed. I also suggest comparing the sedimentary Mn efflux taking into account Mn(III)=L (figure 7) with those calculated without Mn(III)=L as probably done in Żygadłowska et al., 2023.
- d) Some known reactions important in the sedimentary Mn cycle are not modeled or discussed. The revised manuscript should explain and justify why it seems negligible in your site. For example, Mn(III)=L oxidation by nitrite studies (Luther et al., 1997; Luther et al., 2021; Karolewski et al., 2020); debates on Mn-annamox (Hulth et al., 1999; Thamdrup and Dalsgaard, 2000); or the role of adsorbed Mn2+ (Richard et al., 2013; van der Zee et al., 2001; Canfield et al., 1993).
Bibliography
Canfield, D. E., Thamdrup, B., and Hansen, J. W.: The anaerobic degradation of organic matter in Danish coastal sediments: Iron reduction, manganese reduction, and sulfate reduction, Geochim. Cosmochim. Acta, 57, 3867–3883, https://doi.org/10.1016/0016-7037(93)90340-3, 1993.
Hulth, S., Aller, R. C., and Gilbert, F.: Coupled anoxic nitrification/manganese reduction in marine sediments, Geochim. Cosmochim. Acta, 63, 49–66, https://doi.org/10.1016/S0016-7037(98)00285-3, 1999.
Karolewski, J. S., Sutherland, K. M., Hansel, C. M., and Wankel, S. D.: An isotopic study of abiotic nitrite oxidation by ligand-bound manganese (III), Geochim. Cosmochim. Acta, https://doi.org/10.1016/j.gca.2020.11.004, 2020.
Kim, B., Lingappa, U. F., Magyar, J., Monteverde, D., Valentine, J. S., Cho, J., and Fischer, W.: Challenges of Measuring Soluble Mn(III) Species in Natural Samples, Molecules, 27, 1661, https://doi.org/10.3390/molecules27051661, 2022.
Luther, G. W., Sundby, B., Lewis, B. L., Brendel, P. J., and Silverberg, N.: Interactions of manganese with the nitrogen cycle: alternative pathways to dinitrogen, Geochim. Cosmochim. Acta, 61, 4043–4052, 1997.
Luther Iii, G. W., Karolewski, J. S., Sutherland, K. M., Hansel, C. M., and Wankel, S. D.: The Abiotic Nitrite Oxidation by Ligand-Bound Manganese (III): The Chemical Mechanism, Aquat. Geochem., 27, 207–220, https://doi.org/10.1007/s10498-021-09396-0, 2021.
Richard, D., Sundby, B., and Mucci, A.: Kinetics of manganese adsorption, desorption, and oxidation in coastal marine sediments, Limnol. Oceanogr., 58, 987–996, https://doi.org/10.4319/lo.2013.58.3.0987, 2013.
Shaw, T. J., Sholkovitz, E. R., and Klinkhammer, G.: Redox dynamics in the Chesapeake Bay: The effect on sediment/water uranium exchange, Geochim. Cosmochim. Acta, 58, 2985–2995, https://doi.org/10.1016/0016-7037(94)90173-2, 1994.
Thamdrup, B. and Dalsgaard, T.: The fate of ammonium in anoxic manganese oxide-rich marine sediment, Geochim. Cosmochim. Acta, 64, 4157–4164, https://doi.org/10.1016/S0016-7037(00)00496-8, 2000.
Thibault de Chanvalon, A. and Luther, G. W.: Mn speciation at nanomolar concentrations with a porphyrin competitive ligand and UV–vis measurements, Talanta, 200, 15–21, https://doi.org/10.1016/j.talanta.2019.02.069, 2019.
Thibault De Chanvalon, A., Luther, G. W., Estes, E. R., Necker, J., Tebo, B. M., Su, J., and Cai, W.-J.: Influence of manganese cycling on alkalinity in the redox stratified water column of Chesapeake Bay, Biogeosciences, 20, 3053–3071, https://doi.org/10.5194/bg-20-3053-2023, 2023.
van der Zee, C., van Raaphorst, W., and Epping, E.: Absorbed Mn2+ and Mn redox cycling in Iberian continental margin sediments (northeast Atlantic Ocean), J. Mar. Res., 59, 133–166, https://doi.org/10.1357/002224001321237407, 2001.
Żygadłowska, O. M., Venetz, J., Klomp, R., Lenstra, W. K., Van Helmond, N. A. G. M., Röckmann, T., Wallenius, A. J., Martins, P. D., Veraart, A. J., Jetten, M. S. M., and Slomp, C. P.: Pathways of methane removal in the sediment and water column of a seasonally anoxic eutrophic marine basin, Front. Mar. Sci., 10, 1085728, https://doi.org/10.3389/fmars.2023.1085728, 2023.
Citation: https://doi.org/10.5194/egusphere-2024-1706-RC2 -
AC2: 'Reply on RC2', Robin Klomp, 02 Oct 2024
General comment
The manuscript written by Klomp et al. describes the manganese cycle in a coastal sediment with extremely high accumulation rates (> 10 cm yr-1) and overlaid by a seasonally anoxic saline water. The manuscript recycles part of the data published by Żygadłowska et al. (2023) with the addition of new data on manganese speciation (called Mn(II) and Mn(III)=L). The main originality of the manuscript is the use of dissolved manganese speciation measurements into a reactive transport model.
Most of the kinetic parameters concerning Mn(III)=L are deduced from the model fit, which allows the author to discuss the reactivity of Mn(III)=L and its importance in manganese efflux from the sediment. Beside the clear importance of this topic, the real novelty of this approach for Mn speciation and the good quality of the dataset, it seems that in the discussion, the authors feel too confident about the model and the analytical results and avoid discussing the underlying hypothesis and limitations. Ultimately, it leads to a general overinterpretation of the data with many direct affirmative sentences not supported by detailed argumentation. Model results are taken as true while they rely on many cases of hypotheses hidden by the complexity of the model, preventing the reader to appreciate the model's limitations and thus its scope.
In particular,
a) the model fit to Mn(III)=L is not properly discussed while it fits mainly to one unique Mn(III)=L measurement (March, 0-1 cm depth) and fails to fit the deeper part of the Mn(III)=L profile;
b) the analytical demonstration of the true occurrence of Mn(III)=L is not detailed while caution have been published on this method since the Madison et al. (2011) paper (Kim et al., 2022) which requires a particular effort of clarity;
c) it seems that most of the model output are not produced by the modelled chemical reaction but mainly result from the model input i. e. by the strong seasonality of the manganese oxides deposition rate chosen by the authors but not discussed;
d) Important parts of the sedimentary Mn cycle are not discussed, neither mentioned, in particular the interaction with the nitrogen cycle and the role of adsorbed Mn2+. These reservations are detailed below.
Reply: We thank the reviewer for the thorough assessment of our manuscript and for the positive words regarding the novelty of our approach and the quality of the data set. We have done our best to address all comments below and will revise the manuscript accordingly.
Regarding the points raised above, we note that the paper of Żygadłowska et al. (2023) focusses on pathways of methane removal at this site, with the Mn data provided only to allow an assessment of the role of Mn oxides as a potential electron acceptor in methane oxidation. We will expand the model sections to clarify the goals, the assumptions made, the limitations of the model and the boundary conditions. We will expand the section on the analysis of Mn(III)-L. We will also add a section discussing the potential for interactions with the nitrogen cycle and why we can exclude a major role for adsorbed Mn(II). Further details are provided below.
Main reservations
a) The model fit to Mn(III)=L and Mn(II) seems insufficient to deduce fluxes, production and constant rates with a high level of confidence. First, it seems very dangerous to base most of the paper interpretation only on one unique Mn(III)=L (March, 0-1 cm depth) measurement since contamination or analytical errors are always possible.
Reply: Indeed, the peak in Mn(III)-L near the sediment-surface in March is based on one observation only, and we will specifically note this in the revised text and discuss the related uncertainty. We will also indicate that this was the result of triplicate analyses and that the total dissolved Mn concentration, which was determined in an independent procedure, also showed a peak of a similar magnitude at this depth (visible in Figs. S2 and S3, which will be used to illustrate this). Importantly, a very sharp peak in Mn(III) is expected, because the major pathway to produce Mn(III) is oxidation of Mn(II) by O2. Since O2 is only present in the upper 0.6 cm of the sediment (Fig. 3), Mn(III) is expected to be mainly produced in the upper 0.6 cm. Our observations are also in line with other studies on Mn(III)-L in sediments showing a sharp peak of Mn(III)-L near the sediment-surface, recorded in only limited data points (e.g. Madison et al., 2013). Notably, the vertical redox zonation at our site is more compressed than in this previous work, explaining the sharper peak.
Even is the data is validated, the sampling uncertainty on one point should obviously produce important uncertainties in the model results. For example, the authors suppose that the observed maximum is at 0.5 cm depth, while it could also occur at 0.2 cm, given the centimeters resolution of the sampling. Does such difference significantly change the model results ? Many additional sampling uncertainties described in the literature should prevent the author to stand most of their results on only one measurement (spatial heterogeneity from macro organism, erosion during sampling with a gravimetric core, loss of any fluffy layer on the top of the sediment, …).
Reply: The goal of the modeling is to obtain insight in the main drivers of trends in sedimentary Mn cycling in a coastal system that is seasonally euxinic. We will expand the text to explain that limitations related to our sample resolution near the sediment surface do not alter the overall picture of Mn cycling at this location. We note that macrofauna are absent at this site and that our sampling method allows the fluffy layer on top of the sediment to be preserved. This will be added in the text.
Second, the fit favors the Mn(III) maximum in the 0-1 cm depth layer, at the cost of a bad fit at depth. Could it be possible to ignore the high value at the top to favor a good fit at depth ? What would be the model result in this case ? Why do you not select these results?
Reply: The processes that form the Mn(III) peak in the 0–1 cm depth layer are not coupled to those that control the Mn(III)-L profile deeper in the sediment. In the model, the presence of O2 predominantly controls the formation of Mn(III)-L in the surface sediment, while at depth dMn(III)-L is mainly formed via interactions of Mn oxides with Fe and H2S. Therefore, our good fit at the top does not directly impact the Mn model fit deeper in the sediment. We agree with the reviewer that we do not fully capture the Mn(III)-L profile at depth. This is likely because of an incomplete understanding of the processes that impact Mn(III)-L production and formation in anoxic/sulfidic sediments. We will add text to explain this in the revised manuscript.
b) The identification of Mn(III)=L needs to be strengthened since skeptical points of view have been published about this method (Kim et al., 2022). The credibility of the competitive ligand exchange kinetic methods requires more information about the deconvoluting of the kinetic signal and the precise conditions of the essay. In particular, the kinetic of manganese complexation is very sensitive to the chlorite content during the measurement, as detailed in (Thibault de Chanvalon and Luther, 2019). The oxygen concentration during the essay is also critical and should be discussed (Kim et al., 2022). I recommend publishing as supplementary material some examples of the time series of Mn=porphyrin formation rate including the March 0-1 cm depth sample, together with detailed essay conditions (salinity, oxygen), the strategy developed to overcome the method known limitations and the profile of apparent rate constants obtained for Mn(II) and Mn(III)=L.
Reply: To correct for the salt dependence of the kinetics of the Mn(II) – porphyrin reaction, the k-value for the reaction between Mn(II) and porphyrin was determined in triplicate in an aliquot of a sample, which was completely reduced by adding hydroxylamine for 24 hours, as proposed in Oldham et al. (2015). This will be mentioned in the revised manuscript. Graphs of the kinetic lines used to determine the k1 value will be added to the supplements of the paper. In addition to that, graphs of several kinetic curves of samples will be included in the supplements.
Regarding the oxygen concentrations: we followed the protocol of Madison et al. (2011) which allows for its presence during the measurement. During all measurements, the oxygen concentrations were the same, so an effect of O2 on the kinetic measurements would be similar in all analyses and therefore would not cause a difference between the measurements that could result in a drift in the outcome. Oxygen mainly affects Mn(III) bound to strong ligands (DFOB; Kim et al., 2022). It is already known that the method we use does not effectively target Mn(III) bound to strong ligands like DFOB (Madison et al., 2011; Oldham et al., 2015). Therefore, we do not expect our measurements to be affected by O2. We will explicitly mention in the manuscript that O2 was present during the measurements and that this affects the measurement of strong ligands but not those targeted here.
c) Most of the model output is not produced by the modeled chemical reaction, but by the model input: it seems that the Mn-ox concentration in the settling particles varies from 9.6 µmol/g in winter to 0.4 µmol/g during euxinic condition. Such important forcing needs to be discussed in detail, along with the most important geochemical reaction constraining the system. In particular, 1 - Zygadlowska et al. 2023 measured suspended material concentration and demonstrated that the bottom Mn concentration in particles does not change so much between seasons;
Reply: The model output is determined by the chemical reactions and the boundary conditions that are assumed. As indicated in the detailed model description in the supplement (section 1), the input flux of metal oxides was set by fitting the model profiles to the measured profiles. This is the common procedure used in diagenetic modeling (Berg et al., 2003; Reed et al., 2016; Van Cappellen and Wang, 2018) because settling rates of metal oxides from the water column are very difficult to determine accurately. Indeed, Żygadłowska et al. (2023) measured the concentrations of suspended particulate Mn in the water column at 43 m water depth (i.e. 2 meters above the sediment-water interface) but without the settling velocity of the particles, this cannot be translated to a Mn input flux. Moreover, there is likely also lateral transfer of particles near the sediment-water interface along the slopes of this relatively small basin. We will add this in the text.
To address the concerns of the reviewer we also performed a forward simulation of the model using a data set for a range of porewater components and sediment Mn oxide and Mn carbonate for 8 sampling events in 2021, capturing the period between March and October. While we do not have dMn(III)-L and Mn(II) data for 2021, we do have detailed profiles of NH4+, alkalinity, sulfate,H2S, dTFe, dTMn and, for 4 months, Mn oxides and Mn carbonates. The results are shown in figure 1 attached to this reply and illustrate that, without any adjustment, our model is able to capture the major trends in the sedimentary Mn cycle and key porewater constituents over a spring-summer-fall cycle. This gives confidence in the results of the model. The data-model comparison for 2021 (see figure 1 attached to this reply) also supports our finding for 2020 that variations in Mn oxide profiles are strongly impacted by the input of Mn oxides. If the decrease in Mn oxide concentrations in summer would be due to removal by chemical reactions only, an increase in dissolved Mn would be expected at the start of the euxinic conditions. We see, however, that when euxinia establishes in the bottom water (which in 2021 occurred in June; Zygadlowska et al., 2024b), total dissolved Mn concentrations were already decreasing. A short description of the fieldwork campaign performed in 2021 and the model results and data for 2021 will be added to the supplements.
If important Mn oxide consumption in euxinic water is credible (e. g. Thibault De Chanvalon et al., 2023) why should it be the same for Mn carbonate? I was expecting an increase of Mn-carbonate in this case as it occurs in the euxinic sediment and because primary production favors carbonate precipitation;
Reply: The variations in the profile of Mn carbonate in the model are assumed to be the combined result of variations in authigenic Mn carbonate formation and the input of Mn carbonate at the sediment-water interface. As described above for the Mn oxides, the input flux of Mn carbonates at the sediment-water interface was obtained by fitting the model to the data. Importantly, we could not fit the model to the data when assuming that authigenic Mn carbonate formation was responsible for all of the variation in the Mn carbonate profiles.
The relative roles of the two processes can be made visible by turning off authigenic Mn carbonate formation in the model (see figure 2 attached to this reply). Importantly, the observed oscillations are too large to be solely explained by authigenic Mn carbonate formation, as this would require a higher input of Mn oxides and higher concentrations of dissolved Mn(II) concentrations in the sediment than observed. The approach, results and uncertainties in the Mn carbonate modelling will be explained more extensively in the revised manuscript. We will also include the figure below in the supplements.
There is no direct proof of H2S in September; 4 – why assuming that anoxic water is necessary euxinic while transitory period with dominance of dissolved manganese has been observed over months in similar environment (Shaw et al., 1994)?
Reply: The measured bottom water concentration of H2S in September 2020 was 111 µmol L-1, i.e. there is direct proof of H2S. This value was indicated in the supplement (Table S7) and the data point is included in Figure 3 and the data file. We will emphasize this in the main text i.e. emphasize the proof that the bottom waters were euxinic in 2020. Furthermore, we can deduce the presence of H2S in the bottom water in September 2020 from the drawdown of Mo in the bottom waters, as described in Zygadlowska et al. (2023 and 2024a) and from the seasonal accumulation of Mo in the sediment as described by Egger et al. (2016). The presence of H2S in the bottom water in summer was also observed in 2021 as described in Zygadlowska et al. (2024b). We will update the reference and specifically mention this evidence for recurring euxinia in the revised manuscript.
The discussion should clearly underline that most of the seasonality is driven by settling particle composition, while it is currently suggested by the topic discussed in section 4.2 that the “sediment becomes depleted in Mn oxides” because of sediment efflux and OM oxidation.
Reply: We mention that the seasonality is mainly driven by the input of Mn oxides in section 4.2, lines 381 – 382: “The flux is highest and primarily consists of dMn(III)-L under oxic conditions in winter and spring, when the input of Mn oxides and recycling of Mn near the sediment-water interface is highest. We will place more emphasis on this seasonal variation driven by the variation in input in the revised manuscript.
Nice oscillations for Mn-carbonate and Corg content in the sediment are reported and present phase opposition (maximum on one fit with the minimum of the other); is there any possibility to explain them because of geochemical reaction rather than because of input seasonality ?
Reply: Please, see our previous reply regarding the oscillations in Mn carbonate formation and the figure above. We note that the degradation of organic matter in the sediment is strongly constrained by the ammonium profiles which show only modest variations between the different months in 2020 and 2021. This implies that we are capturing the seasonal changes in organic matter degradation. In this case, we can only model the variations in Corg burial by invoking a variation in input. Such a variation is in line with variations in primary productivity and organic matter supply from the North Sea known for this system (e.g. Hagens et al., 2015). We will clarify this in the text.
Something in the model switches the oscillation from phase opposition to in-phase oscillation below 65 cm depth, what it is ?
Reply: Indeed, there is such a switch at 65 cm depth. Because it is visible in the data, we also changed the forcing in the model by changing the boundary conditions for the input of Mn oxides, Corg and the sedimentation rate at this time in the simulation. We do not completely understand the mechanisms that drive this change in oscillation phase at depth, although we do see that this is a zone of increased sediment compaction. This switch does not change our findings regarding the Mn cycle, however, because the active cycling of Mn occurs mainly in the upper 25 cm of the sediment (see in the supplement Fig. S6) and the change referred to occurs below 65 cm.
The observed loss of 4 umol/g of Mn oxides between March and September in the top 5 cm sediment would require a sediment efflux of approximately 4 umol/g x 2 600 g/dm3 x (1-0.90) x 0.5 dm / 6 months = 87 umol/dm2/month = 290 umol/m2/d which approximately fits with the observed gradient and the calculated flux of 210 umol/m2/d in March. So, on one hand, the rapid calculation suggests biogeochemical processes strong enough to produce the observed MnO2 depletion between March and September. And on the other hand, the elaborated model requires very strong forcing in the Mn content input to fit the data. Is the rapid calculation I propose wrong ? why ?
Reply: We remind the reviewer of the exceptionally high sedimentation rate of around 20 cm per year at this site. This implies that the top 5 cm of the sediment has been deposited in approximately 3 months instead of the 6 months assumed in the calculation. This would mean a doubling of the flux to 580 µmol m-2 d-1, which no longer fits with any of the fluxes calculated from the data.
The rapid sedimentation rate also implies that sediment that was at the sediment-water interface in March is located at a depth of around 10 cm in September. The concentration of Mn oxide at the sediment surface in March is comparable to the concentration at 10 cm in September, both around 4.5 µmol g-1, which means that, even when the removal processes in March would be strong enough to remove 4 µmol g-1 Mn oxide, such strong removal is not observed in the data.
Why does the published model decrease the benthic efflux as soon as the water column becomes anoxic? I expected the opposite, the absence of oxygen should favor Mn efflux (since there is no more MnO2 precipitating in the oxygenated layer) until reactive MnO2 is consumed (which should take about 6 months). Why is it not modelled ? This counterintuitive result should be underlined and discussed.
Reply: The benthic flux of dissolved Mn decreases relatively fast upon the occurrence of bottom water euxinia, because the highly reactive Mn oxides in this system are removed very quickly. As a consequence, there is not much highly reactive Mn oxide remaining when the system becomes euxinic. Therefore, the increase in benthic flux of Mn due to the release of dissolved Mn from Mn oxides that are reduced when the system becomes euxinic is smaller than expected. This is confirmed by the forward modeling for 2021 (see figure 1 attached to this reply), where total dissolved Mn in the porewater decreases with a factor of around 4 from March to April and remains low throughout the euxinic period that lasted from June to September in that year. As indicated above, these data and the model fits will be included in the supplements.
I also suggest comparing the sedimentary Mn efflux taking into account Mn(III)=L (figure 7) with those calculated without Mn(III)=L as probably done in Żygadłowska et al., 2023.
Reply: As suggested by the reviewer, we calculated the benthic fluxes based on the porewater profiles assuming the diffusion coefficient for Mn(II). In March, the benthic flux calculated without correcting for the Mn(III)-L concentration is about 10x larger than the modelled flux where Mn(III)-L is taken into account. In September, the calculated benthic flux is about 3 times larger than the modelled flux. These calculations will be incorporated in the methods and discussion.
d) Some known reactions important in the sedimentary Mn cycle are not modeled or discussed. The revised manuscript should explain and justify why it seems negligible in your site. For example, Mn(III)=L oxidation by nitrite studies (Luther et al., 1997; Luther et al., 2021; Karolewski et al., 2020); debates on Mn-annamox (Hulth et al., 1999; Thamdrup and Dalsgaard, 2000); or the role of adsorbed Mn2+ (Richard et al., 2013; van der Zee et al., 2001; Canfield et al., 1993).
Reply: Explanations for the exclusion of the mentioned processes in the model will be included in the revised model description in the supplements. We have the following argumentation:
We measured nitrite concentrations in the upper 10cm of the sediment, both in March and September 2020 with a Gallery™ Automated Chemistry Analyzer type 861 (Thermo Fisher Scientific). In March, nitrite was absent from the porewater. In September, concentrations of nitrite never exceeded 0.7 µM and showed no trend with depth. Given these results, we do not expect an effect of interactions of nitrite with Mn cycling at this site. The nitrite data will be included in the supplements and we will explain why the processes mentioned were not included in the model.
The relevance of Mn-annamox in marine environments is still under debate, as pointed out in the two references given (Hulth et al., 1999; Thamdrup and Dalsgaard, 2000). Because we are not certain whether this process can take place, we did not include it in the reactive transport model. We will mention this in the model description.
The primary effect of sorption of dissolved Mn(II) on Mn cycling is the enhanced transport related to mixing of the sediment through bioturbation (Slomp et al., 1997). At sites with little or no bioturbation, as is the case at our study site (macrofauna were absent, see above), the impact of Mn(II) sorption will be very small. We will add this in the model description in the supplements.
Bibliography
Canfield, D. E., Thamdrup, B., and Hansen, J. W.: The anaerobic degradation of organic matter in Danish coastal sediments: Iron reduction, manganese reduction, and sulfate reduction, Geochim. Cosmochim. Acta, 57, 3867–3883, https://doi.org/10.1016/0016-7037(93)90340-3, 1993.
Egger, M., Lenstra, W., Jong, D., Meysman, F. J. R., Sapart, C. J., Van Der Veen, C., Röckmann, T., Gonzalez, S., & Slomp, C. P.: Rapid sediment accumulation results in high methane effluxes from coastal sediments. PLoS ONE, 11(8), 1–22. https://doi.org/10.1371/journal.pone.0161609 , 2016
Hagens, M., Slomp, C. P., Meysman, F. J. R. , Seitaj, D. , Harlay, J. , Borges, a V., and Middelburg, J. J.: Biogeochemical processes and buffering capacity concurrently affect acidification in a seasonally hypoxic coastal marine basin. Biogeosciences 12: 1561–1583, 2015.
Hulth, S., Aller, R. C., and Gilbert, F.: Coupled anoxic nitrification/manganese reduction in marine sediments, Geochim. Cosmochim. Acta, 63, 49–66, https://doi.org/10.1016/S0016-7037(98)00285-3, 1999.
Karolewski, J. S., Sutherland, K. M., Hansel, C. M., and Wankel, S. D.: An isotopic study of abiotic nitrite oxidation by ligand-bound manganese (III), Geochim. Cosmochim. Acta, https://doi.org/10.1016/j.gca.2020.11.004, 2020.
Kim, B., Lingappa, U. F., Magyar, J., Monteverde, D., Valentine, J. S., Cho, J., and Fischer, W.: Challenges of Measuring Soluble Mn(III) Species in Natural Samples, Molecules, 27, 1661, https://doi.org/10.3390/molecules27051661, 2022.
Luther, G. W., Sundby, B., Lewis, B. L., Brendel, P. J., and Silverberg, N.: Interactions of manganese with the nitrogen cycle: alternative pathways to dinitrogen, Geochim. Cosmochim. Acta, 61, 4043–4052, 1997.
Luther III, G. W., Karolewski, J. S., Sutherland, K. M., Hansel, C. M., and Wankel, S. D.: The Abiotic Nitrite Oxidation by Ligand-Bound Manganese (III): The Chemical Mechanism, Aquat. Geochem., 27, 207–220, https://doi.org/10.1007/s10498-021-09396-0, 2021.
Madison, A. S., Tebo, B. M., & Luther, G. W.: Simultaneous determination of soluble manganese(III), manganese(II) and total manganese in natural (pore)waters. Talanta, 84(2), 374–381. https://doi.org/10.1016/j.talanta.2011.01.025, 2011
Madison, A. S., Tebo, B. M., Mucci, A., Sundby, B., & Luther III, G. W.: Abundant Porewater Mn(III) Is a Major Component of the Sedimentary Redox System. Science, 341(August), 875–878. https://doi.org/10.5040/9780755621101.0007, 2013
Oldham, V. E., Owings, S. M., Jones, M. R., Tebo, B. M., & Luther, G. W.: Evidence for the presence of strong Mn(III)-binding ligands in the water column of the Chesapeake Bay. Mar. Chem., 171, 58–66. https://doi.org/10.1016/j.marchem.2015.02.008, 2015
Reed, D.C., Gustafsson, B.G. and Slomp, C.P.: "Shelf-to-basin iron shuttling enhances vivianite formation in deep Baltic Sea sediments." Earth and Planet. Sci. Lett. 434: 241-251, 2016.
Richard, D., Sundby, B., and Mucci, A.: Kinetics of manganese adsorption, desorption, and oxidation in coastal marine sediments, Limnol. Oceanogr., 58, 987–996, https://doi.org/10.4319/lo.2013.58.3.0987, 2013.
Shaw, T. J., Sholkovitz, E. R., and Klinkhammer, G.: Redox dynamics in the Chesapeake Bay: The effect on sediment/water uranium exchange, Geochim. Cosmochim. Acta, 58, 2985–2995, https://doi.org/10.1016/0016-7037(94)90173-2, 1994.
Slomp, C. P., Malschaert, J. F. P., Lohse, L., & Van Raaphorst, W.: Iron and manganese cycling in different sedimentary environments on the North Sea continental margin. Science, 17(9), 1083–1117, 1997.
Thamdrup, B. and Dalsgaard, T.: The fate of ammonium in anoxic manganese oxide-rich marine sediment, Geochim. Cosmochim. Acta, 64, 4157–4164, https://doi.org/10.1016/S0016-7037(00)00496-8, 2000.
Thibault de Chanvalon, A. and Luther, G. W.: Mn speciation at nanomolar concentrations with a porphyrin competitive ligand and UV–vis measurements, Talanta, 200, 15–21, https://doi.org/10.1016/j.talanta.2019.02.069, 2019.
Thibault De Chanvalon, A., Luther, G. W., Estes, E. R., Necker, J., Tebo, B. M., Su, J., and Cai, W.-J.: Influence of manganese cycling on alkalinity in the redox stratified water column of Chesapeake Bay, Biogeosciences, 20, 3053–3071, https://doi.org/10.5194/bg-20-3053-2023, 2023.
van der Zee, C., van Raaphorst, W., and Epping, E.: Absorbed Mn2+ and Mn redox cycling in Iberian continental margin sediments (northeast Atlantic Ocean), J. Mar. Res., 59, 133–166, https://doi.org/10.1357/002224001321237407, 2001.
Żygadłowska, O. M., Venetz, J., Klomp, R., Lenstra, W. K., Van Helmond, N. A. G. M., Röckmann, T., Wallenius, A. J., Martins, P. D., Veraart, A. J., Jetten, M. S. M., and Slomp, C. P.: Pathways of methane removal in the sediment and water column of a seasonally anoxic eutrophic marine basin, Front. Mar. Sci., 10, 1085728, https://doi.org/10.3389/fmars.2023.1085728, 2023.
Żygadlowska, O. M., van Helmond, N.A.G.M., Lenstra, W.K., Klomp, R., Accou, R., Puyk, R., Dickson, A.J., Jetten, M.S.M., and Slomp, C.P.: Seasonal euxinia in a coastal basin: Impact on sedimentary molybdenum enrichments and isotope signatures, Chem. Geol., https://doi.org/10.1016/j.chemgeo.2024.122430, 2024a
Żygadlowska, O. M., Venetz, J., Lenstra, W. K., van Helmond, N.A.G.M., Klomp, R., Röckmann, T., Veraart, A. J., Jetten, M. S. M., and Slomp, C. P.: Ebullition drives high methane emissions from a eutrophic coastal basin, Geochim. Cosmochim. Acta, https://doi.org/10.1016/j.gca.2024.08.028, 2024b.
Status: closed
-
RC1: 'Comment on egusphere-2024-1706', Anonymous Referee #1, 02 Aug 2024
Review Klomp et al.
General comments
Klomp et al. investigate seasonal differences in the sedimentary Mn cycling in a eutrophic coastal basin in the Netherlands. Their study provides a comprehensive data set and combines geochemical water column, pore-water and solid-phase data with reactive transport modeling.
The authors emphasize that dissolved Mn(III)-L is an important component of the Mn cycle throughout the year. Depending on the seasonal oxygen concentration in the bottom water and the influx of Mn oxides, Mn is predominantly released to the overlying water as dissolved Mn(III)-L in winter (oxic bottom water and higher influx of Mn oxides), whereas in summer (euxinic bottom water and lower influx of Mn oxides), both dissolved Mn(II) and dMn(III)-L are released, but Mn(II) is the dominant species. In contrast to dissolved Mn(II), the relatively stable and mobile dMn(III)-L may be transported from coastal areas into the open ocean. In addition, the biogeochemical processes leading to the formation and removal of dMn(III)-L in the sediment are strongly linked to the Fe cycle. Therefore, this study is very helpful in improving our understanding of sedimentary Mn and Fe cycling, especially the coupling between the two cycles, in coastal areas.
Overall, the manuscript is well written, however some of the figures could be improved (see specific comments). I would recommend the publication of this manuscript after some minor revisions.
Specific comments
Abstract
Line 11: It is better to write “depth profiles of water, pore-water and solid-phase data” instead of just “depth profiles”. Otherwise, it is not clear which comprehensive data set has been the basis for this study.
Methods
Figure 1a: The map of Lake Grevelingen including the overview map of the Netherlands in the top right corner is too small. It is very difficult to identify the water depths, especially in the Scharendijke basin.
Results
Figure 3: In contrast to the water column profiles, the pore water profiles are relatively small. As some of the discussion relates specifically to the upper centimeters, it would be helpful if these were shown in more detail. In addition, the light grey dots and lines are hard to see.
Figure 3h-j: If I see it correctly, the sum of dMn(III)-L and Mn(II) does not correspond to the TD Mn concentrations. Is there any explanation for this?
Figure 4: Again, the light grey dots and lines are hard to see.
Line 237 + Figure 4a: As the Corg content usually refers to the total sediment weight, it is better to speak of Corg contents rather than Corg concentrations.
Line 245-311: The description of the model results already contains many possible interpretations or approaches for discussion. It would be good if these were directly linked with the discussion part.
Figure 5: At first glance, the figure is a little confusing due to many numbers/rates. One suggestion here would be to split the figure in two: (a) situation in March, (b) situation in September. In this way, the size of the arrows could be adjusted to the respective rates to better highlight the differences in the rates between March and September. In addition, the oxygen conditions in the bottom water could then be integrated for both situations.
Discussion
Line 334: Again, it would be very good if the top 0 to 10 cm were shown in more detail to see the strong counter gradients.
Line 394-395: At this point it should be described in more detail to what extent the transport of trace metals is decoupled from that of Mn when it is mainly present as dMn(III)-L. Were trace metals measured in the water column samples that could confirm the mentioned hypothesis?
Technical corrections
Line 16: Since manganese (Mn) is introduced in line 11, oxygen should be introduced in the same way: Our model indicates that, when oxygen (O2) is present in the bottom water, there are three major sources of pore water dMn(III)-L (…).
Line 43: Again, please also introduce oxygen for consistency.
Line 307-308: Here the authors are certainly referring to Figure 8, as Figure 9 does not exist.
Line 370: The sentence structure is a bit awkward. It would be better: When bottom water O2 re-establish in October, the influx of Mn and Fe oxides, the rates of sedimentary Mn cycling and the benthic flux of Mn all increase.
Line 388: Instead of saying “when it meets O2”, it is more appropriate to write “when it is exposed to O2”.
Citation: https://doi.org/10.5194/egusphere-2024-1706-RC1 -
AC1: 'Reply on RC1', Robin Klomp, 02 Oct 2024
Klomp et al. investigate seasonal differences in the sedimentary Mn cycling in a eutrophic coastal basin in the Netherlands. Their study provides a comprehensive data set and combines geochemical water column, pore-water and solid-phase data with reactive transport modeling.
The authors emphasize that dissolved Mn(III)-L is an important component of the Mn cycle throughout the year. Depending on the seasonal oxygen concentration in the bottom water and the influx of Mn oxides, Mn is predominantly released to the overlying water as dissolved Mn(III)-L in winter (oxic bottom water and higher influx of Mn oxides), whereas in summer (euxinic bottom water and lower influx of Mn oxides), both dissolved Mn(II) and dMn(III)-L are released, but Mn(II) is the dominant species. In contrast to dissolved Mn(II), the relatively stable and mobile dMn(III)-L may be transported from coastal areas into the open ocean. In addition, the biogeochemical processes leading to the formation and removal of dMn(III)-L in the sediment are strongly linked to the Fe cycle. Therefore, this study is very helpful in improving our understanding of sedimentary Mn and Fe cycling, especially the coupling between the two cycles, in coastal areas.
Overall, the manuscript is well written, however some of the figures could be improved (see specific comments). I would recommend the publication of this manuscript after some minor revisions.
Reply: We thank the reviewer for the positive words and comments and suggestions. Below, we indicate how we will address the comments in the revised manuscript.
Specific comments
Abstract
Line 11: It is better to write “depth profiles of water, pore-water and solid-phase data” instead of just “depth profiles”. Otherwise, it is not clear which comprehensive data set has been the basis for this study.
Reply: We will modify the text as suggested.
Methods
Figure 1a: The map of Lake Grevelingen including the overview map of the Netherlands in the top right corner is too small. It is very difficult to identify the water depths, especially in the Scharendijke basin.
Reply: We will modify the figure as suggested.
Results
Figure 3: In contrast to the water column profiles, the pore water profiles are relatively small. As some of the discussion relates specifically to the upper centimeters, it would be helpful if these were shown in more detail. In addition, the light grey dots and lines are hard to see.
Reply: A zoom of the upper 20 cm of the porewater profiles will be added to the supplements. We will change the color of the light grey dots and lines to a different color.
Figure 3h-j: If I see it correctly, the sum of dMn(III)-L and Mn(II) does not correspond to the TD Mn concentrations. Is there any explanation for this?
Reply: TD Mn is measured independently of dMn(III)-L and Mn(II), therefore these values can differ slightly. Furthermore, the kinetic Mn(II)/Mn(III) method that we use does not include the Mn(III) that is bound to strong ligands, which are included in the ICP-OES results for TD Mn (Oldham et al., 2015). We assume that the difference between dMn(III)-L + dMn(II) and the TD Mn measured by ICP-OES is the fraction of dMn(III)-L that is bound to strong ligands. In the supplements figures S2 and S3 it is visible that in March TD Mn and the sum of dMn(III)-L and Mn(II) overlap well, but that in September TD Mn exceeds the sum of dMn(III)-L and Mn(II) in many measurements, indicating that the strongly bound dMn(III)-L makes up for a larger part of the dMn(III)-L in September. We will revise the text to point this out.
Figure 4: Again, the light grey dots and lines are hard to see.
Reply: The light grey color in the figure will be changed to a different color.
Line 237 + Figure 4a: As the Corg content usually refers to the total sediment weight, it is better to speak of Corg contents rather than Corg concentrations.
Reply: We will modify the text as suggested.
Line 245-311: The description of the model results already contains many possible interpretations or approaches for discussion. It would be good if these were directly linked with the discussion part.
Reply: We agree that the points regarding the degassing and the seasonality in the solid phase profiles (line 246-251) could be interpreted as discussion. These lines will be removed from the results. The remainder of the section describes the model results and does not include interpretation of the model results. We will modify the text to clarify that we are presenting the outcome of the model.
Figure 5: At first glance, the figure is a little confusing due to many numbers/rates. One suggestion here would be to split the figure in two: (a) situation in March, (b) situation in September. In this way, the size of the arrows could be adjusted to the respective rates to better highlight the differences in the rates between March and September. In addition, the oxygen conditions in the bottom water could then be integrated for both situations.
Reply: We appreciate the suggestion of the reviewer but find that comparing the two seasons becomes difficult when the figure is split into two panels, because the outcome for the two seasons is then presented separate from each other. The seasonal contrast is a key aspect of this section of the discussion, therefore we prefer to keep the one panel figure. We will add the redox state of the bottom water in both seasons in the panel as suggested.
Discussion
Line 334: Again, it would be very good if the top 0 to 10 cm were shown in more detail to see the strong counter gradients.
Reply: Please, see our reply to the earlier comment on Figure 3. A zoom of the top 20 cm of the porewater profiles will be added to the supplements.
Line 394-395: At this point it should be described in more detail to what extent the transport of trace metals is decoupled from that of Mn when it is mainly present as dMn(III)-L. Were trace metals measured in the water column samples that could confirm the mentioned hypothesis?
Reply: We will expand the section of the text on the trace metals as suggested, referencing the relevant literature on this topic. Unfortunately, we do not have water column profiles of cobalt, nickel and zinc for March and September 2020 to include in this paper.
Technical corrections
Line 16: Since manganese (Mn) is introduced in line 11, oxygen should be introduced in the same way: Our model indicates that, when oxygen (O2) is present in the bottom water, there are three major sources of pore water dMn(III)-L (…).
Reply : The text will be modified as suggested.
Line 43: Again, please also introduce oxygen for consistency.
Reply : The text will be modified as suggested.
Line 307-308: Here the authors are certainly referring to Figure 8, as Figure 9 does not exist.
Reply : We will correct this in the text.
Line 370: The sentence structure is a bit awkward. It would be better: When bottom water O2 re-establish in October, the influx of Mn and Fe oxides, the rates of sedimentary Mn cycling and the benthic flux of Mn all increase.
Reply : The text will be modified as suggested.
Line 388: Instead of saying “when it meets O2”, it is more appropriate to write “when it is exposed to O2”.
Reply : The text will be modified as suggested.
Bibliography
Oldham, V. E., Owings, S. M., Jones, M. R., Tebo, B. M., & Luther, G. W.: Evidence for the presence of strong Mn(III)-binding ligands in the water column of the Chesapeake Bay. Mar. Chem., 171, 58–66. https://doi.org/10.1016/j.marchem.2015.02.008, 2015
Citation: https://doi.org/10.5194/egusphere-2024-1706-AC1
-
AC1: 'Reply on RC1', Robin Klomp, 02 Oct 2024
-
RC2: 'Comment on egusphere-2024-1706', Aubin Thibault de chanvalon, 30 Aug 2024
General comment
The manuscript written by Klomp et al. describes the manganese cycle in a coastal sediment with extremely high accumulation rates (> 10 cm yr-1) and overlaid by a seasonally anoxic saline water. The manuscript recycles part of the data published by Żygadłowska et al. (2023) with the addition of new data on manganese speciation (called Mn(II) and Mn(III)=L). The main originality of the manuscript is the use of dissolved manganese speciation measurements into a reactive transport model. Most of the kinetic parameters concerning Mn(III)=L are deduced from the model fit, which allows the author to discuss the reactivity of Mn(III)=L and its importance in manganese efflux from the sediment. Beside the clear importance of this topic, the real novelty of this approach for Mn speciation and the good quality of the dataset, it seems that in the discussion, the authors feel too confident about the model and the analytical results and avoid discussing the underlying hypothesis and limitations. Ultimately, it leads to a general overinterpretation of the data with many direct affirmative sentences not supported by detailed argumentation. Model results are taken as true while they rely on many cases of hypotheses hidden by the complexity of the model, preventing the reader to appreciate the model's limitations and thus its scope.
In particular, a) the model fit to Mn(III)=L is not properly discussed while it fits mainly to one unique Mn(III)=L measurement (March, 0-1 cm depth) and fails to fit the deeper part of the Mn(III)=L profile; b) the analytical demonstration of the true occurrence of Mn(III)=L is not detailed while caution have been published on this method since the Madison et al. (2011) paper (Kim et al., 2022) which requires a particular effort of clarity; c) it seems that most of the model output are not produced by the modelled chemical reaction but mainly result from the model input i. e. by the strong seasonality of the manganese oxides deposition rate chosen by the authors but not discussed; d) Important parts of the sedimentary Mn cycle are not discussed, neither mentioned, in particular the interaction with the nitrogen cycle and the role of adsorbed Mn2+. These reservations are detailed below.
Main reservations
- a) The model fit to Mn(III)=L and Mn(II) seems insufficient to deduce fluxes, production and constant rates with a high level of confidence. First, it seems very dangerous to base most of the paper interpretation only on one unique Mn(III)=L (March, 0-1 cm depth) measurement since contamination or analytical errors are always possible. Even is the data is validated, the sampling uncertainty on one point should obviously produce important uncertainties in the model results. For example, the authors suppose that the observed maximum is at 0.5 cm depth, while it could also occur at 0.2 cm, given the centimeters resolution of the sampling. Does such difference significantly change the model results ? Many additional sampling uncertainties described in the literature should prevent the author to stand most of their results on only one measurement (spatial heterogeneity from macro organism, erosion during sampling with a gravimetric core, loss of any fluffy layer on the top of the sediment, …). Second, the fit favors the Mn(III) maximum in the 0-1 cm depth layer, at the cost of a bad fit at depth. Could it be possible to ignore the high value at the top to favor a good fit at depth ? What would be the model result in this case ? Why do you not select these results?
- b) The identification of Mn(III)=L needs to be strengthened since skeptical points of view have been published about this method (Kim et al., 2022). The credibility of the competitive ligand exchange kinetic methods requires more information about the deconvoluting of the kinetic signal and the precise conditions of the essay. In particular, the kinetic of manganese complexation is very sensitive to the chlorite content during the measurement, as detailed in (Thibault de Chanvalon and Luther, 2019). The oxygen concentration during the essay is also critical and should be discussed (Kim et al., 2022). I recommend publishing as supplementary material some examples of the time series of Mn=porphyrin formation rate including the March 0-1 cm depth sample, together with detailed essay conditions (salinity, oxygen), the strategy developed to overcome the method known limitations and the profile of apparent rate constants obtained for Mn(II) and Mn(III)=L.
- c) Most of the model output is not produced by the modeled chemical reaction, but by the model input: it seems that the Mn-ox concentration in the settling particles varies from 9.6 µmol/g in winter to 0.4 µmol/g during euxinic condition. Such important forcing needs to be discussed in detail, along with the most important geochemical reaction constraining the system. In particular, 1 - Zygadlowska et al. 2023 measured suspended material concentration and demonstrated that the bottom Mn concentration in particles does not change so much between seasons; 2 - if important Mn oxide consumption in euxinic water is credible (e.g. Thibault De Chanvalon et al., 2023) why should it be the same for Mn carbonate ? I was expecting an increase of Mn-carbonate in this case as it occurs in the euxinic sediment and because primary production favors carbonate precipitation; 3 – there is no direct proof of H2S in September; 4 – why assuming that anoxic water is necessary euxinic while transitory period with dominance of dissolved manganese has been observed over months in similar environment (Shaw et al., 1994) ? 5 – the discussion should clearly underline that most of the seasonality is driven by settling particle composition, while it is currently suggested by the topic discussed in section 4.2 that the “sediment becomes depleted in Mn oxides” because of sediment efflux and OM oxidation. 6 - Nice oscillations for Mn-carbonate and Corg content in the sediment are reported and present phase opposition (maximum on one fit with the minimum of the other); is there any possibility to explain them because of geochemical reaction rather than because of input seasonality ? Something in the model switches the oscillation from phase opposition to in-phase oscillation below 65 cm depth, what it is ? 7 - The observed loss of 4 umol/g of Mn oxides between March and September in the top 5 cm sediment would require a sediment efflux of approximately 4 umol/g x 2 600 g/dm3 x (1-0.90) x 0.5 dm / 6 months = 87 umol/dm2/month = 290 umol/m2/d which approximately fits with the observed gradient and the calculated flux of 210 umol/m2/d in March. So, on one hand, the rapid calculation suggests biogeochemical processes strong enough to produce the observed MnO2 depletion between March and September. And on the other hand, the elaborated model requires very strong forcing in the Mn content input to fit the data. Is the rapid calculation I propose wrong ? why ? 8 - Why does the published model decrease the benthic efflux as soon as the water column becomes anoxic? I expected the opposite, the absence of oxygen should favor Mn efflux (since there is no more MnO2 precipitating in the oxygenated layer) until reactive MnO2 is consumed (which should take about 6 months). Why is it not modelled ? This counterintuitive result should be underlined and discussed. I also suggest comparing the sedimentary Mn efflux taking into account Mn(III)=L (figure 7) with those calculated without Mn(III)=L as probably done in Żygadłowska et al., 2023.
- d) Some known reactions important in the sedimentary Mn cycle are not modeled or discussed. The revised manuscript should explain and justify why it seems negligible in your site. For example, Mn(III)=L oxidation by nitrite studies (Luther et al., 1997; Luther et al., 2021; Karolewski et al., 2020); debates on Mn-annamox (Hulth et al., 1999; Thamdrup and Dalsgaard, 2000); or the role of adsorbed Mn2+ (Richard et al., 2013; van der Zee et al., 2001; Canfield et al., 1993).
Bibliography
Canfield, D. E., Thamdrup, B., and Hansen, J. W.: The anaerobic degradation of organic matter in Danish coastal sediments: Iron reduction, manganese reduction, and sulfate reduction, Geochim. Cosmochim. Acta, 57, 3867–3883, https://doi.org/10.1016/0016-7037(93)90340-3, 1993.
Hulth, S., Aller, R. C., and Gilbert, F.: Coupled anoxic nitrification/manganese reduction in marine sediments, Geochim. Cosmochim. Acta, 63, 49–66, https://doi.org/10.1016/S0016-7037(98)00285-3, 1999.
Karolewski, J. S., Sutherland, K. M., Hansel, C. M., and Wankel, S. D.: An isotopic study of abiotic nitrite oxidation by ligand-bound manganese (III), Geochim. Cosmochim. Acta, https://doi.org/10.1016/j.gca.2020.11.004, 2020.
Kim, B., Lingappa, U. F., Magyar, J., Monteverde, D., Valentine, J. S., Cho, J., and Fischer, W.: Challenges of Measuring Soluble Mn(III) Species in Natural Samples, Molecules, 27, 1661, https://doi.org/10.3390/molecules27051661, 2022.
Luther, G. W., Sundby, B., Lewis, B. L., Brendel, P. J., and Silverberg, N.: Interactions of manganese with the nitrogen cycle: alternative pathways to dinitrogen, Geochim. Cosmochim. Acta, 61, 4043–4052, 1997.
Luther Iii, G. W., Karolewski, J. S., Sutherland, K. M., Hansel, C. M., and Wankel, S. D.: The Abiotic Nitrite Oxidation by Ligand-Bound Manganese (III): The Chemical Mechanism, Aquat. Geochem., 27, 207–220, https://doi.org/10.1007/s10498-021-09396-0, 2021.
Richard, D., Sundby, B., and Mucci, A.: Kinetics of manganese adsorption, desorption, and oxidation in coastal marine sediments, Limnol. Oceanogr., 58, 987–996, https://doi.org/10.4319/lo.2013.58.3.0987, 2013.
Shaw, T. J., Sholkovitz, E. R., and Klinkhammer, G.: Redox dynamics in the Chesapeake Bay: The effect on sediment/water uranium exchange, Geochim. Cosmochim. Acta, 58, 2985–2995, https://doi.org/10.1016/0016-7037(94)90173-2, 1994.
Thamdrup, B. and Dalsgaard, T.: The fate of ammonium in anoxic manganese oxide-rich marine sediment, Geochim. Cosmochim. Acta, 64, 4157–4164, https://doi.org/10.1016/S0016-7037(00)00496-8, 2000.
Thibault de Chanvalon, A. and Luther, G. W.: Mn speciation at nanomolar concentrations with a porphyrin competitive ligand and UV–vis measurements, Talanta, 200, 15–21, https://doi.org/10.1016/j.talanta.2019.02.069, 2019.
Thibault De Chanvalon, A., Luther, G. W., Estes, E. R., Necker, J., Tebo, B. M., Su, J., and Cai, W.-J.: Influence of manganese cycling on alkalinity in the redox stratified water column of Chesapeake Bay, Biogeosciences, 20, 3053–3071, https://doi.org/10.5194/bg-20-3053-2023, 2023.
van der Zee, C., van Raaphorst, W., and Epping, E.: Absorbed Mn2+ and Mn redox cycling in Iberian continental margin sediments (northeast Atlantic Ocean), J. Mar. Res., 59, 133–166, https://doi.org/10.1357/002224001321237407, 2001.
Żygadłowska, O. M., Venetz, J., Klomp, R., Lenstra, W. K., Van Helmond, N. A. G. M., Röckmann, T., Wallenius, A. J., Martins, P. D., Veraart, A. J., Jetten, M. S. M., and Slomp, C. P.: Pathways of methane removal in the sediment and water column of a seasonally anoxic eutrophic marine basin, Front. Mar. Sci., 10, 1085728, https://doi.org/10.3389/fmars.2023.1085728, 2023.
Citation: https://doi.org/10.5194/egusphere-2024-1706-RC2 -
AC2: 'Reply on RC2', Robin Klomp, 02 Oct 2024
General comment
The manuscript written by Klomp et al. describes the manganese cycle in a coastal sediment with extremely high accumulation rates (> 10 cm yr-1) and overlaid by a seasonally anoxic saline water. The manuscript recycles part of the data published by Żygadłowska et al. (2023) with the addition of new data on manganese speciation (called Mn(II) and Mn(III)=L). The main originality of the manuscript is the use of dissolved manganese speciation measurements into a reactive transport model.
Most of the kinetic parameters concerning Mn(III)=L are deduced from the model fit, which allows the author to discuss the reactivity of Mn(III)=L and its importance in manganese efflux from the sediment. Beside the clear importance of this topic, the real novelty of this approach for Mn speciation and the good quality of the dataset, it seems that in the discussion, the authors feel too confident about the model and the analytical results and avoid discussing the underlying hypothesis and limitations. Ultimately, it leads to a general overinterpretation of the data with many direct affirmative sentences not supported by detailed argumentation. Model results are taken as true while they rely on many cases of hypotheses hidden by the complexity of the model, preventing the reader to appreciate the model's limitations and thus its scope.
In particular,
a) the model fit to Mn(III)=L is not properly discussed while it fits mainly to one unique Mn(III)=L measurement (March, 0-1 cm depth) and fails to fit the deeper part of the Mn(III)=L profile;
b) the analytical demonstration of the true occurrence of Mn(III)=L is not detailed while caution have been published on this method since the Madison et al. (2011) paper (Kim et al., 2022) which requires a particular effort of clarity;
c) it seems that most of the model output are not produced by the modelled chemical reaction but mainly result from the model input i. e. by the strong seasonality of the manganese oxides deposition rate chosen by the authors but not discussed;
d) Important parts of the sedimentary Mn cycle are not discussed, neither mentioned, in particular the interaction with the nitrogen cycle and the role of adsorbed Mn2+. These reservations are detailed below.
Reply: We thank the reviewer for the thorough assessment of our manuscript and for the positive words regarding the novelty of our approach and the quality of the data set. We have done our best to address all comments below and will revise the manuscript accordingly.
Regarding the points raised above, we note that the paper of Żygadłowska et al. (2023) focusses on pathways of methane removal at this site, with the Mn data provided only to allow an assessment of the role of Mn oxides as a potential electron acceptor in methane oxidation. We will expand the model sections to clarify the goals, the assumptions made, the limitations of the model and the boundary conditions. We will expand the section on the analysis of Mn(III)-L. We will also add a section discussing the potential for interactions with the nitrogen cycle and why we can exclude a major role for adsorbed Mn(II). Further details are provided below.
Main reservations
a) The model fit to Mn(III)=L and Mn(II) seems insufficient to deduce fluxes, production and constant rates with a high level of confidence. First, it seems very dangerous to base most of the paper interpretation only on one unique Mn(III)=L (March, 0-1 cm depth) measurement since contamination or analytical errors are always possible.
Reply: Indeed, the peak in Mn(III)-L near the sediment-surface in March is based on one observation only, and we will specifically note this in the revised text and discuss the related uncertainty. We will also indicate that this was the result of triplicate analyses and that the total dissolved Mn concentration, which was determined in an independent procedure, also showed a peak of a similar magnitude at this depth (visible in Figs. S2 and S3, which will be used to illustrate this). Importantly, a very sharp peak in Mn(III) is expected, because the major pathway to produce Mn(III) is oxidation of Mn(II) by O2. Since O2 is only present in the upper 0.6 cm of the sediment (Fig. 3), Mn(III) is expected to be mainly produced in the upper 0.6 cm. Our observations are also in line with other studies on Mn(III)-L in sediments showing a sharp peak of Mn(III)-L near the sediment-surface, recorded in only limited data points (e.g. Madison et al., 2013). Notably, the vertical redox zonation at our site is more compressed than in this previous work, explaining the sharper peak.
Even is the data is validated, the sampling uncertainty on one point should obviously produce important uncertainties in the model results. For example, the authors suppose that the observed maximum is at 0.5 cm depth, while it could also occur at 0.2 cm, given the centimeters resolution of the sampling. Does such difference significantly change the model results ? Many additional sampling uncertainties described in the literature should prevent the author to stand most of their results on only one measurement (spatial heterogeneity from macro organism, erosion during sampling with a gravimetric core, loss of any fluffy layer on the top of the sediment, …).
Reply: The goal of the modeling is to obtain insight in the main drivers of trends in sedimentary Mn cycling in a coastal system that is seasonally euxinic. We will expand the text to explain that limitations related to our sample resolution near the sediment surface do not alter the overall picture of Mn cycling at this location. We note that macrofauna are absent at this site and that our sampling method allows the fluffy layer on top of the sediment to be preserved. This will be added in the text.
Second, the fit favors the Mn(III) maximum in the 0-1 cm depth layer, at the cost of a bad fit at depth. Could it be possible to ignore the high value at the top to favor a good fit at depth ? What would be the model result in this case ? Why do you not select these results?
Reply: The processes that form the Mn(III) peak in the 0–1 cm depth layer are not coupled to those that control the Mn(III)-L profile deeper in the sediment. In the model, the presence of O2 predominantly controls the formation of Mn(III)-L in the surface sediment, while at depth dMn(III)-L is mainly formed via interactions of Mn oxides with Fe and H2S. Therefore, our good fit at the top does not directly impact the Mn model fit deeper in the sediment. We agree with the reviewer that we do not fully capture the Mn(III)-L profile at depth. This is likely because of an incomplete understanding of the processes that impact Mn(III)-L production and formation in anoxic/sulfidic sediments. We will add text to explain this in the revised manuscript.
b) The identification of Mn(III)=L needs to be strengthened since skeptical points of view have been published about this method (Kim et al., 2022). The credibility of the competitive ligand exchange kinetic methods requires more information about the deconvoluting of the kinetic signal and the precise conditions of the essay. In particular, the kinetic of manganese complexation is very sensitive to the chlorite content during the measurement, as detailed in (Thibault de Chanvalon and Luther, 2019). The oxygen concentration during the essay is also critical and should be discussed (Kim et al., 2022). I recommend publishing as supplementary material some examples of the time series of Mn=porphyrin formation rate including the March 0-1 cm depth sample, together with detailed essay conditions (salinity, oxygen), the strategy developed to overcome the method known limitations and the profile of apparent rate constants obtained for Mn(II) and Mn(III)=L.
Reply: To correct for the salt dependence of the kinetics of the Mn(II) – porphyrin reaction, the k-value for the reaction between Mn(II) and porphyrin was determined in triplicate in an aliquot of a sample, which was completely reduced by adding hydroxylamine for 24 hours, as proposed in Oldham et al. (2015). This will be mentioned in the revised manuscript. Graphs of the kinetic lines used to determine the k1 value will be added to the supplements of the paper. In addition to that, graphs of several kinetic curves of samples will be included in the supplements.
Regarding the oxygen concentrations: we followed the protocol of Madison et al. (2011) which allows for its presence during the measurement. During all measurements, the oxygen concentrations were the same, so an effect of O2 on the kinetic measurements would be similar in all analyses and therefore would not cause a difference between the measurements that could result in a drift in the outcome. Oxygen mainly affects Mn(III) bound to strong ligands (DFOB; Kim et al., 2022). It is already known that the method we use does not effectively target Mn(III) bound to strong ligands like DFOB (Madison et al., 2011; Oldham et al., 2015). Therefore, we do not expect our measurements to be affected by O2. We will explicitly mention in the manuscript that O2 was present during the measurements and that this affects the measurement of strong ligands but not those targeted here.
c) Most of the model output is not produced by the modeled chemical reaction, but by the model input: it seems that the Mn-ox concentration in the settling particles varies from 9.6 µmol/g in winter to 0.4 µmol/g during euxinic condition. Such important forcing needs to be discussed in detail, along with the most important geochemical reaction constraining the system. In particular, 1 - Zygadlowska et al. 2023 measured suspended material concentration and demonstrated that the bottom Mn concentration in particles does not change so much between seasons;
Reply: The model output is determined by the chemical reactions and the boundary conditions that are assumed. As indicated in the detailed model description in the supplement (section 1), the input flux of metal oxides was set by fitting the model profiles to the measured profiles. This is the common procedure used in diagenetic modeling (Berg et al., 2003; Reed et al., 2016; Van Cappellen and Wang, 2018) because settling rates of metal oxides from the water column are very difficult to determine accurately. Indeed, Żygadłowska et al. (2023) measured the concentrations of suspended particulate Mn in the water column at 43 m water depth (i.e. 2 meters above the sediment-water interface) but without the settling velocity of the particles, this cannot be translated to a Mn input flux. Moreover, there is likely also lateral transfer of particles near the sediment-water interface along the slopes of this relatively small basin. We will add this in the text.
To address the concerns of the reviewer we also performed a forward simulation of the model using a data set for a range of porewater components and sediment Mn oxide and Mn carbonate for 8 sampling events in 2021, capturing the period between March and October. While we do not have dMn(III)-L and Mn(II) data for 2021, we do have detailed profiles of NH4+, alkalinity, sulfate,H2S, dTFe, dTMn and, for 4 months, Mn oxides and Mn carbonates. The results are shown in figure 1 attached to this reply and illustrate that, without any adjustment, our model is able to capture the major trends in the sedimentary Mn cycle and key porewater constituents over a spring-summer-fall cycle. This gives confidence in the results of the model. The data-model comparison for 2021 (see figure 1 attached to this reply) also supports our finding for 2020 that variations in Mn oxide profiles are strongly impacted by the input of Mn oxides. If the decrease in Mn oxide concentrations in summer would be due to removal by chemical reactions only, an increase in dissolved Mn would be expected at the start of the euxinic conditions. We see, however, that when euxinia establishes in the bottom water (which in 2021 occurred in June; Zygadlowska et al., 2024b), total dissolved Mn concentrations were already decreasing. A short description of the fieldwork campaign performed in 2021 and the model results and data for 2021 will be added to the supplements.
If important Mn oxide consumption in euxinic water is credible (e. g. Thibault De Chanvalon et al., 2023) why should it be the same for Mn carbonate? I was expecting an increase of Mn-carbonate in this case as it occurs in the euxinic sediment and because primary production favors carbonate precipitation;
Reply: The variations in the profile of Mn carbonate in the model are assumed to be the combined result of variations in authigenic Mn carbonate formation and the input of Mn carbonate at the sediment-water interface. As described above for the Mn oxides, the input flux of Mn carbonates at the sediment-water interface was obtained by fitting the model to the data. Importantly, we could not fit the model to the data when assuming that authigenic Mn carbonate formation was responsible for all of the variation in the Mn carbonate profiles.
The relative roles of the two processes can be made visible by turning off authigenic Mn carbonate formation in the model (see figure 2 attached to this reply). Importantly, the observed oscillations are too large to be solely explained by authigenic Mn carbonate formation, as this would require a higher input of Mn oxides and higher concentrations of dissolved Mn(II) concentrations in the sediment than observed. The approach, results and uncertainties in the Mn carbonate modelling will be explained more extensively in the revised manuscript. We will also include the figure below in the supplements.
There is no direct proof of H2S in September; 4 – why assuming that anoxic water is necessary euxinic while transitory period with dominance of dissolved manganese has been observed over months in similar environment (Shaw et al., 1994)?
Reply: The measured bottom water concentration of H2S in September 2020 was 111 µmol L-1, i.e. there is direct proof of H2S. This value was indicated in the supplement (Table S7) and the data point is included in Figure 3 and the data file. We will emphasize this in the main text i.e. emphasize the proof that the bottom waters were euxinic in 2020. Furthermore, we can deduce the presence of H2S in the bottom water in September 2020 from the drawdown of Mo in the bottom waters, as described in Zygadlowska et al. (2023 and 2024a) and from the seasonal accumulation of Mo in the sediment as described by Egger et al. (2016). The presence of H2S in the bottom water in summer was also observed in 2021 as described in Zygadlowska et al. (2024b). We will update the reference and specifically mention this evidence for recurring euxinia in the revised manuscript.
The discussion should clearly underline that most of the seasonality is driven by settling particle composition, while it is currently suggested by the topic discussed in section 4.2 that the “sediment becomes depleted in Mn oxides” because of sediment efflux and OM oxidation.
Reply: We mention that the seasonality is mainly driven by the input of Mn oxides in section 4.2, lines 381 – 382: “The flux is highest and primarily consists of dMn(III)-L under oxic conditions in winter and spring, when the input of Mn oxides and recycling of Mn near the sediment-water interface is highest. We will place more emphasis on this seasonal variation driven by the variation in input in the revised manuscript.
Nice oscillations for Mn-carbonate and Corg content in the sediment are reported and present phase opposition (maximum on one fit with the minimum of the other); is there any possibility to explain them because of geochemical reaction rather than because of input seasonality ?
Reply: Please, see our previous reply regarding the oscillations in Mn carbonate formation and the figure above. We note that the degradation of organic matter in the sediment is strongly constrained by the ammonium profiles which show only modest variations between the different months in 2020 and 2021. This implies that we are capturing the seasonal changes in organic matter degradation. In this case, we can only model the variations in Corg burial by invoking a variation in input. Such a variation is in line with variations in primary productivity and organic matter supply from the North Sea known for this system (e.g. Hagens et al., 2015). We will clarify this in the text.
Something in the model switches the oscillation from phase opposition to in-phase oscillation below 65 cm depth, what it is ?
Reply: Indeed, there is such a switch at 65 cm depth. Because it is visible in the data, we also changed the forcing in the model by changing the boundary conditions for the input of Mn oxides, Corg and the sedimentation rate at this time in the simulation. We do not completely understand the mechanisms that drive this change in oscillation phase at depth, although we do see that this is a zone of increased sediment compaction. This switch does not change our findings regarding the Mn cycle, however, because the active cycling of Mn occurs mainly in the upper 25 cm of the sediment (see in the supplement Fig. S6) and the change referred to occurs below 65 cm.
The observed loss of 4 umol/g of Mn oxides between March and September in the top 5 cm sediment would require a sediment efflux of approximately 4 umol/g x 2 600 g/dm3 x (1-0.90) x 0.5 dm / 6 months = 87 umol/dm2/month = 290 umol/m2/d which approximately fits with the observed gradient and the calculated flux of 210 umol/m2/d in March. So, on one hand, the rapid calculation suggests biogeochemical processes strong enough to produce the observed MnO2 depletion between March and September. And on the other hand, the elaborated model requires very strong forcing in the Mn content input to fit the data. Is the rapid calculation I propose wrong ? why ?
Reply: We remind the reviewer of the exceptionally high sedimentation rate of around 20 cm per year at this site. This implies that the top 5 cm of the sediment has been deposited in approximately 3 months instead of the 6 months assumed in the calculation. This would mean a doubling of the flux to 580 µmol m-2 d-1, which no longer fits with any of the fluxes calculated from the data.
The rapid sedimentation rate also implies that sediment that was at the sediment-water interface in March is located at a depth of around 10 cm in September. The concentration of Mn oxide at the sediment surface in March is comparable to the concentration at 10 cm in September, both around 4.5 µmol g-1, which means that, even when the removal processes in March would be strong enough to remove 4 µmol g-1 Mn oxide, such strong removal is not observed in the data.
Why does the published model decrease the benthic efflux as soon as the water column becomes anoxic? I expected the opposite, the absence of oxygen should favor Mn efflux (since there is no more MnO2 precipitating in the oxygenated layer) until reactive MnO2 is consumed (which should take about 6 months). Why is it not modelled ? This counterintuitive result should be underlined and discussed.
Reply: The benthic flux of dissolved Mn decreases relatively fast upon the occurrence of bottom water euxinia, because the highly reactive Mn oxides in this system are removed very quickly. As a consequence, there is not much highly reactive Mn oxide remaining when the system becomes euxinic. Therefore, the increase in benthic flux of Mn due to the release of dissolved Mn from Mn oxides that are reduced when the system becomes euxinic is smaller than expected. This is confirmed by the forward modeling for 2021 (see figure 1 attached to this reply), where total dissolved Mn in the porewater decreases with a factor of around 4 from March to April and remains low throughout the euxinic period that lasted from June to September in that year. As indicated above, these data and the model fits will be included in the supplements.
I also suggest comparing the sedimentary Mn efflux taking into account Mn(III)=L (figure 7) with those calculated without Mn(III)=L as probably done in Żygadłowska et al., 2023.
Reply: As suggested by the reviewer, we calculated the benthic fluxes based on the porewater profiles assuming the diffusion coefficient for Mn(II). In March, the benthic flux calculated without correcting for the Mn(III)-L concentration is about 10x larger than the modelled flux where Mn(III)-L is taken into account. In September, the calculated benthic flux is about 3 times larger than the modelled flux. These calculations will be incorporated in the methods and discussion.
d) Some known reactions important in the sedimentary Mn cycle are not modeled or discussed. The revised manuscript should explain and justify why it seems negligible in your site. For example, Mn(III)=L oxidation by nitrite studies (Luther et al., 1997; Luther et al., 2021; Karolewski et al., 2020); debates on Mn-annamox (Hulth et al., 1999; Thamdrup and Dalsgaard, 2000); or the role of adsorbed Mn2+ (Richard et al., 2013; van der Zee et al., 2001; Canfield et al., 1993).
Reply: Explanations for the exclusion of the mentioned processes in the model will be included in the revised model description in the supplements. We have the following argumentation:
We measured nitrite concentrations in the upper 10cm of the sediment, both in March and September 2020 with a Gallery™ Automated Chemistry Analyzer type 861 (Thermo Fisher Scientific). In March, nitrite was absent from the porewater. In September, concentrations of nitrite never exceeded 0.7 µM and showed no trend with depth. Given these results, we do not expect an effect of interactions of nitrite with Mn cycling at this site. The nitrite data will be included in the supplements and we will explain why the processes mentioned were not included in the model.
The relevance of Mn-annamox in marine environments is still under debate, as pointed out in the two references given (Hulth et al., 1999; Thamdrup and Dalsgaard, 2000). Because we are not certain whether this process can take place, we did not include it in the reactive transport model. We will mention this in the model description.
The primary effect of sorption of dissolved Mn(II) on Mn cycling is the enhanced transport related to mixing of the sediment through bioturbation (Slomp et al., 1997). At sites with little or no bioturbation, as is the case at our study site (macrofauna were absent, see above), the impact of Mn(II) sorption will be very small. We will add this in the model description in the supplements.
Bibliography
Canfield, D. E., Thamdrup, B., and Hansen, J. W.: The anaerobic degradation of organic matter in Danish coastal sediments: Iron reduction, manganese reduction, and sulfate reduction, Geochim. Cosmochim. Acta, 57, 3867–3883, https://doi.org/10.1016/0016-7037(93)90340-3, 1993.
Egger, M., Lenstra, W., Jong, D., Meysman, F. J. R., Sapart, C. J., Van Der Veen, C., Röckmann, T., Gonzalez, S., & Slomp, C. P.: Rapid sediment accumulation results in high methane effluxes from coastal sediments. PLoS ONE, 11(8), 1–22. https://doi.org/10.1371/journal.pone.0161609 , 2016
Hagens, M., Slomp, C. P., Meysman, F. J. R. , Seitaj, D. , Harlay, J. , Borges, a V., and Middelburg, J. J.: Biogeochemical processes and buffering capacity concurrently affect acidification in a seasonally hypoxic coastal marine basin. Biogeosciences 12: 1561–1583, 2015.
Hulth, S., Aller, R. C., and Gilbert, F.: Coupled anoxic nitrification/manganese reduction in marine sediments, Geochim. Cosmochim. Acta, 63, 49–66, https://doi.org/10.1016/S0016-7037(98)00285-3, 1999.
Karolewski, J. S., Sutherland, K. M., Hansel, C. M., and Wankel, S. D.: An isotopic study of abiotic nitrite oxidation by ligand-bound manganese (III), Geochim. Cosmochim. Acta, https://doi.org/10.1016/j.gca.2020.11.004, 2020.
Kim, B., Lingappa, U. F., Magyar, J., Monteverde, D., Valentine, J. S., Cho, J., and Fischer, W.: Challenges of Measuring Soluble Mn(III) Species in Natural Samples, Molecules, 27, 1661, https://doi.org/10.3390/molecules27051661, 2022.
Luther, G. W., Sundby, B., Lewis, B. L., Brendel, P. J., and Silverberg, N.: Interactions of manganese with the nitrogen cycle: alternative pathways to dinitrogen, Geochim. Cosmochim. Acta, 61, 4043–4052, 1997.
Luther III, G. W., Karolewski, J. S., Sutherland, K. M., Hansel, C. M., and Wankel, S. D.: The Abiotic Nitrite Oxidation by Ligand-Bound Manganese (III): The Chemical Mechanism, Aquat. Geochem., 27, 207–220, https://doi.org/10.1007/s10498-021-09396-0, 2021.
Madison, A. S., Tebo, B. M., & Luther, G. W.: Simultaneous determination of soluble manganese(III), manganese(II) and total manganese in natural (pore)waters. Talanta, 84(2), 374–381. https://doi.org/10.1016/j.talanta.2011.01.025, 2011
Madison, A. S., Tebo, B. M., Mucci, A., Sundby, B., & Luther III, G. W.: Abundant Porewater Mn(III) Is a Major Component of the Sedimentary Redox System. Science, 341(August), 875–878. https://doi.org/10.5040/9780755621101.0007, 2013
Oldham, V. E., Owings, S. M., Jones, M. R., Tebo, B. M., & Luther, G. W.: Evidence for the presence of strong Mn(III)-binding ligands in the water column of the Chesapeake Bay. Mar. Chem., 171, 58–66. https://doi.org/10.1016/j.marchem.2015.02.008, 2015
Reed, D.C., Gustafsson, B.G. and Slomp, C.P.: "Shelf-to-basin iron shuttling enhances vivianite formation in deep Baltic Sea sediments." Earth and Planet. Sci. Lett. 434: 241-251, 2016.
Richard, D., Sundby, B., and Mucci, A.: Kinetics of manganese adsorption, desorption, and oxidation in coastal marine sediments, Limnol. Oceanogr., 58, 987–996, https://doi.org/10.4319/lo.2013.58.3.0987, 2013.
Shaw, T. J., Sholkovitz, E. R., and Klinkhammer, G.: Redox dynamics in the Chesapeake Bay: The effect on sediment/water uranium exchange, Geochim. Cosmochim. Acta, 58, 2985–2995, https://doi.org/10.1016/0016-7037(94)90173-2, 1994.
Slomp, C. P., Malschaert, J. F. P., Lohse, L., & Van Raaphorst, W.: Iron and manganese cycling in different sedimentary environments on the North Sea continental margin. Science, 17(9), 1083–1117, 1997.
Thamdrup, B. and Dalsgaard, T.: The fate of ammonium in anoxic manganese oxide-rich marine sediment, Geochim. Cosmochim. Acta, 64, 4157–4164, https://doi.org/10.1016/S0016-7037(00)00496-8, 2000.
Thibault de Chanvalon, A. and Luther, G. W.: Mn speciation at nanomolar concentrations with a porphyrin competitive ligand and UV–vis measurements, Talanta, 200, 15–21, https://doi.org/10.1016/j.talanta.2019.02.069, 2019.
Thibault De Chanvalon, A., Luther, G. W., Estes, E. R., Necker, J., Tebo, B. M., Su, J., and Cai, W.-J.: Influence of manganese cycling on alkalinity in the redox stratified water column of Chesapeake Bay, Biogeosciences, 20, 3053–3071, https://doi.org/10.5194/bg-20-3053-2023, 2023.
van der Zee, C., van Raaphorst, W., and Epping, E.: Absorbed Mn2+ and Mn redox cycling in Iberian continental margin sediments (northeast Atlantic Ocean), J. Mar. Res., 59, 133–166, https://doi.org/10.1357/002224001321237407, 2001.
Żygadłowska, O. M., Venetz, J., Klomp, R., Lenstra, W. K., Van Helmond, N. A. G. M., Röckmann, T., Wallenius, A. J., Martins, P. D., Veraart, A. J., Jetten, M. S. M., and Slomp, C. P.: Pathways of methane removal in the sediment and water column of a seasonally anoxic eutrophic marine basin, Front. Mar. Sci., 10, 1085728, https://doi.org/10.3389/fmars.2023.1085728, 2023.
Żygadlowska, O. M., van Helmond, N.A.G.M., Lenstra, W.K., Klomp, R., Accou, R., Puyk, R., Dickson, A.J., Jetten, M.S.M., and Slomp, C.P.: Seasonal euxinia in a coastal basin: Impact on sedimentary molybdenum enrichments and isotope signatures, Chem. Geol., https://doi.org/10.1016/j.chemgeo.2024.122430, 2024a
Żygadlowska, O. M., Venetz, J., Lenstra, W. K., van Helmond, N.A.G.M., Klomp, R., Röckmann, T., Veraart, A. J., Jetten, M. S. M., and Slomp, C. P.: Ebullition drives high methane emissions from a eutrophic coastal basin, Geochim. Cosmochim. Acta, https://doi.org/10.1016/j.gca.2024.08.028, 2024b.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 363 | 111 | 222 | 696 | 28 | 12 | 15 |
- HTML: 363
- PDF: 111
- XML: 222
- Total: 696
- Supplement: 28
- BibTeX: 12
- EndNote: 15
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1