the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 24 Jun 2026
Cryogenic Fractionation and Thaw-Gradient Reorganization of Carbon, Nutrient and Trace Element Pools in Permafrost Peatlands
Abstract. Permafrost peatlands are highly vulnerable to climate warming because active layer deepening can mobilize previously frozen solutes into suprapermafrost flow paths and regional hydrological networks. Yet the composition of the frozen porewater reservoir, i.e., pore ice, and its role in regulating C, nutrient and trace-element release during thaw, remain poorly constrained. Here, we characterized dissolved organic matter (DOM), nutrients, and major and trace solutes across the active layer–permafrost interface in a continuous (> 90 % ice) polygonal peatland of the subarctic tundra. Four microtopes representing a thaw gradient (convex polygon → concave polygon→ peat transitional fen → peat-mineral fen) were sampled, with porewater and pore ice (< 0.45 µm) analyzed in active and frozen layers. Pore ice DOM showed a predominantly microbial and aliphatic signature, with elevated dissolved organic carbon (DOC) concentrations, indicating selective preservation of microbially processed, low-aromatic compounds during freezing. Freeze–thaw cycling homogenized active layer porewater chemistry, whereas pore ice retained stronger site-specific geochemical signatures. Overall, DOC, dissolved N and P, and exchangeable cations (Ca, Mn, Ba, Sr) preferentially accumulated in pore ice, while lithogenic elements (Zr, Ga, Hf, Ge, Nb, Cr, Y and REE) were more concentrated in active layer porewaters. Frozen horizons also exhibited lower C:N:P ratios than active layers, highlighting nutrient-enriched stoichiometry in the permafrost compartment. A coupled stoichiometric and decay-rate model suggests that phosphorus is largely consumed in situ during progressive thaw, thereby constraining further carbon and nitrogen processing, whereas substantial fractions of DOC and dissolved N remain available for lateral export. Viewed as a space-for-time thaw sequence, the transition from polygonal bog to fen indicates that permafrost degradation may initially enhance mobilization of labile carbon and nutrients before hydrological redistribution dominates. These results identify pore ice as both a cryogenic archive of past biogeochemical processing and a reactive reservoir capable of amplifying Arctic carbon–nutrient fluxes under continued warming.
- Preprint
(1185 KB) - Metadata XML
-
Supplement
(4145 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2026-3058', Anonymous Referee #1, 28 Jul 2026
-
RC2: 'Reply on RC1', Anonymous Referee #2, 28 Sep 2026
Overview
The manuscript “Cryogenic Fractionation and Thaw-Gradient Reorganization of Carbon, Nutrient and Trace Element Pools in Permafrost Peatlands” addresses a timely and important topic appropriate for publication in Biogeosciences. Four sediment cores (180 cm) from a permafrost peatland in Western Siberia were analyzed for dissolved organic matter, nutrients, and major and trace solutes, comparing frozen and active soil layers. The manuscript includes mass balance projections and a principal component analysis. The dataset and analysis are impressive and figures informative. The manuscript would benefit from a clearer presentation of the research questions, a more consistent separation (or combination) of results and discussion, shortening of some result sections, and simplification of figures. I recommend accepting the article after the following comments have been addressed.
Comments by line
Improve the graphical abstract by making a one row figure. Remove the two maps and replace with simplified version of Fig 1b (remove 6-11 in legend). Instead of the table for DOC and TDN add horizonal lines similar to Y-REE. Add a legend with all abbreviations below the figure. Consider subheadings for individual parts (e.g., permafrost peatland, sediment cores, …). Without reading the paper the climate change part is not very intuitive. Maybe have “present” and “2100” to show changes. Or remove to simplify.
L40 (185 ± 66 Pg C, (Hugelius et al., 2020) remove second (
L41-43 “One of the largest…” Seems a bit out of place. Would move to third paragraph and start with general trends and significance.
L45-46 Why is that relevant?
L56 Add space before “(“
Sometimes you abbreviate and sometimes you spell out WSL. Be consistent. Personally, I would spell out everywhere. Avoid unnecessary abbreviations (AL, FL etc.) when possible.
It would make the need for your study clearer if your introduction would include some specific open research questions leading up to your aims.
L90 What does Tz stand for? It would be more intuitive to use habitat types instead (e.g., concave bog, convex bog, transitional fen, peat–mineral fen, if that is what you mean) in text and in figures.
L95 All cores 180 cm long?
L97-112 Shorten this text. It is impossible to rule out within-site variability. Your results are valuable regardless of the limited core number, but variability should be considered in the discussion.
L134 How was pH measured?
Use past tense for the entire result section (expect projects, which should be future tense).
3.1 While well written, this section reads like a mixed results and discussion section, whereas other sections are pure results with a separate discussion section. Choose one style. Personally, I would combine all and shorten result details.
Sections 3.2, 3.3 and 3.5 could be shortened slightly to focus on overall trends and only separate between sites or depths if highly relevant.
L204 – 205 Report (average ± SD) after each parameter. Same for other parameters. This enables easy comparison. And/or add a table with all parameters (average ± SD, min – max) for all cores and overall in manuscript or SI.
Fig. 2 More distinct blue and green shades would enhance the readability. Slightly smaller and lightly transparent points as well. Put Fig 2f in SI and use an unform x-axis here to enable easy comparison. Add thin and light lines (same color as points) with lower active boundary layers of all cores in each plot with small text above each. Remove the general lines instead. No shaded background needed. Correct sub/superscripts in the caption. This applies for all similar figures.
L225 Avoid informal language and remove “As for C derived from organic acids”.
L261 There is a word missing. Do you mean: were significantly lower? What does this number stand for? Tz-4 = 7. Would calculate something like on average x-times/% lower ….Would remove p values everywhere and put full stats in a SI table.
Citation: https://doi.org/10.5194/egusphere-2026-3058-RC2 -
AC1: 'Reply on RC1', Lucia Perez-Serrano, 27 Oct 2026
We thank the reviewer for the positive assessment of our study and the constructive comments. We have revised the manuscript to improve the focus of the Discussion and clarify our interpretations, as detailed in the point-by-point responses below.
1. The concept of “selective preservation” of certain compounds during freezing or after thawing is not clear to me. It is quite often mentioned (for example L23, 287, 348, L517f) throughout the manuscript. The term would imply that certain properties of the analytes themselves cause different inclusion into the ice or preferential liberation after thawing while other analytes are left in the ice. However, differences between the chemical composition of active layer porewater and pore ice are also largely because especially labile compounds and limiting nutrients are not depleted while they are frozen, whereas biogeochemical cycling continues in the active layer. The finding that more aromatic DOC is found in the active layer than in frozen pore ice is therefore expected. I suggest to rethink this term of “selective preservation” or at least cite direct experimental evidence of it, if you think it is necessary to include it.
Regarding the concept of “selective preservation”, we thank the reviewer for this clarification. Our use of “selective” referred to the relative enrichment of low-aromatic, low-molecular-weight compounds within the dissolved carbon pool of frozen horizons. We agree that “preservation” is the more appropriate term, as “selective preservation” could imply a demonstrated mechanism of compound-specific incorporation into ice or release during thawing. We will therefore replace “selective preservation” with “preservation” throughout the manuscript. We will also clarify that the observed differences between active-layer porewater and pore ice can reflect reduced microbial consumption in frozen horizons, together with continued processing of labile organic compounds and nutrients in the active layer. These differences in turnover can contribute to the greater relative aromaticity of active-layer DOM. Accordingly, the revised text describes preservation in frozen horizons and presents differences in biological processing as one of the explanations for the observed compositional contrasts.
2. The first section of the results (Peatland formation and permafrost installation) does not fit in with the research aims in my opinion. Moreover, the section starts out with citing and discussing other literature when it should be about the results of this study. It is also not directly clear to me how the age of the peat and the peat accumulation can give direct evidence on freezing histories, especially since that information was not well connected to the other results later. If there is a connection, and the peat age over depth somehow informs the concentrations of analytes in the pore ice, then this could be mentioned in later sections with referral to the results of the peat age in the SI (as it is already). I therefore recommend to leave out this section but involve some arguments in the later discussion, if needed.
Concerning the first section of the results, we agree and we will shift essential part of his subsection from the Results to the Supplement, retaining a brief explanation for the most significant peat accumulation ages. The radiocarbon data and age–depth models in the Supplement (Table S1 and Fig. S1) will be complemented with former detailed chronology and regional comparisons. The relevant text on peat accumulation will be now briefly addressed in the Discussion, with clarification that peat ages do not directly establish the timing of permafrost formation. The age–depth models document contrasting peat accumulation histories among the four microtopes (Table S1, Fig. S1). These histories provide context for differences in the organic matrix hosting porewaters and pore ice and may contribute to variations in solute composition alongside contemporary hydrological conditions and biological processing. The radiocarbon measurements date the peat matrix and do not directly constrain pore-ice age or the timing of freezing.
3. It is not clear to me how the PCA was performed, since it is stated in L325 that a PCA was performed separately for active layer and frozen layer porewater data. But in that case, there should be two different PCA plots as the outcome. The explanation goes on with “for all samples combined…”, implying that there was also a PCA performed for all samples. It is therefore not clear for me what is displayed in Figure 6. I suggest to clarify this issue.
The PCA described in section 3.6 was fit for both active and frozen layers samples for all sites (Tz-4, Tz-3, Tz-2 and Tz-1) together. No other PCA involving separation for active or frozen layer samples was performed or included in this study. Thus, variance (PC1, PC2, PC3) describes a single PCA. Active and frozen layers labels were applied afterwards only to ease discussion. This will be clarified in section 3.6.
4. In general, some parts of the discussion read more like introduction and it is not always clear how the explanations are directly linked to the result of the manuscript (e.g. L493-496 or L497 to 512). I suggest to carefully revise those sections, always starting with the own results and afterwards including explanations with citations. In contrast, section 4.3 is written quite clearly
Sections L493-496 or L497 to 512 of the discussion will be updated as following: “Water residence time increases from transitional fen to polygons (Tz-2 < Tz-3 < Tz-4), promoting initial flushing followed by depletion of newly available solutes (Loiko et al., 2019). Even if DOC export and consumption variability are highly influenced by microhabitat (Table 1), as formerly observed for of C dynamics responses upon thawing in discontinuous permafrost peatlands in Western Siberia (Halaś et al., 2025), increased groundwater contribution may enhance transport of labile solutes in the hydrological continuums (Krickov et al., 2026). Overall, the gradual active layer deepening is likely to enhance DOC and TDN export while maintaining partial in situ retention of P. The convex polygon (Tz-4) represents a long-stabilized system with strong vertical compartmentalization and longer residence times, favoring in situ processing; the intermediate thaw stages (concave polygon Tz-3 and transitional fen Tz-2) pore ice reservoirs are strongly mobilized, potentially enhancing export pulses. The peat-mineral fen (Tz-1), characterized by greater hydrological connectivity, reflects a more advanced thaw state in which redistribution and flushing reduce compartmental contrasts and increase the likelihood of lateral solute transfer. Thus, the thaw chronosequence suggests a transient amplification of nutrient-rich export at intermediate stages, followed by stabilization or redistribution under advanced thaw conditions, ultimately controlled by adjacent variability inside each micro-habitat, attenuation of colloidal fraction by DOC-trace metals complexes adsorption on mineral horizons (Lim et al., 2022), and microbial adaptation to post-thaw conditions (Li et al., 2024).”
5. Data availability: the Zenodo repository code does not work yet. Please include a working hyperlink in the next version.
All data supporting this study are publicly available via the Zenodo repository at the hyperlink 10.5281/zenodo.20059630
L16: The use of “because” in that sentence is not logical. Permafrost peatlands are vulnerable to climate warming which can lead to active layer deepening, etc. I suggest to rearrange that sentence.
We propose the following sentence:
“Permafrost peatlands vulnerability to climate warming can result in active layer deepening and ultimately, the of release frozen solutes into suprapermafrost flow paths and regional hydrological networks."
L39: Is the amount of C stock that is mentioned relating to the amount of released C from permafrost peatlands or the amount of sequestered C? The first one would be more logical, but in the current way in which the sentence is written, it is not clear. I suggest to rearrange that sentence.The amount of C stock mentioned referred to the amount of sequestered C. We will revise the sentence indicating the expected amount of released C from permafrost peatlands, as following:
“As the Arctic has warmed nearly four times the rate of the global average the last four decades (Rantanen et al., 2022), destabilization of the sequestered C and N in permafrost-affected peatlands upon thawing (10-30 Pg C, 0.4-1.1 Pg N, Hugelius et al., 2020) through emissions from inland water surfaces and export to major river systems is likely to occur (Frey and Smith, 2005).”
L80: It is not clear to me what the “controls on pore ice formation” would be. It sounds like the authors would aim to study how easily ice forms in the different microhabitats. I suggest to clarify this aim.Regarding the “controls on pore ice formation”, we agree that “formation” was ambiguous and will be removed it from the stated aim. Our focus is on differences in the geochemical composition of pore ice among microhabitats and their implications for solute mobilization during thaw. Pore-ice chemistry is characterized using water extracted from thawed core intervals, without distinguishing individual ice structures or investigating the physical mechanisms of ice formation.
In response to this comment, we will revise the sentence at L80 as following: “Particular attention is given to differences in pore-ice composition among microtopographic settings and their implications for the storage and potential mobilization of carbon, nutrients and trace elements during thaw.”L116: Was this whole procedure done directly on-site or after transport to some kind of lab?
Regarding the soil-water ice sampling procedure, samples were cut using a ceramic blade into 5-cm and 10-cm layers for horizons deeper than 100 cm on-site, and thawed at room temperature in the field laboratory 2-3 h after sampling. We will add these details in the manuscript.
L240: The use of the word “contribution” implies that the fraction of organic acids to total DOC are displayed in the figure. However, the data is given in absolute concentrations of mg /L. The authors could consider either changing the wording here or referring to another figure in which the contribution is visible, or simply summarizing the contribution in the text.We propose the following change to highlight the C concentration from Organic Acids:
“Regarding the C pool, pore ice in peat-dominated microsites (convex polygon Tz-4 → concave polygon Tz-3 → peat transitional fen Tz-2) showed a significantly higher C concentration from organic acids compared to the frozen mineral layer of Tz-1 (peat-mineral fen) (Fig. 3a).”
L251: What are carboxylic compounds in this case and how were they measured? In case the authors mean organic acids, as shown in Fig S2, then I would stick with that term and not introduce a new one.Carboxylic acids is an equivalent term for organic acids (measured by high-performance ionic chromatography Dionex Ics-5000, uncertainty of 2 %). Previous references to term ‘carboxylic acids’ will be changed to ‘organic acids’ in the text.
L335f: The authors interpret the higher variation of frozen compared to active layer samples in the PCA as a greater variability in pore ice chemistry. However, in the next sentence an example is given that Tz-4 frozen samples show minimal variation along PC1 which shows lower variation within this site. Is that not a contradiction to the previous sentence?
Regarding the higher variation of frozen compared to active layer samples in the PCA explained as a greater variability in pore ice chemistry and Tz-4 examples, we agree that Tz-4 frozen samples show minimal variation along PC1, thus lower variation within this site.
Tz-4 shows an exceptional minimal variation along PC1 linked to lower influence of FTC cycles in convex polygons (mounds) during permafrost installation, since rapid freezing inhibited homogenization through repeated FTC cycles observed for frozen layers Tz-3 to Tz-1 samples. We will change the sentence as:
“Active layer samples show lower dispersion around their centroid (3.27 ± 2.69) compared to frozen samples (5.26 ± 1.98) (p < 0.001), suggesting greater compositional homogenization within active layer porewaters contrary to frozen samples, reflecting greater vertical and site-specific variability in pore ice chemistry (Fig. S8, Fig. 6). For example, most Tz-4 frozen samples show minimal variation along PC1, indicating compositional stability within this microsite.”
L358f: I do not necessarily agree with the statement that the different chemical signatures between active layer porewater and frozen pore ice are due to freeze-thaw cycles. The successive freezing and thawing should only affect the active layer and not the permanently frozen compartments?Regarding the different chemical signatures between active layer porewater and frozen pore ice linked to the freeze-thaw cycles, we agree that the successive freezing and thawing impact the contemporary active layer. Through the permafrost installation at the spatially asynchronous peat growth and freezing histories across the thaw gradient, ancient active layers vary in depth depending on the micro-habitat, thus influencing the geochemical composition of the pore ice by the impact of historical freeze-thaw cycles (Tz-4 older, long-stabilized frost mound vs Tz-3 to Tz-1 recent or ongoing peat formation and hydrological connectivity).
L361: Which compounds do you mean with “certain compound peaks”?Regarding the “certain compound peaks”, we refer to the absence of the DOC, TDN, and TDP peaks in Tz-4 frozen layer, in comparison to samples Tz-3 to Tz-1 (Fig. 2), in relation to the peat accumulation rates and seasonal freezing. We will add the aforementioned examples in the text:
“For example, Tz-4 peat accumulation rates and seasonal freezing may explain the absence of DOC, TDN, and TDP (Fig. 2), along with the compositional homogenization for trace elements in the frozen layer (Fig. 6), as polygonal mounds can lengthen solute residence time (less water loss and reduced hydraulic conductivity) and thus concentrate pore fluids (Raudina et al., 2017), […]”.
L412f: It is not directly clear how is this reference about the amplification of FTC on microbial N turnover is connected to the results of the manuscript. Please consider making the connection to the previous sentence clearer.We will clear the connection between the amplification of FTC on microbial N turnover to the results of the manuscript as in:
“Frozen horizons store a chemically labile and readily metabolizable DOM pool (FI > 1.4, higher concentrations of C from organic acids, and SUVA254 < 2), and rapid reactivation of the N cycling at the thaw front along with C mineralization processes (Fouché et al., 2020), enhanced by microbial metabolism driven by NH4+ pool consumption observed in simulated FTC (Yan et al., 2025) suggest a rapid biogeochemical reactivity.”
L414f: There is a lot of interpretation in this one sentence and I think it would be clearer for the potential readers of this manuscript if you would explain this step-by-step. First, you could start this new paragraph stating that you will talk about the bioavailability of P. Then, you could consider to mention that P can also be organic-bound, mainly by aromatic groups, and that thus the DOM composition affects the bioavailability of P. Finally, you can mention that the more aliphatic-dominated DOM in the pore ice likely leads to more bioavailable P upon thaw.We will update P bioavailability sentences as: “However, cycling of C and N is constrained by P bioavailability dependent on DOM composition, since organic-bound P binding with aromatics organic may inhibit mobilization (Hamdan et al., 2012). Thus, P in aliphatic-dominated pore ice fraction may be more bioavailable upon thawing.”
L525: What does “maximize the coupling” mean? I suggest to avoid such wordy terms.Concerning the “maximize the coupling” sentence, we agree and will replace “maximize the coupling” with “show a stronger association,” referring to the co-variation of DOM and exchangeable cation concentrations in pore ice.
Figures
L121, Fig. 1: Based on what reference did you identify the different ice forms? Or is this based on the sampled cores?The ice forms shown in Fig. 1b were identified from our field observations at the Tazovsky site over the past decade and from the examination of multiple cores collected for other investigations. These observations extend beyond the four cores analyzed in the present study. We have clarified the basis of this identification in the revised figure caption.
We will add to Fig. 1 caption the following: “The cryostructures illustrated in panel (b) are based on field observations at the Tazovsky site over the past decade, including examination of additional cores collected for other investigations.”
L221, Fig. 2: The given information on the active layer thickness in the caption is confusing to me. First, it is stated that the lower limit of the active layer thickness is from 35 to 65 cm, which corresponds to the active layer light blue band in the figure. Based on this, I would think that the maximum extent of the active layer is 35 to 65 cm across the cores, but we do not know which core has which extent. Then, different depths are given for the different cores in brackets, which I assumed were the maximum extents of active layer thickness. However, for Tz-1, it says 155 cm, which is well above the 65 cm given before. It would be helpful if the authors clarify that.We will clarify the active layer thickness representing the active layer depth at the time of sampling is represented for each site (Tz-4 = 40 ± 5 cm, Tz-3 = 60 ± 5, Tz-2 = 40 ± 5 cm and Tz-1 = 55 ± 5 cm).
(Please see supplement for the updated of the Figure 2). The corresponding text under the figure will be updated as:
"Figure 2. (Log)-profiles of dissolved (< 0.45 µm) (a) DOC, (b) TDN, (c) TDP, (d) C from Organic Acids (OA), (e) N-NH4+ and (f) P-PO43- (mg L-1) for porewater at the active layers and pore ice at the frozen layers. The lower limit of the active layer depth at the time of sampling is represented for each site (Tz-4 = 40 ± 5 cm, Tz-3 = 60 ± 5, Tz-2 = 40 ± 5 cm and Tz-1 = 55 ± 5 cm)."
Supplementary information
Table S4: Please consider adding short explanations on how the N and P limitation factors are calculated.
An explanation on N and P limitation factors will be added in the section “Note on the mass balance projections of C, N and P consumption and export under active layer deepening calculations” preceding the Table S4:
"Limitation factors are constrained by the amount of C that available N or P pools can support (Eq. 3).
Eq. 3 C_(cap,N (P)) = N (P)Total × C:N (P) consumption
Consumption of N and P fractions regarding the C pool is determined by ratios f_N and f_P (Eq. 4), limited at 1 (i.e., f_N = 1 implies a full mobilization of C pool carried by N sufficiency, whereas f_N = 0.2 only supports 20 % consumption of C pool).
Eq. 4 f_N (P) = min (1, (C(cap,N (P)) / C_Total)" -
AC2: 'Reply on RC2', Lucia Perez-Serrano, 27 Oct 2026
We thank the reviewer for the positive assessment of our study and the constructive recommendations. We have clarified the research questions, improved the separation of Results and Discussion, shortened relevant sections, and revised the figures for clarity, as detailed in our point-by-point responses below. We also have improved the graphical abstract.
Comments by lineImprove the graphical abstract by making a one row figure. Remove the two maps and replace with simplified version of Fig 1b (remove 6-11 in legend). Instead of the table for DOC and TDN add horizonal lines similar to Y-REE. Add a legend with all abbreviations below the figure. Consider subheadings for individual parts (e.g., permafrost peatland, sediment cores, …). Without reading the paper the climate change part is not very intuitive. Maybe have “present” and “2100” to show changes. Or remove to simplify.
Please see the updated abstract in the Supplement attached.
L41-43 “One of the largest…” Seems a bit out of place. Would move to third paragraph and start with general trends and significance.
L45-46 Why is that relevant?We will change the introduction considering the remarks pointed in L41-43 and L45-46. Regarding the 45-46 line, the statement “increase in the area of negative forms of relief (hollows and subsidences) compared to the positive ones (mounds and polygons) (Pastukhov et al., 2017)”, we agree that is not necessarily relevant.
“As the Arctic has warmed nearly four times the rate of the global average the last four decades (Rantanen et al., 2022), destabilization of the sequestered C and N in permafrost-affected peatlands upon thawing (10-30 Pg C, 0.4-1.1 Pg N, Hugelius et al., 2020) through emissions from inland water surfaces and export to major river systems is likely to occur (Frey and Smith, 2005). Precisely, climate warming may retreat the permafrost boundary to northern latitudes (Frey et al., 2007; Sherstyukov and Sherstyukov, 2015; Fewster et al., 2023), leading to a deepening of the active layer thickness (Åkerman and Johansson, 2008; Romanovsky et al., 2010). The active layer undergoes seasonal thawing-freezing and is responsible for water transport along the permafrost table (Woo, 2012; Liao and Zhuang, 2017; Koch et al., 2024), with suprapermafrost flow being the main source of incoming solutes to thermokarst lakes during summer periods, and a complementary source for Arctic rivers (Raudina et al., 2017a; Krickov et al., 2026).
Solutes from permafrost soils are rich in dissolved organic carbon (DOC), macro and micro-nutrients, and toxicants, especially if issued from the peat layer (Kuzmina et al., 2023). Peat organic soil is a highly complex porous medium which host variable ground ice structures in terms of size and time scale (Fu et al., 2022), and thawing of pore ice can result in the opening of isolated (non-connected) pores that contain solutes and bacteria (Morgalev et al., 2023). Besides, thawing of the permanent frozen layer may alter hydrological pathways, which will also come with a mobilization of unaccounted dissolved organic matter (DOM) pools, resulting in an enrichment in DOC, nutrients and metals of aquatic ecosystems (Vonk et al., 2015; O’Donnell et al., 2024), which can significantly affect C, N and P cycling.
Losses of peat C could have significant implications for aquatic biogeochemistry and ecosystems, particularly in the Western Siberia Lowland (WSL), which holds 1.05 million km2 permafrost-bearing soils, with an estimated peatland coverage of 592,440 km2 and a total C pool of 70.21 Pg C (Sheng et al., 2004), and a high vulnerability of the frozen peat linked to its average temperature slightly below 0 °C (Romanovsky et al., 2010; Raudina et al., 2025a).”
Sometimes you abbreviate and sometimes you spell out WSL. Be consistent. Personally, I would spell out everywhere. Avoid unnecessary abbreviations (AL, FL etc.) when possible.Regarding the spelling of WSL and abbreviations (AL, FL), we agree and will spell out “Western Siberia Lowland,” “active layer,” and “frozen layer” in the main text. We will maintain the AL (active layer) and FL (frozen layer) abbreviations only in Figures 3 and 4 to distinguish layers per site.
It would make the need for your study clearer if your introduction would include some specific open research questions leading up to your aims.Regarding the suggestion for the introduction to include some specific open research questions leading up to the aims of the study, we thank the reviewer for this suggestion, we will add three specific research questions immediately before the study aims. These address differences among microtopographic settings, contrasting distributions of solute groups, and the implications of nutrient stoichiometry for projected consumption and potential export during thaw:
“Three questions remain important for understanding how these reservoirs may respond to peatland thaw. First, are the differences in DOM composition and dissolved C:N ratios between active layer porewaters and pore ice consistent across polygon and fen microtopes? Second, do DOC, nutrients and major cations show similar patterns of accumulation in frozen horizons, and how do these patterns compare with those of lithogenic trace elements? Third, under a specified active layer deepening scenario, how does the measured C:N stoichiometry constrain projected in situ consumption and potential lateral export? These questions connect the chemical composition of the frozen reservoir with the potential fate of its solutes upon thaw.”
L90 What does Tz stand for? It would be more intuitive to use habitat types instead (e.g., concave bog, convex bog, transitional fen, peat–mineral fen, if that is what you mean) in text and in figures.Concerning the abbreviation “Tz”, we will clarify in Section 2.1 that “Tz” stands for Tazovsky, the study site, and explicitly linked each site code to its microhabitat. We retain Tz-1 to Tz-4 as concise identifiers of the sampled profiles, allowing consistent comparisons between active and frozen layers across the text, figures and tables. Habitat names are used when interpreting differences among microtopographic settings.
L95 All cores 180 cm long?Yes, all cores length is 180 cm. Details are found in section 2.1 (L88) and Figure 1d.
L97-112 Shorten this text. It is impossible to rule out within-site variability. Your results are valuable regardless of the limited core number, but variability should be considered in the discussion.The text will be shortened as following:
“One core was collected per microsite along the thaw gradient (polygonal bog, transitional fen, peat–mineral fen) since previous multi-core investigations at the Tazovsky site demonstrated relatively low small-scale spatial variability within comparable geomorphological units (Loiko et al., 2019, 2023). Across multiple mineral and peat soils in the region, reproducible depth-dependent hydrochemical profiles indicated coherent pedogenic structure and comparable solute migration regimes within each microtopographic class (Lim et al., 2025a), thus the selected cores are considered representative of their respective microsites in terms of vegetation, soil type, ground ice characteristics and active layer thickness.”
Variability will be now considered in the discussion sections.
L134 How was pH measured?The pH was measured within 1 hour after pore water extraction using Hanna pH-meter, with an uncertainty of ± 0.02 pH units. We will update the pH measurement details in section 2.1.
3.1 While well written, this section reads like a mixed results and discussion section, whereas other sections are pure results with a separate discussion section. Choose one style. Personally, I would combine all and shorten result details.
Regarding the sections 3.1, we will retain separate Results and Discussion sections and update the subsection on peatland formation and permafrost installation only with a brief explanation for the most significant peat accumulation ages. The relevant interpretation will be moved to the Discussion and the detailed chronology is presented in the Supplementary Information.
L204 – 205 Report (average ± SD) after each parameter. Same for other parameters. This enables easy comparison. And/or add a table with all parameters (average ± SD, min – max) for all cores and overall in manuscript or SI.We will update the Table S2 with the average ± SD, min – max and add the parameters in the text as following:
“The vertical pattern of the major nutrient distribution exhibited systematically lower average concentrations across all microsites at the active vs frozen layers (20–60 cm below the active layer boundary) for DOC (233.06 ± 91.15 vs 442.96 ± 281.59 mg L-1), TDN (8.76 ± 12.31 vs 44.71 ± 35.06 mg L-1) and TDP (0.23 ± 0.27 vs 1.023 ± 1.30 mg L-1) (Table S2).”
Fig. 2 More distinct blue and green shades would enhance the readability. Slightly smaller and lightly transparent points as well. Put Fig 2f in SI and use an unform x-axis here to enable easy comparison. Add thin and light lines (same color as points) with lower active boundary layers of all cores in each plot with small text above each. Remove the general lines instead. No shaded background needed. Correct sub/superscripts in the caption. This applies for all similar figures.Figure 2 (and similar figures) will be updated for the active layer thickness, clarifying the depth at the time of sampling is represented for each site (Tz-4 = 40 ± 5 cm, Tz-3 = 60 ± 5, Tz-2 = 40 ± 5 cm and Tz-1 = 55 ± 5 cm). Since the active layer depth is represented as a single band per site, and the former shaded background is removed, the blue and green shades for samples Tz-4 to Tz-1 are more easily readable.
Please see attached supplement for Figure 2 and Figure 5 (example of a similar figure) update.
L261: There is a word missing. Do you mean: were significantly lower? What does this number stand for? Tz-4 = 7. Would calculate something like on average x-times/% lower ….Would remove p values everywhere and put full stats in a SI table.
Average for active layer (53.25) and frozen layer (16.75) medians show that frozen layer median C:N ratios are 3.2 times lower. The sentence will be updated as following:
“Median C:N ratios were on average 3.2 lower in frozen layers compared to active layers at every site (Fig. 2, Table S3).”
-
AC1: 'Reply on RC1', Lucia Perez-Serrano, 27 Oct 2026
-
AC1: 'Reply on RC1', Lucia Perez-Serrano, 27 Oct 2026
We thank the reviewer for the positive assessment of our study and the constructive comments. We have revised the manuscript to improve the focus of the Discussion and clarify our interpretations, as detailed in the point-by-point responses below.
1. The concept of “selective preservation” of certain compounds during freezing or after thawing is not clear to me. It is quite often mentioned (for example L23, 287, 348, L517f) throughout the manuscript. The term would imply that certain properties of the analytes themselves cause different inclusion into the ice or preferential liberation after thawing while other analytes are left in the ice. However, differences between the chemical composition of active layer porewater and pore ice are also largely because especially labile compounds and limiting nutrients are not depleted while they are frozen, whereas biogeochemical cycling continues in the active layer. The finding that more aromatic DOC is found in the active layer than in frozen pore ice is therefore expected. I suggest to rethink this term of “selective preservation” or at least cite direct experimental evidence of it, if you think it is necessary to include it.
Regarding the concept of “selective preservation”, we thank the reviewer for this clarification. Our use of “selective” referred to the relative enrichment of low-aromatic, low-molecular-weight compounds within the dissolved carbon pool of frozen horizons. We agree that “preservation” is the more appropriate term, as “selective preservation” could imply a demonstrated mechanism of compound-specific incorporation into ice or release during thawing. We will therefore replace “selective preservation” with “preservation” throughout the manuscript. We will also clarify that the observed differences between active-layer porewater and pore ice can reflect reduced microbial consumption in frozen horizons, together with continued processing of labile organic compounds and nutrients in the active layer. These differences in turnover can contribute to the greater relative aromaticity of active-layer DOM. Accordingly, the revised text describes preservation in frozen horizons and presents differences in biological processing as one of the explanations for the observed compositional contrasts.
2. The first section of the results (Peatland formation and permafrost installation) does not fit in with the research aims in my opinion. Moreover, the section starts out with citing and discussing other literature when it should be about the results of this study. It is also not directly clear to me how the age of the peat and the peat accumulation can give direct evidence on freezing histories, especially since that information was not well connected to the other results later. If there is a connection, and the peat age over depth somehow informs the concentrations of analytes in the pore ice, then this could be mentioned in later sections with referral to the results of the peat age in the SI (as it is already). I therefore recommend to leave out this section but involve some arguments in the later discussion, if needed.
Concerning the first section of the results, we agree and we will shift essential part of his subsection from the Results to the Supplement, retaining a brief explanation for the most significant peat accumulation ages. The radiocarbon data and age–depth models in the Supplement (Table S1 and Fig. S1) will be complemented with former detailed chronology and regional comparisons. The relevant text on peat accumulation will be now briefly addressed in the Discussion, with clarification that peat ages do not directly establish the timing of permafrost formation. The age–depth models document contrasting peat accumulation histories among the four microtopes (Table S1, Fig. S1). These histories provide context for differences in the organic matrix hosting porewaters and pore ice and may contribute to variations in solute composition alongside contemporary hydrological conditions and biological processing. The radiocarbon measurements date the peat matrix and do not directly constrain pore-ice age or the timing of freezing.
3. It is not clear to me how the PCA was performed, since it is stated in L325 that a PCA was performed separately for active layer and frozen layer porewater data. But in that case, there should be two different PCA plots as the outcome. The explanation goes on with “for all samples combined…”, implying that there was also a PCA performed for all samples. It is therefore not clear for me what is displayed in Figure 6. I suggest to clarify this issue.
The PCA described in section 3.6 was fit for both active and frozen layers samples for all sites (Tz-4, Tz-3, Tz-2 and Tz-1) together. No other PCA involving separation for active or frozen layer samples was performed or included in this study. Thus, variance (PC1, PC2, PC3) describes a single PCA. Active and frozen layers labels were applied afterwards only to ease discussion. This will be clarified in section 3.6.
4. In general, some parts of the discussion read more like introduction and it is not always clear how the explanations are directly linked to the result of the manuscript (e.g. L493-496 or L497 to 512). I suggest to carefully revise those sections, always starting with the own results and afterwards including explanations with citations. In contrast, section 4.3 is written quite clearly
Sections L493-496 or L497 to 512 of the discussion will be updated as following: “Water residence time increases from transitional fen to polygons (Tz-2 < Tz-3 < Tz-4), promoting initial flushing followed by depletion of newly available solutes (Loiko et al., 2019). Even if DOC export and consumption variability are highly influenced by microhabitat (Table 1), as formerly observed for of C dynamics responses upon thawing in discontinuous permafrost peatlands in Western Siberia (Halaś et al., 2025), increased groundwater contribution may enhance transport of labile solutes in the hydrological continuums (Krickov et al., 2026). Overall, the gradual active layer deepening is likely to enhance DOC and TDN export while maintaining partial in situ retention of P. The convex polygon (Tz-4) represents a long-stabilized system with strong vertical compartmentalization and longer residence times, favoring in situ processing; the intermediate thaw stages (concave polygon Tz-3 and transitional fen Tz-2) pore ice reservoirs are strongly mobilized, potentially enhancing export pulses. The peat-mineral fen (Tz-1), characterized by greater hydrological connectivity, reflects a more advanced thaw state in which redistribution and flushing reduce compartmental contrasts and increase the likelihood of lateral solute transfer. Thus, the thaw chronosequence suggests a transient amplification of nutrient-rich export at intermediate stages, followed by stabilization or redistribution under advanced thaw conditions, ultimately controlled by adjacent variability inside each micro-habitat, attenuation of colloidal fraction by DOC-trace metals complexes adsorption on mineral horizons (Lim et al., 2022), and microbial adaptation to post-thaw conditions (Li et al., 2024).”
5. Data availability: the Zenodo repository code does not work yet. Please include a working hyperlink in the next version.
All data supporting this study are publicly available via the Zenodo repository at the hyperlink 10.5281/zenodo.20059630
L16: The use of “because” in that sentence is not logical. Permafrost peatlands are vulnerable to climate warming which can lead to active layer deepening, etc. I suggest to rearrange that sentence.
We propose the following sentence:
“Permafrost peatlands vulnerability to climate warming can result in active layer deepening and ultimately, the of release frozen solutes into suprapermafrost flow paths and regional hydrological networks."
L39: Is the amount of C stock that is mentioned relating to the amount of released C from permafrost peatlands or the amount of sequestered C? The first one would be more logical, but in the current way in which the sentence is written, it is not clear. I suggest to rearrange that sentence.The amount of C stock mentioned referred to the amount of sequestered C. We will revise the sentence indicating the expected amount of released C from permafrost peatlands, as following:
“As the Arctic has warmed nearly four times the rate of the global average the last four decades (Rantanen et al., 2022), destabilization of the sequestered C and N in permafrost-affected peatlands upon thawing (10-30 Pg C, 0.4-1.1 Pg N, Hugelius et al., 2020) through emissions from inland water surfaces and export to major river systems is likely to occur (Frey and Smith, 2005).”
L80: It is not clear to me what the “controls on pore ice formation” would be. It sounds like the authors would aim to study how easily ice forms in the different microhabitats. I suggest to clarify this aim.Regarding the “controls on pore ice formation”, we agree that “formation” was ambiguous and will be removed it from the stated aim. Our focus is on differences in the geochemical composition of pore ice among microhabitats and their implications for solute mobilization during thaw. Pore-ice chemistry is characterized using water extracted from thawed core intervals, without distinguishing individual ice structures or investigating the physical mechanisms of ice formation.
In response to this comment, we will revise the sentence at L80 as following: “Particular attention is given to differences in pore-ice composition among microtopographic settings and their implications for the storage and potential mobilization of carbon, nutrients and trace elements during thaw.”L116: Was this whole procedure done directly on-site or after transport to some kind of lab?
Regarding the soil-water ice sampling procedure, samples were cut using a ceramic blade into 5-cm and 10-cm layers for horizons deeper than 100 cm on-site, and thawed at room temperature in the field laboratory 2-3 h after sampling. We will add these details in the manuscript.
L240: The use of the word “contribution” implies that the fraction of organic acids to total DOC are displayed in the figure. However, the data is given in absolute concentrations of mg /L. The authors could consider either changing the wording here or referring to another figure in which the contribution is visible, or simply summarizing the contribution in the text.We propose the following change to highlight the C concentration from Organic Acids:
“Regarding the C pool, pore ice in peat-dominated microsites (convex polygon Tz-4 → concave polygon Tz-3 → peat transitional fen Tz-2) showed a significantly higher C concentration from organic acids compared to the frozen mineral layer of Tz-1 (peat-mineral fen) (Fig. 3a).”
L251: What are carboxylic compounds in this case and how were they measured? In case the authors mean organic acids, as shown in Fig S2, then I would stick with that term and not introduce a new one.Carboxylic acids is an equivalent term for organic acids (measured by high-performance ionic chromatography Dionex Ics-5000, uncertainty of 2 %). Previous references to term ‘carboxylic acids’ will be changed to ‘organic acids’ in the text.
L335f: The authors interpret the higher variation of frozen compared to active layer samples in the PCA as a greater variability in pore ice chemistry. However, in the next sentence an example is given that Tz-4 frozen samples show minimal variation along PC1 which shows lower variation within this site. Is that not a contradiction to the previous sentence?
Regarding the higher variation of frozen compared to active layer samples in the PCA explained as a greater variability in pore ice chemistry and Tz-4 examples, we agree that Tz-4 frozen samples show minimal variation along PC1, thus lower variation within this site.
Tz-4 shows an exceptional minimal variation along PC1 linked to lower influence of FTC cycles in convex polygons (mounds) during permafrost installation, since rapid freezing inhibited homogenization through repeated FTC cycles observed for frozen layers Tz-3 to Tz-1 samples. We will change the sentence as:
“Active layer samples show lower dispersion around their centroid (3.27 ± 2.69) compared to frozen samples (5.26 ± 1.98) (p < 0.001), suggesting greater compositional homogenization within active layer porewaters contrary to frozen samples, reflecting greater vertical and site-specific variability in pore ice chemistry (Fig. S8, Fig. 6). For example, most Tz-4 frozen samples show minimal variation along PC1, indicating compositional stability within this microsite.”
L358f: I do not necessarily agree with the statement that the different chemical signatures between active layer porewater and frozen pore ice are due to freeze-thaw cycles. The successive freezing and thawing should only affect the active layer and not the permanently frozen compartments?Regarding the different chemical signatures between active layer porewater and frozen pore ice linked to the freeze-thaw cycles, we agree that the successive freezing and thawing impact the contemporary active layer. Through the permafrost installation at the spatially asynchronous peat growth and freezing histories across the thaw gradient, ancient active layers vary in depth depending on the micro-habitat, thus influencing the geochemical composition of the pore ice by the impact of historical freeze-thaw cycles (Tz-4 older, long-stabilized frost mound vs Tz-3 to Tz-1 recent or ongoing peat formation and hydrological connectivity).
L361: Which compounds do you mean with “certain compound peaks”?Regarding the “certain compound peaks”, we refer to the absence of the DOC, TDN, and TDP peaks in Tz-4 frozen layer, in comparison to samples Tz-3 to Tz-1 (Fig. 2), in relation to the peat accumulation rates and seasonal freezing. We will add the aforementioned examples in the text:
“For example, Tz-4 peat accumulation rates and seasonal freezing may explain the absence of DOC, TDN, and TDP (Fig. 2), along with the compositional homogenization for trace elements in the frozen layer (Fig. 6), as polygonal mounds can lengthen solute residence time (less water loss and reduced hydraulic conductivity) and thus concentrate pore fluids (Raudina et al., 2017), […]”.
L412f: It is not directly clear how is this reference about the amplification of FTC on microbial N turnover is connected to the results of the manuscript. Please consider making the connection to the previous sentence clearer.We will clear the connection between the amplification of FTC on microbial N turnover to the results of the manuscript as in:
“Frozen horizons store a chemically labile and readily metabolizable DOM pool (FI > 1.4, higher concentrations of C from organic acids, and SUVA254 < 2), and rapid reactivation of the N cycling at the thaw front along with C mineralization processes (Fouché et al., 2020), enhanced by microbial metabolism driven by NH4+ pool consumption observed in simulated FTC (Yan et al., 2025) suggest a rapid biogeochemical reactivity.”
L414f: There is a lot of interpretation in this one sentence and I think it would be clearer for the potential readers of this manuscript if you would explain this step-by-step. First, you could start this new paragraph stating that you will talk about the bioavailability of P. Then, you could consider to mention that P can also be organic-bound, mainly by aromatic groups, and that thus the DOM composition affects the bioavailability of P. Finally, you can mention that the more aliphatic-dominated DOM in the pore ice likely leads to more bioavailable P upon thaw.We will update P bioavailability sentences as: “However, cycling of C and N is constrained by P bioavailability dependent on DOM composition, since organic-bound P binding with aromatics organic may inhibit mobilization (Hamdan et al., 2012). Thus, P in aliphatic-dominated pore ice fraction may be more bioavailable upon thawing.”
L525: What does “maximize the coupling” mean? I suggest to avoid such wordy terms.Concerning the “maximize the coupling” sentence, we agree and will replace “maximize the coupling” with “show a stronger association,” referring to the co-variation of DOM and exchangeable cation concentrations in pore ice.
Figures
L121, Fig. 1: Based on what reference did you identify the different ice forms? Or is this based on the sampled cores?The ice forms shown in Fig. 1b were identified from our field observations at the Tazovsky site over the past decade and from the examination of multiple cores collected for other investigations. These observations extend beyond the four cores analyzed in the present study. We have clarified the basis of this identification in the revised figure caption.
We will add to Fig. 1 caption the following: “The cryostructures illustrated in panel (b) are based on field observations at the Tazovsky site over the past decade, including examination of additional cores collected for other investigations.”
L221, Fig. 2: The given information on the active layer thickness in the caption is confusing to me. First, it is stated that the lower limit of the active layer thickness is from 35 to 65 cm, which corresponds to the active layer light blue band in the figure. Based on this, I would think that the maximum extent of the active layer is 35 to 65 cm across the cores, but we do not know which core has which extent. Then, different depths are given for the different cores in brackets, which I assumed were the maximum extents of active layer thickness. However, for Tz-1, it says 155 cm, which is well above the 65 cm given before. It would be helpful if the authors clarify that.We will clarify the active layer thickness representing the active layer depth at the time of sampling is represented for each site (Tz-4 = 40 ± 5 cm, Tz-3 = 60 ± 5, Tz-2 = 40 ± 5 cm and Tz-1 = 55 ± 5 cm).
(Please see supplement for the updated of the Figure 2). The corresponding text under the figure will be updated as:
"Figure 2. (Log)-profiles of dissolved (< 0.45 µm) (a) DOC, (b) TDN, (c) TDP, (d) C from Organic Acids (OA), (e) N-NH4+ and (f) P-PO43- (mg L-1) for porewater at the active layers and pore ice at the frozen layers. The lower limit of the active layer depth at the time of sampling is represented for each site (Tz-4 = 40 ± 5 cm, Tz-3 = 60 ± 5, Tz-2 = 40 ± 5 cm and Tz-1 = 55 ± 5 cm)."
Supplementary information
Table S4: Please consider adding short explanations on how the N and P limitation factors are calculated.
An explanation on N and P limitation factors will be added in the section “Note on the mass balance projections of C, N and P consumption and export under active layer deepening calculations” preceding the Table S4:
"Limitation factors are constrained by the amount of C that available N or P pools can support (Eq. 3).
Eq. 3 C_(cap,N (P)) = N (P)Total × C:N (P) consumption
Consumption of N and P fractions regarding the C pool is determined by ratios f_N and f_P (Eq. 4), limited at 1 (i.e., f_N = 1 implies a full mobilization of C pool carried by N sufficiency, whereas f_N = 0.2 only supports 20 % consumption of C pool).
Eq. 4 f_N (P) = min (1, (C(cap,N (P)) / C_Total)"
-
RC2: 'Reply on RC1', Anonymous Referee #2, 28 Sep 2026
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 250 | 131 | 37 | 418 | 62 | 26 | 30 |
- HTML: 250
- PDF: 131
- XML: 37
- Total: 418
- Supplement: 62
- BibTeX: 26
- EndNote: 30
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Overview
In this manuscript, the authors analyzed porewater chemistry in high-depth resolution along a permafrost thaw gradient in Siberia, targeting both active layer porewaters and pore ice. They found that there are substantial differences in DOC concentration and composition, as well as nutrient pools, in the permafrost layers compared to the active layers. Consumption and potential export rates of nutrients were modeled in order to evaluate the effect of active layer deepening in the future.
In general, I find the results presented in this manuscript very interesting and valuable for the field. The applied methods are described clearly and the figures are in general well understandable. However, the discussion of the results is sometimes not very focused and could use some revisions for clarity. I have some general questions and specific comments for revising the manuscript and I hope that these are helpful to the authors. As most of my questions can be addressed with textual changes but likely require some time, I recommend major revisions.
General comments and questions
Specific comments
L16: The use of “because” in that sentence is not logical. Permafrost peatlands are vulnerable to climate warming which can lead to active layer deepening, etc. I suggest to rearrange that sentence.
L39: Is the amount of C stock that is mentioned relating to the amount of released C from permafrost peatlands or the amount of sequestered C? The first one would be more logical, but in the current way in which the sentence is written, it is not clear. I suggest to rearrange that sentence.
L80: It is not clear to me what the “controls on pore ice formation” would be. It sounds like the authors would aim to study how easily ice forms in the different microhabitats. I suggest to clarify this aim.
L116: Was this whole procedure done directly on-site or after transport to some kind of lab?
L240: The use of the word “contribution” implies that the fraction of organic acids to total DOC are displayed in the figure. However, the data is given in absolute concentrations of mg /L. The authors could consider either changing the wording here or referring to another figure in which the contribution is visible, or simply summarizing the contribution in the text.
L251: What are carboxylic compounds in this case and how were they measured? In case the authors mean organic acids, as shown in Fig S2, then I would stick with that term and not introduce a new one.
L335f: The authors interpret the higher variation of frozen compared to active layer samples in the PCA as a greater variability in pore ice chemistry. However, in the next sentence an example is given that Tz-4 frozen samples show minimal variation along PC1 which shows lower variation within this site. Is that not a contradiction to the previous sentence?
L358f: I do not necessarily agree with the statement that the different chemical signatures between active layer porewater and frozen pore ice are due to freeze-thaw cycles. The successive freezing and thawing should only affect the active layer and not the permanently frozen compartments?
L361: Which compounds do you mean with “certain compound peaks”?
L409: include “potentially chemically labile DOM pool” since this discussion is based only on indicator variables and not on the direct DOM composition or a direct measure of DOM mineralization
L412f: It is not directly clear how is this reference about the amplification of FTC on microbial N turnover is connected to the results of the manuscript. Please consider making the connection to the previous sentence clearer.
L414f: There is a lot of interpretation in this one sentence and I think it would be clearer for the potential readers of this manuscript if you would explain this step-by-step. First, you could start this new paragraph stating that you will talk about the bioavailability of P. Then, you could consider to mention that P can also be organic-bound, mainly by aromatic groups, and that thus the DOM composition affects the bioavailability of P. Finally, you can mention that the more aliphatic-dominated DOM in the pore ice likely leads to more bioavailable P upon thaw.
L525: What does “maximize the coupling” mean? I suggest to avoid such wordy terms.
Figures:
L121, Figure 1: Based on what reference did you identify the different ice forms? Or is this based on the sampled cores?
Figure 1, subplot b: You can barely see the illustrated stream in the figure on the right side (legend item 11). I suggest to elongate this part so you can clearly see the stream if you want to include it.
Figure 1, subplot d: I suggest to specify in the figure caption that the frozen layers are corresponding to the blue overlay in the figure.
L221, Figure 2: The given information on the active layer thickness in the caption is confusing to me. First, it is stated that the lower limit of the active layer thickness is from 35 to 65 cm, which corresponds to the active layer light blue band in the figure. Based on this, I would think that the maximum extent of the active layer is 35 to 65 cm across the cores, but we do not know which core has which extent. Then, different depths are given for the different cores in brackets, which I assumed were the maximum extents of active layer thickness. However, for Tz-1, it says 155 cm, which is well above the 65 cm given before. It would be helpful if the authors clarify that.
L221, Figure 2: put “Tz-1” instead of “Tz-“ (also in L310)
Supplementary information:
Table S4: Please consider adding short explanations on how the N and P limitation factors are calculated.