the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 21 Jan 2026
Dissolved organic carbon-mediated controls dominate soil carbon mineralization in response to freeze-thaw cycles
Abstract. Soil freeze-thaw cycles (FTCs) exert substantial effects on the mineralization of soil organic carbon (SOC), particularly in high-altitude and -latitude cold regions. Ongoing climate change is altering FTC frequency and duration, yet the responses of SOC mineralization to such changes remain poorly understood, limiting our ability to predict carbon cycle-climate feedbacks. Here, we incubated soils from two depths across three sites to quantify how FTC regimes regulate SOC mineralization and explore underlying controls. Across all treatments, we observed a pronounced thaw-induced pulse of CO2 release, but more frequent freeze-thaw cycles led to more cumulative CO2 release, given the same length of cumulative thaw days. Across treatments, mineralization was most strongly correlated with DOC and hydrolytic/oxidative enzyme activities, while being suppressed by mineralogical (free and amorphous Fe/Al oxides) and physical (aggregate-protected carbon) constraints. Partial correlations and path analyses revealed that DOC was the single most consistent predictor of mineralization, retaining its influence even when enzymatic, substrate quality, or mineralogical variables were controlled. Subsoil SOC mineralization was additionally shaped by molecular carbon composition and mineral protection. These findings reveal a vertical shift from DOC-mediated substrate accessibility to molecularly and physically constrained decomposition. Accounting for these depth-specific mechanisms will improve prediction of SOC-climate feedbacks under FTC shifts due to climate change.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(2178 KB)
-
Supplement
(1850 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2178 KB) - Metadata XML
-
Supplement
(1850 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2026-139', Anonymous Referee #1, 27 Feb 2026
-
AC1: 'Reply on RC1', Zhongkui Luo, 26 Mar 2026
Dear Reviewer 1,
We would like to thank you for your time and patience in giving constructive criticism and suggestion to improve the manuscript. Your comments have help us improve the clarity and scope of the manuscript.Â
Our complete response to the reviewer's comments are in the Supplement.
-
AC1: 'Reply on RC1', Zhongkui Luo, 26 Mar 2026
-
RC2: 'Comment on egusphere-2026-139', Anonymous Referee #2, 05 Mar 2026
I read "Dissolved organic carbon-mediated controls dominate soil carbon mineralization in response to freeze-thaw cycles" by Jiaxin Yan et al. with great interest. The topic is timely, the experimental design is well-conceived, and the interdisciplinary approach, combining soil physics, biogeochemistry, and microbiology, is particularly valuable. I believe this work will be of broad interest to the community. However, three main concerns should be addressed before publication.
MAIN COMMENTS
- Coherence between objectives, analyses, and conclusions, including statistical transparency
This is my primary concern. The stated objectives focus on how freeze-thaw cycle duration and frequency regulate SOC mineralization. However, a substantial part of the analytical framework (correlation analyses, partial correlations, path model) pools observations across all freeze-thaw treatments to identify general predictors of Rs, effectively setting aside the treatment structure. These analyses address a different question: what controls SOC mineralization in general, not specifically how freeze-thaw regimes drive it. When the discussion then uses these treatment-blind results to make mechanistic claims about freeze-thaw effects, a logical gap emerges that is not acknowledged. The analytical framework as presented does not allow the reader to clearly assess whether DOC responses differ systematically between freeze-thaw regimes.
This confusion is compounded by a lack of clarity on how temporal dependency was handled across analyses. While I am not a statistical expert, several analyses, most notably the MANOVA and the PLS-PM, include repeated measurements from the same incubation jars across multiple time points, and it is not clear how non-independence among observations was accounted for. This is particularly relevant for the PLS-PM, where all thaw-phase observations are explicitly pooled, which may inflate the effective sample size and produce overly optimistic significance levels. Reported n values also differ across figures without explicit justification, making it difficult to assess the unit of replication in each case.
The core conclusions, that FTC frequency drives cumulative mineralization and that DOC is the dominant predictor of Rs, appear robust. However, the conclusion that freeze-thaw regimes influence mineralization indirectly through DOC is difficult to evaluate as presented. The authors may well have handled these issues appropriately, but the PLS-PM as described pools repeated measurements from the same jars across time points, and it is not clear to the reader how temporal variance was separated from treatment variance in estimating the reported path coefficients. Without this clarification, it is difficult to assess whether the treatment pathways reflect genuine differences between freeze-thaw regimes or partly capture within-jar temporal dynamics. The authors should clarify how temporal dependency was handled and clearly delineate which findings speak to general soil biogeochemistry versus freeze-thaw-specific mechanisms.
- Enzyme activity versus enzyme concentration
The manuscript reports extracellular enzyme "activities" measured using commercial ELISA kits. ELISA assays primarily quantify enzyme protein abundance through antibody binding rather than catalytic activity. Unless the standard curve was explicitly constructed from enzymes of known catalytic activity, the use of the term "activity" throughout the text and figure legends may be misleading. Furthermore, extracellular enzymes in soils can be stabilized on mineral surfaces or organic matter, and enzymes such as β-glucosidase comprise multiple classes across diverse microbial taxa that may not be equally recognized by the antibodies used. The authors should clarify what is actually being quantified, whether any conversion to catalytic units was applied, and discuss these methodological limitations.
- Contextualisation with in situ manipulation literature
The manuscript would benefit from a more systematic comparison with the large body of literature on in situ snow cover manipulation and soil translocation experiments, which increase freeze-thaw frequency and have documented consequences for soil carbon cycling. This literature is largely absent and would strengthen the contextualisation of the findings. In particular, the authors should discuss whether the treatment effects observed in their controlled incubation are consistent in direction and magnitude with what has been reported in field manipulation studies, and where discrepancies exist, propose possible explanations.
MINOR COMMENTS
- L34: Does increased FTC frequency consistently increase SOC decomposition in the literature? Snow cover typically protects soil from freezing and may actually reduce decomposition. Please clarify or add appropriate references.
- L44: "Negative effects on microbial community composition" is vague. Please rephrase.
- L124: Soil moisture was maintained at 60% water-holding capacity, please specify how WHC was measured or estimated, particularly for sieved soils.
- L149: Please specify centrifugation duration and relative centrifugal force.
- Figure 1: Please add the treatment acronym (e.g. LFLT, SFST) directly as a panel title to facilitate reading.
- Figure 3: If error bars reflect temporal variability within a single jar rather than biological replication, this should be stated explicitly in the caption.
Citation: https://doi.org/10.5194/egusphere-2026-139-RC2 -
AC2: 'Reply on RC2', Zhongkui Luo, 26 Mar 2026
Dear Reviewer 2,
We sincerely appreciate the time and effort you devoted to reviewing our manuscript. Your constructive comments and suggestions have been very valuable in improving the clarity and scope of the work.
Our point-by-point responses to your comments are included in the Supplementary Material.
Peer review completion
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2026-139', Anonymous Referee #1, 27 Feb 2026
General comment
This paper presents the response of three different soils (representing different altitudes) taken at two depths (0–10 and 70–80 cm) to four different kinds of freeze-thaw cycles (FTCs): long freeze/long thaw, long freeze/short thaw, short freeze/long thaw and short freeze/short thaw. The study, including a detailed experimental part and rigorous statistical analyses, highlights the differences in SOC mineralization across the altitudes, depths, and FTC types. Specifically, more frequent FTCs lead to higher SOC mineralization. Further investigation was conducted to determine the predictors of SOC mineralization. Dissolved organic carbon appears to be the most important of these predictors, both in topsoil and subsoil.
I found the article very interesting: the experimental design, while presenting limits, is both simple and efficient, and the results and conclusions are overall well presented. However, I got a bit lost in the statistical analysis part. This will certainly be useful for statistics enthusiasts, but it is a bit difficult to follow for the average reader (which I am). Be careful not to lose us in the way, because the paper is of great interest for everyone and the findings must be clear even for those who don’t master fully how you obtain them.
Overall, this paper is a nice addition to the knowledge on the effects of climate change – and specifically frequent FTCs – on soils and SOC mineralization.
Â
Specific comments
L55: please specify the depth you consider for subsoil.
L65: please define ‘labile’ (in terms of residence time for instance). Also, you could use a reference to establish DOC as a labile pool (while it is generally thought of as labile without more precision, several studies showed that part of it can persist for decades, see Kalbitz & Kaiser, 2008 (https://doi.org/10.1002/jpln.200700043) for example).
L109: what do you mean by ‘quality’?
L145: where does the 0.45 value come from?
L166: what are substrates A and B? If it is not important, it may be better not to mention it.
L250: ‘On the first thawing day (Fig. 3; Fig. 4), DOC concentrations varied significantly among treatments’ → if I am not mistaken, this is not visible on the figures you indicate; the uppercase letters in Fig.3 for DOC are all A and do not show the first thawing day.
L255: if I interpret it well, the MBC indicates that there are more microbes in the topsoil horizon than in subsoil (which is not surprising). How do you explain that enzyme activities are higher in subsoil (although not always significantly)?
L262: it is not completely clear to me how to read the partial correlations. What do PC x DOC = 0.66 (bottom left corner) and DOC x PC = 0.79 (top right corner) correspond to? Is 0.66 the partial correlation between SOC mineralization and PC with DOC controlled, or the contrary? Sorry if this is usual for this type of graph; perhaps adding a word about it would help. Also, why aren’t all the boxes filled?
L290: I am not familiar with path analysis, but I don’t see how ‘subsoil mineralization exhibited strong additional associations with C molecular composition’ (that we indeed see on Fig. S7) shows on Fig.6b. All values related to the molecular structure seem quite low. Shouldn’t we also see this correlation result on the path analysis?
Â
Technical comments
L31: you need spaces when using ‘ - ‘, also I think it needs a longer dash. Same later in the sentence.
L43: ‘and are sensitive to FTCs’
L44: ‘can recover rapidly’
L87: typo in ‘physicochemical’, and is a word missing? Perhaps ‘physicochemical properties’?
L96–99: the sentence seems to be repeated.
L148: vertexing → vortexing?
L248: spaces and longer dash needed when using ‘ - ‘.
L343: ‘due to the fact that DOC’
Citation: https://doi.org/10.5194/egusphere-2026-139-RC1 -
AC1: 'Reply on RC1', Zhongkui Luo, 26 Mar 2026
Dear Reviewer 1,
We would like to thank you for your time and patience in giving constructive criticism and suggestion to improve the manuscript. Your comments have help us improve the clarity and scope of the manuscript.Â
Our complete response to the reviewer's comments are in the Supplement.
-
AC1: 'Reply on RC1', Zhongkui Luo, 26 Mar 2026
-
RC2: 'Comment on egusphere-2026-139', Anonymous Referee #2, 05 Mar 2026
I read "Dissolved organic carbon-mediated controls dominate soil carbon mineralization in response to freeze-thaw cycles" by Jiaxin Yan et al. with great interest. The topic is timely, the experimental design is well-conceived, and the interdisciplinary approach, combining soil physics, biogeochemistry, and microbiology, is particularly valuable. I believe this work will be of broad interest to the community. However, three main concerns should be addressed before publication.
MAIN COMMENTS
- Coherence between objectives, analyses, and conclusions, including statistical transparency
This is my primary concern. The stated objectives focus on how freeze-thaw cycle duration and frequency regulate SOC mineralization. However, a substantial part of the analytical framework (correlation analyses, partial correlations, path model) pools observations across all freeze-thaw treatments to identify general predictors of Rs, effectively setting aside the treatment structure. These analyses address a different question: what controls SOC mineralization in general, not specifically how freeze-thaw regimes drive it. When the discussion then uses these treatment-blind results to make mechanistic claims about freeze-thaw effects, a logical gap emerges that is not acknowledged. The analytical framework as presented does not allow the reader to clearly assess whether DOC responses differ systematically between freeze-thaw regimes.
This confusion is compounded by a lack of clarity on how temporal dependency was handled across analyses. While I am not a statistical expert, several analyses, most notably the MANOVA and the PLS-PM, include repeated measurements from the same incubation jars across multiple time points, and it is not clear how non-independence among observations was accounted for. This is particularly relevant for the PLS-PM, where all thaw-phase observations are explicitly pooled, which may inflate the effective sample size and produce overly optimistic significance levels. Reported n values also differ across figures without explicit justification, making it difficult to assess the unit of replication in each case.
The core conclusions, that FTC frequency drives cumulative mineralization and that DOC is the dominant predictor of Rs, appear robust. However, the conclusion that freeze-thaw regimes influence mineralization indirectly through DOC is difficult to evaluate as presented. The authors may well have handled these issues appropriately, but the PLS-PM as described pools repeated measurements from the same jars across time points, and it is not clear to the reader how temporal variance was separated from treatment variance in estimating the reported path coefficients. Without this clarification, it is difficult to assess whether the treatment pathways reflect genuine differences between freeze-thaw regimes or partly capture within-jar temporal dynamics. The authors should clarify how temporal dependency was handled and clearly delineate which findings speak to general soil biogeochemistry versus freeze-thaw-specific mechanisms.
- Enzyme activity versus enzyme concentration
The manuscript reports extracellular enzyme "activities" measured using commercial ELISA kits. ELISA assays primarily quantify enzyme protein abundance through antibody binding rather than catalytic activity. Unless the standard curve was explicitly constructed from enzymes of known catalytic activity, the use of the term "activity" throughout the text and figure legends may be misleading. Furthermore, extracellular enzymes in soils can be stabilized on mineral surfaces or organic matter, and enzymes such as β-glucosidase comprise multiple classes across diverse microbial taxa that may not be equally recognized by the antibodies used. The authors should clarify what is actually being quantified, whether any conversion to catalytic units was applied, and discuss these methodological limitations.
- Contextualisation with in situ manipulation literature
The manuscript would benefit from a more systematic comparison with the large body of literature on in situ snow cover manipulation and soil translocation experiments, which increase freeze-thaw frequency and have documented consequences for soil carbon cycling. This literature is largely absent and would strengthen the contextualisation of the findings. In particular, the authors should discuss whether the treatment effects observed in their controlled incubation are consistent in direction and magnitude with what has been reported in field manipulation studies, and where discrepancies exist, propose possible explanations.
MINOR COMMENTS
- L34: Does increased FTC frequency consistently increase SOC decomposition in the literature? Snow cover typically protects soil from freezing and may actually reduce decomposition. Please clarify or add appropriate references.
- L44: "Negative effects on microbial community composition" is vague. Please rephrase.
- L124: Soil moisture was maintained at 60% water-holding capacity, please specify how WHC was measured or estimated, particularly for sieved soils.
- L149: Please specify centrifugation duration and relative centrifugal force.
- Figure 1: Please add the treatment acronym (e.g. LFLT, SFST) directly as a panel title to facilitate reading.
- Figure 3: If error bars reflect temporal variability within a single jar rather than biological replication, this should be stated explicitly in the caption.
Citation: https://doi.org/10.5194/egusphere-2026-139-RC2 -
AC2: 'Reply on RC2', Zhongkui Luo, 26 Mar 2026
Dear Reviewer 2,
We sincerely appreciate the time and effort you devoted to reviewing our manuscript. Your constructive comments and suggestions have been very valuable in improving the clarity and scope of the work.
Our point-by-point responses to your comments are included in the Supplementary Material.
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 1,448 | 839 | 130 | 2,417 | 354 | 106 | 174 |
- HTML: 1,448
- PDF: 839
- XML: 130
- Total: 2,417
- Supplement: 354
- BibTeX: 106
- EndNote: 174
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Jiaxin Yan
Jinyang Zheng
Shuai Zhang
Mingming Wang
Ting Sun
Jiajun Mao
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2178 KB) - Metadata XML
-
Supplement
(1850 KB) - BibTeX
- EndNote
- Final revised paper
General comment
This paper presents the response of three different soils (representing different altitudes) taken at two depths (0–10 and 70–80 cm) to four different kinds of freeze-thaw cycles (FTCs): long freeze/long thaw, long freeze/short thaw, short freeze/long thaw and short freeze/short thaw. The study, including a detailed experimental part and rigorous statistical analyses, highlights the differences in SOC mineralization across the altitudes, depths, and FTC types. Specifically, more frequent FTCs lead to higher SOC mineralization. Further investigation was conducted to determine the predictors of SOC mineralization. Dissolved organic carbon appears to be the most important of these predictors, both in topsoil and subsoil.
I found the article very interesting: the experimental design, while presenting limits, is both simple and efficient, and the results and conclusions are overall well presented. However, I got a bit lost in the statistical analysis part. This will certainly be useful for statistics enthusiasts, but it is a bit difficult to follow for the average reader (which I am). Be careful not to lose us in the way, because the paper is of great interest for everyone and the findings must be clear even for those who don’t master fully how you obtain them.
Overall, this paper is a nice addition to the knowledge on the effects of climate change – and specifically frequent FTCs – on soils and SOC mineralization.
Â
Specific comments
L55: please specify the depth you consider for subsoil.
L65: please define ‘labile’ (in terms of residence time for instance). Also, you could use a reference to establish DOC as a labile pool (while it is generally thought of as labile without more precision, several studies showed that part of it can persist for decades, see Kalbitz & Kaiser, 2008 (https://doi.org/10.1002/jpln.200700043) for example).
L109: what do you mean by ‘quality’?
L145: where does the 0.45 value come from?
L166: what are substrates A and B? If it is not important, it may be better not to mention it.
L250: ‘On the first thawing day (Fig. 3; Fig. 4), DOC concentrations varied significantly among treatments’ → if I am not mistaken, this is not visible on the figures you indicate; the uppercase letters in Fig.3 for DOC are all A and do not show the first thawing day.
L255: if I interpret it well, the MBC indicates that there are more microbes in the topsoil horizon than in subsoil (which is not surprising). How do you explain that enzyme activities are higher in subsoil (although not always significantly)?
L262: it is not completely clear to me how to read the partial correlations. What do PC x DOC = 0.66 (bottom left corner) and DOC x PC = 0.79 (top right corner) correspond to? Is 0.66 the partial correlation between SOC mineralization and PC with DOC controlled, or the contrary? Sorry if this is usual for this type of graph; perhaps adding a word about it would help. Also, why aren’t all the boxes filled?
L290: I am not familiar with path analysis, but I don’t see how ‘subsoil mineralization exhibited strong additional associations with C molecular composition’ (that we indeed see on Fig. S7) shows on Fig.6b. All values related to the molecular structure seem quite low. Shouldn’t we also see this correlation result on the path analysis?
Â
Technical comments
L31: you need spaces when using ‘ - ‘, also I think it needs a longer dash. Same later in the sentence.
L43: ‘and are sensitive to FTCs’
L44: ‘can recover rapidly’
L87: typo in ‘physicochemical’, and is a word missing? Perhaps ‘physicochemical properties’?
L96–99: the sentence seems to be repeated.
L148: vertexing → vortexing?
L248: spaces and longer dash needed when using ‘ - ‘.
L343: ‘due to the fact that DOC’