Urbanization drives coupled shifts in soil-carbon stocks, sources, and stability across natural and restored mangroves and tidal flats

Huang, Minde; Guo, Fen; Gao, Xueqin; Ouyang, Xiaoguang; Zhang, Yuan

doi:10.5194/egusphere-2026-2512

Preprints

https://doi.org/10.5194/egusphere-2026-2512

Preprints

13 May 2026

| 13 May 2026

Urbanization drives coupled shifts in soil-carbon stocks, sources, and stability across natural and restored mangroves and tidal flats

Minde Huang, Fen Guo, Xueqin Gao, Xiaoguang Ouyang, and Yuan Zhang

Abstract. Urbanization reshapes coastal blue carbon, but its effects on how much carbon is stored, who supplies it, and how long it persists remain poorly integrated. We investigated natural and restored mangroves and adjacent tidal flats along an urbanization gradient, and quantified soil organic carbon stocks, burial from ²¹⁰Pb profiles, source composition using isotope end-member mixing, and stability from turnover metrics. Our results showed a coordinated triad response to urbanization. Carbon stocks and burial declined, sources shifted away from mangrove detritus toward planktonic and algal inputs, and turnover accelerated, lowering stability. Responses were habitat dependent and nonlinear. Natural mangroves in low urbanization settings maintained the highest sequestration with mangrove-dominated inputs and slower turnover. Restored stands and tidal flats showed steeper stock losses, stronger source substitution, and faster cycling under higher urban pressure. We introduced a triad framework that treats stocks, sources, and stability as coupled state variables along the urbanization gradient and identified two reproducible system states: carbon anchors in low urbanization natural mangroves and instability fronts in restored stands and tidal flats. Shifts in sources and stability precede stock losses, providing clear early warnings of urban impact. A simple diagnostic that combines connectivity and accretion with source composition and a composite stability index guides anchor protection and stability-first restoration. These results link urban growth to blue-carbon performance and define actionable thresholds for sustaining coastal carbon.

Received: 30 Apr 2026 – Discussion started: 13 May 2026

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Minde Huang, Fen Guo, Xueqin Gao, Xiaoguang Ouyang, and Yuan Zhang

Status: final response (author comments only)

CC1: 'Comment on egusphere-2026-2512', Li Gang, 27 May 2026

AC5: 'Reply on CC1', Fen Guo, 14 Jul 2026

General Comments

Comments

Responses

This well-written manuscript investigates soil organic carbon stocks, burial rate, source composition, and turnover rate along an urbanization gradient in natural (low urbanization) and restored mangroves (moderate urbanization), as well as adjacent tidal flats (high urbanization) in Zhanjiang, Huizhou, and Shantou, Guangdong Province, Southern China. The study found that urbanization reduces carbon stocks and burial rates and shifts carbon sources from mangrove detritus toward planktonic and algal inputs. Linking urban growth to blue carbon performance is interesting and useful for defining actionable thresholds to sustain coastal carbon.

We sincerely thank the community reviewer for the positive evaluation of our manuscript. We appreciate your encouraging comments and support.

Specific Comments

Comments

Responses

My main concern is that the investigation in areas with low, moderate, and high levels of urbanization was conducted in three different cities in China. Whether or to what extent do the substantial differences in environmental factors among these cities affect the conclusions?

Thank you for raising this important concern. We agree that the three cities differ not only in urbanization intensity but also in environmental and geomorphic setting. Because each urbanization level was represented by a single city, city identity and urbanization level are confounded in the present study design. Consequently, the available data do not allow us to quantify precisely how much of the observed variation in soil organic carbon stocks, source contributions, burial, and turnover is attributable to urbanization rather than to differences in hydrodynamics, tidal regime, sediment supply, salinity, sediment texture, or natural accretion processes among cities.

All study regions are located along the subtropical coast of Guangdong Province and therefore share a broadly comparable climatic setting. This regional consistency reduces large-scale climatic variation, but it does not eliminate local environmental and geomorphic differences. We have revised the Discussion to make this distinction explicit.

We therefore interpret the observed patterns as associations between soil-carbon properties and contrasting urbanization contexts, rather than as causal effects of urbanization isolated from city-specific environmental conditions. The broadly concordant changes observed across several carbon indicators and habitat categories increase confidence that urbanization may contribute to the reported patterns, but they do not independently separate urbanization from regional environmental heterogeneity. We have tempered the causal language throughout the manuscript accordingly.

We have also added this limitation to the Discussion and Conclusions. A more rigorous test would require several independent cities within each urbanization category, comparable habitat types within each city, and direct measurements of tidal dynamics, sediment supply, geomorphic setting, and other environmental covariates. Such a replicated design would allow urbanization effects to be separated from regional variation using hierarchical models or variance-partitioning approaches.

As shown in the manuscript, annual sunshine duration increases from less than 1,500 hours in Shantou (north) to over 2,100 hours in Zhanjiang (south), indicating that environmental conditions, climate, and biological processes differ greatly. The impacts of these differences on carbon sources cannot be ignored. The manuscript should address and discuss the extent to which these differences affect carbon sources.

Thank you for raising this important point. We agree that the substantial difference in annual sunshine duration among the three regions represents more than background geographic variation and may influence both the production and the isotopic characteristics of potential carbon sources.

Differences in solar radiation and associated climatic conditions may affect soil organic carbon source contributions through several pathways. Greater light availability can alter mangrove productivity, litter production, and belowground carbon inputs, while also influencing phytoplankton productivity and the supply of aquatic organic matter to coastal sediments. Regional climatic differences may additionally affect organic matter decomposition and preservation. Moreover, light availability, temperature, salinity, growth conditions, and community composition can influence the δ¹³C and δ¹⁵N signatures of both vascular plants and phytoplankton. These differences may therefore affect not only the actual supply of carbon sources but also their estimated contributions in the isotope mixing model.

Because each urbanization level is represented by a different city, the present study cannot quantitatively separate the influence of sunshine duration and other regional environmental conditions from that of urbanization. We have therefore revised the Discussion to clarify that the observed changes in organic carbon source contributions are associated with contrasting urbanization and regional environmental contexts, rather than being attributable exclusively to urbanization. We have also tempered causal statements throughout the manuscript accordingly.

The concordant shift from C3 vascular plant-derived carbon toward phytoplankton-derived carbon across several habitat categories suggests that urbanization-related changes may contribute to the observed pattern. However, regional differences in solar radiation, hydrodynamics, sediment supply, salinity, and primary productivity may also have influenced the magnitude of this shift. We now identify this geographic confounding explicitly as a limitation of the study.

A rigorous quantification of these relative effects would require multiple independent cities within each urbanization category, together with direct measurements of solar radiation, primary productivity, hydrodynamics, and sediment transport. Future studies should also use locally resolved source end-members to determine whether geographic variation in isotope signatures affects source-apportionment estimates.

Citation: https://doi.org/10.5194/egusphere-2026-2512-AC5

RC1: 'Comment on egusphere-2026-2512', Anonymous Referee #1, 01 Jun 2026

In my opinion, this paper addresses relevant scientific questions within the scope of BG and presents novel ideas and tools. I believe the topics addressed are well represented in the manuscript title and abstract and are important for the field of environmental science as a whole. I suggest accepting this manuscript with some revisions.

General comments:

Figures: All of your figures look very different from each other. Similar categories (Natural mangroves, Restored mangroves, Tidal flats or High, Moderate, Low-urbanization) are shown in several different colors or layouts throughout your figures. It might be easier for the reader if you picked a consistent color scheme throughout (example: applying the green, yellow, red scheme you used to depict High, Moderate, and Low-urbanizations in Figure 6 to all of your relevant figures) or consistent layout (like the clearly differentiated column and row headings in Figure 5).

Specific comments:

line 28: Do you want to include this image as Figure 1? If so, please add a figure description and adjust figure numbering throughout the manuscript accordingly.

line 72: Both “blue-carbon” and “blue carbon” have been used in the manuscript so far. Please choose one format and apply throughout the whole manuscript.

line 90: Figure 1. Please make some adjustments to Figure 1. Inset C is very clear and easy to understand. Insets B and D are not as clear. It would be helpful to have the first views for Insets B and D zoomed out more so we can tell where we are (as you did with Inset A). It would also be helpful to keep the color scheme consistent throughout all of the maps (water always shown as blue, adjoining land always the same gray or white as they are in Inset A).

lines 91-92: Figure 1 description. I do not see urbanization gradients depicted in your Figure 1. Please either add urbanization gradients to the figure (or make them more obvious if I have missed them) or adjust this figure description.

lines 106-107: Is this soil bulk density dry or wet? Based on information further down in the manuscript, it seems like you looked at dry bulk density, specifically. If this is the case, please clarify here.

lines 107-108: Please explain your leaf sample methodology further. What do you mean by “according to the local mangrove community composition”?

line 110: Please provide information on what kind of soil corer you used as you mention compaction issues further down.

lines 111-112: Please provide details on why you adjusted your soil core section intervals as you got deeper.

line 117: At what temperature did you refrigerate the samples?

lines 120-121: Please provide an explanation and a citation for a publication to support this assumption.

line 121: Please provide more information in this section of the manuscript on why you did not assess sediment accumulation rates for tidal flats.

lines 125-135: Please provide citations for your soil parameter measurement protocols.

lines 134-135: It might be worthwhile to define SBD as DBD (dry bulk density) to be more specific about the kind of soil bulk density your measured.

lines 158-159: I think this methodology is probably fine. Ideally, data exploration would involve a few more steps than listed here. A paper that provides a detailed protocol for data exploration with ecological data is Zuur, A. F., Ieno, E.N., and Elphick, C. S. 2010. A protocol for data exploration to avoid common statistical problems. Methods in Ecology and Evolution. https://doi.org/10.1111/j.2041-210X.2009.00001.x.

line 191: Table 1, line 1, column 3 (High-urbanization natural mangroves, Salinity): 52 g kg ^-1 seems high for this part of the world. Please double-check this is not a typo. If this is correct, please discuss the high salinity in more depth in your discussion section. If this salinity level is a known feature of this particular area, please include that in section 2.1 Study area.

lines 194-196: Figure 2 description: Please include units in all of your figure descriptions.

lines 198-199: Figure 3 description: Please include units in all of your figure descriptions. What do you mean by "different letters indicated significant differences in soil organic carbon among urbanization levels within the same ecosystem type..."? Define what the letters are and what each of them means, exactly.

line 208: Figure 4: It seems you have two main rows for these figure insets (Natural Mangroves and Restored Mangroves) and three main columns (High-urbanization, Moderate-urbanization, and Low-urbanization). Can you adjust the figure so it is more obvious that these insets all fit inside these different categories? Simply making the row and column titles bigger may achieve this or you could add a background grid or ribbons like you did in Figure 6.

lines 209-210: Figure 4 description: Please provide more details in your figure description, such as units. This would also be a good place to clarify which of the Inset Figures are natural mangroves vs restored mangroves and high vs moderate vs low-urbanization.

lines 208-210: What locations do Inset Figures A, B, C, D, and E correspond with (ex: Shantou, Zhanjiang, or Huizhou)? Please provide the location names in either the figure or in the figure description.

lines 231-232: Figure 5 description: Please include more details in your figure description. What do the different colored boxes for your different depth intervals represent? Do Insets A, B, C, D, and E correspond with the same Insets in Figure 4? If so, it may be helpful to include the locations for each of these either in the figure or in the figure description.

lines 246-247: Figure 6 description: Please provide more details in this figure description. What do you mean by "different letters indicated significant differences in Beta values among urbanization levels within the same ecosystem type..."? Define what the letters are and what each of them means, exactly.

line 270: “Once those limits are crossed, previously buried horizons return to active cycling and carbon stock decline rapidly…” Provide citation.

Citation: https://doi.org/10.5194/egusphere-2026-2512-RC1

AC1: 'Reply on RC1', Fen Guo, 14 Jul 2026

General Comments

Comments	Responses
In my opinion, this paper addresses relevant scientific questions within the scope of BG and presents novel ideas and tools. I believe the topics addressed are well represented in the manuscript title and abstract and are important for the field of environmental science as a whole.	We sincerely thank you for your positive evaluation of our manuscript.
Figures: All of your figures look very different from each other. Similar categories (Natural mangroves, Restored mangroves, Tidal flats or High, Moderate, Low-urbanization) are shown in several different colors or layouts throughout your figures. It might be easier for the reader if you picked a consistent color scheme throughout (example: applying the green, yellow, red scheme you used to depict High, Moderate, and Low-urbanizations in Figure 6 to all of your relevant figures) or consistent layout (like the clearly differentiated column and row headings in Figure 5).	Thank you very much for your valuable suggestion. We have standardized all figures in the revised manuscript by adopting a consistent color scheme to represent different urbanization levels, thereby improving consistency across figures and enhancing overall readability and clarity.

Specific Comments

Comments	Responses
line 28: Do you want to include this image as Figure 1? If so, please add a figure description and adjust figure numbering throughout the manuscript accordingly.	Thank you very much for your valuable and constructive suggestion. In response, we have carefully checked and unified the figure numbering in the revised manuscript. This figure was not included as Figure 1 and has been renumbered as Figure 7 according to its position in the main text. Meanwhile, we have added the corresponding caption: “Figure 7. Variations in soil organic carbon stocks, source composition, and turnover rates in natural mangroves, restored mangroves, and unvegetated tidal flats across different urbanization gradients.” In addition, we have thoroughly checked all figure numbering throughout the manuscript to ensure consistency between figures and in-text citations.
line 72: Both “blue-carbon” and “blue carbon” have been used in the manuscript so far. Please choose one format and apply throughout the whole manuscript.	Thank you very much for your valuable suggestion. In response, we have carefully revised the entire manuscript to ensure consistency in terminology, replacing “blue-carbon” with “blue carbon” throughout the text to improve standardization and linguistic consistency.
line 90: Figure 1. Please make some adjustments to Figure 1. Inset C is very clear and easy to understand. Insets B and D are not as clear. It would be helpful to have the first views for Insets B and D zoomed out more so we can tell where we are (as you did with Inset A). It would also be helpful to keep the color scheme consistent throughout all of the maps (water always shown as blue, adjoining land always the same gray or white as they are in Inset A).	Thank you very much for your valuable comments. In response to your suggestions, we have revised Figure 1 accordingly (as shown below). Specifically, panels (b) and (d) have been appropriately adjusted by expanding the map extents to improve spatial context and clarity. In addition, we have unified the color scheme across all maps, with water bodies consistently represented in blue, thereby enhancing the overall consistency, readability, and standardization of the figures.
lines 91-92: Figure 1 description. I do not see urbanization gradients depicted in your Figure 1. Please either add urbanization gradients to the figure (or make them more obvious if I have missed them) or adjust this figure description.	Thank you very much for your valuable suggestion. In response, we have revised the caption of Figure 1 accordingly. The updated caption is as follows:“Figure 1. Location of sampling sites across high, moderate, and low urbanization gradients in southern China, showing natural mangroves, restored mangroves, and tidal flats. Panels (b), (c), and (d) show the sampling sites in Shantou, Zhanjiang, and Huizhou, representing moderate, low, and high urbanization levels, respectively.” This revision further clarifies the information presented in the figure, ensures consistency between the figure and its caption, and more accurately specifies the urbanization levels corresponding to the three study regions.
lines 106-107: Is this soil bulk density dry or wet? Based on information further down in the manuscript, it seems like you looked at dry bulk density, specifically. If this is the case, please clarify here.	Thank you very much for your valuable comment. In this study, soil bulk density refers to dry soil bulk density. In response to your suggestion, we have revised the Methods section accordingly by changing “soil bulk density (SBD)” to “dry soil bulk density (SBD)” throughout the manuscript to avoid potential ambiguity and to improve the accuracy and completeness of the methodological description.
lines 107-108: Please explain your leaf sample methodology further. What do you mean by “according to the local mangrove community composition”?	Thank you very much for your valuable comment. Due to the lack of clarity in the original description, it may have led to ambiguity. In the revised manuscript, we have further clarified the leaf sampling procedure as follows: within each plot, all mangrove species present were identified according to the actual community composition, and mature and healthy leaves were collected from each species. Three replicate samples were set for each species to more comprehensively characterize the stable isotope signatures of mangrove plant leaves in the study area.
line 110: Please provide information on what kind of soil corer you used as you mention compaction issues further down.	Thank you very much for your valuable comment. In response, we have further clarified the soil sampling method in the Methods section of the revised manuscript. Specifically, soil samples were collected using a custom-made stainless-steel coring device (inner diameter: 5 cm). The design of the corer and the sampling procedure follow the protocol recommended by the International Blue Carbon Initiative, as described in the “Coastal Blue Carbon: Methods for Assessing Carbon Stocks and Emissions in Mangroves, Salt Marshes, and Seagrasses” manual (https://www.thebluecarboninitiative.org/manual), ensuring consistency with internationally recognized standards.
lines 111-112: Please provide details on why you adjusted your soil core section intervals as you got deeper.	Thank you very much for your valuable comments. In the Methods section of the revised manuscript, we have further clarified that soil samples were collected using a custom-made stainless-steel coring device (inner diameter: 5 cm). The design of the corer and the sampling procedures follow the guidelines of the International Blue Carbon Initiative, specifically the “Coastal Blue Carbon: Methods for Assessing Carbon Stocks and Emissions in Mangroves, Salt Marshes, and Seagrasses” manual (https://www.thebluecarboninitiative.org/manual). Many coastal sediment studies adopt a similar sampling strategy characterized by high-resolution sampling in surface layers and coarser intervals at deeper depths (Hu et al., 2024; Rani et al., 2023; Santos-Andrade et al., 2021; Sasmito et al., 2020; Yu et al., 2021; Zhang et al., 2012). Therefore, this widely used stratification approach was adopted in our study to ensure methodological comparability and improve analytical efficiency.
line 117: At what temperature did you refrigerate the samples?	Thank you very much for your valuable comment. In response to your suggestion, we have further clarified the statement in the Methods section of the revised manuscript by replacing “All samples were labeled and refrigerated prior to laboratory processing and analysis” with “All samples were labeled and stored at −80 °C prior to laboratory processing and analysis” to ensure greater accuracy and methodological clarity.
lines 120-121: Please provide an explanation and a citation for a publication to support this assumption.	Thank you very much for your valuable comment. We clarify that this hypothesis was primarily intended to explain the rationale for selecting one representative sediment core for ²¹⁰Pb dating within each ecosystem type and urbanization level, and subsequently using the derived sediment accretion rate to estimate soil organic carbon burial rates for the corresponding ecosystem. Due to the relatively high cost, labor intensity, and technical requirements associated with ²¹⁰Pb dating analyses, regional-scale blue carbon studies often require the selection of representative sediment cores for estimating sediment accretion rates based on specific research objectives. Previous studies have demonstrated that sediment accretion rates may show relatively limited variation among different locations or restoration conditions within the same mangrove ecosystem under similar environmental settings. For example, MacKenzie et al. (2016) reported no significant differences in sediment accretion rates among disturbed and undisturbed mangroves, edge and interior mangrove habitats, as well as naturally recovering and artificially restored mangroves, suggesting relatively consistent sedimentation dynamics within comparable mangrove systems. In addition, Huang et al.(2025b) adopted a similar sampling strategy for sediment accretion rate estimation in mangrove ecosystems. Following your suggestion, we have added these relevant references (Huang et al., 2025b; MacKenzie et al., 2016) to the revised manuscript to provide further support for our sampling approach.
line 121: Please provide more information in this section of the manuscript on why you did not assess sediment accumulation rates for tidal flats.	Thank you very much for your valuable comment. In this study, ²¹⁰Pb dating and sediment accretion rate estimation were not conducted in unvegetated tidal flats, mainly because tidal flats are strongly influenced by dynamic tidal processes and generally lack stable and continuous sedimentary sequences. The frequent resuspension and disturbance of sediments in these environments may compromise the assumptions required for ²¹⁰Pb dating and reduce the reliability of sediment accretion rate estimates. Furthermore, previous studies have demonstrated that the long-term carbon storage capacity of unvegetated tidal flats is generally lower than that of vegetated mangrove ecosystems (Alongi, 2014; Bouillon et al., 2008; Pendleton et al., 2012). Considering the current research priorities related to global mangrove restoration and blue carbon management, we prioritized sediment accretion rate assessments in mangrove ecosystems during the initial sampling design. In future studies, we will further improve the sampling framework by incorporating sediment accretion measurements in tidal flats, enabling a more comprehensive comparison of sedimentary processes and carbon burial contributions among different coastal ecosystems.
lines 125-135: Please provide citations for your soil parameter measurement protocols.	Thank you very much for your valuable comment. In the Methods section of the revised manuscript, we have added relevant references for the determination of soil physicochemical properties (Huang et al., 2025a, 2024) to clearly specify the standard analytical methods used for each parameter. In addition, all soil analyses were conducted by a certified third-party laboratory, while only sample pre-treatment was performed by our team, ensuring the standardization, reliability, and reproducibility of the data.
lines 134-135: It might be worthwhile to define SBD as DBD (dry bulk density) to be more specific about the kind of soil bulk density your measured.	Thank you very much for your valuable suggestion. In response, we have further clarified in the Methods section of the revised manuscript that “soil bulk density (SBD)” has been revised to “dry bulk density (DBD)” to more accurately describe the measured soil property. This revision helps avoid potential ambiguity and improves the clarity and consistency of the methodological description throughout the manuscript.
lines 158-159: I think this methodology is probably fine. Ideally, data exploration would involve a few more steps than listed here. A paper that provides a detailed protocol for data exploration with ecological data is Zuur, A. F., Ieno, E.N., and Elphick, C. S. 2010. A protocol for data exploration to avoid common statistical problems. Methods in Ecology and Evolution. https://doi.org/10.1111/j.2041-210X.2009.00001.x.	Thank you very much for your valuable comment. In response to your suggestion, we further consulted the ecological data exploration guidelines proposed by Zuur et al. (2010) and carefully reviewed our data analysis workflow. Prior to statistical analyses, we had already examined data distributions, potential outliers, and model assumptions to ensure the appropriateness of the applied statistical approaches. The data preprocessing and statistical analysis procedures were conducted following approaches described in previous studies (Xia et al., 2021; Zhang et al., 2022). In the revised manuscript, we have added the relevant references and further clarified these procedures in the Methods section to improve the transparency and reproducibility of the statistical analysis workflow.
line 191: Table 1, line 1, column 3 (High-urbanization natural mangroves, Salinity): 52 g kg ^-1 seems high for this part of the world. Please double-check this is not a typo. If this is correct, please discuss the high salinity in more depth in your discussion section. If this salinity level is a known feature of this particular area, please include that in section 2.1 Study area.	Thank you very much for your valuable comment. We have further checked the soil salinity data for highly urbanized natural mangroves in Table 1 and confirmed that the reported value is correct. The relatively high salinity may be related to the unique geographical and hydrological characteristics of the study area. Specifically, the Huizhou study site is located within a semi-enclosed drowned valley bay formed by the inland extension of Honghai Bay, where local water exchange is relatively restricted. These conditions may enhance evaporation–concentration processes and salt retention, leading to greater accumulation of salts in sediments. In addition, previous studies have reported that seawater salinity in this region generally exceeds 30 ppt (Aslam and Wang, 2025), indicating a naturally high-salinity background. The long-term presence of salt production activities in this area further reflects its distinctive saline environmental conditions. Therefore, although the measured soil salinity value is relatively high compared with other mangrove regions, it is reasonable considering the local hydrological setting and regional salinity characteristics. Following your suggestion, we have added relevant environmental information in Section 2.1 “Study area” of the revised manuscript to provide further context for interpreting this high salinity value.
lines 194-196: Figure 2 description: Please include units in all of your figure descriptions.	Thank you very much for your valuable suggestion. In response, we have revised the caption of Figure 2 in the revised manuscript by adding the units of all relevant variables to improve clarity, consistency, and standardization of figure presentation.
lines 198-199: Figure 3 description: Please include units in all of your figure descriptions. What do you mean by "different letters indicated significant differences in soil organic carbon among urbanization levels within the same ecosystem type..."? Define what the letters are and what each of them means, exactly.	Thank you very much for your comment. In the revised manuscript, we have added the units of all relevant variables in the caption of Figure 3 to improve completeness, clarity, and standardization of figure presentation. In addition, we have briefly clarified the meaning of the significance letters. Specifically, in Figure 3a, letters should be interpreted within each ecosystem type. For example, in restored mangroves, different letters (a and b) under high and moderate urbanization indicate a statistically significant difference in soil organic carbon between these two levels. The same interpretation applies to natural mangroves and bare tidal flats, where letters should be understood within each ecosystem category rather than across different ecosystem types.
line 208: Figure 4: It seems you have two main rows for these figure insets (Natural Mangroves and Restored Mangroves) and three main columns (High-urbanization, Moderate-urbanization, and Low-urbanization). Can you adjust the figure so it is more obvious that these insets all fit inside these different categories? Simply making the row and column titles bigger may achieve this or you could add a background grid or ribbons like you did in Figure 6.	Thank you very much for your valuable suggestion. In response, we have revised Figure 4 by enhancing the visual emphasis of row and column headers to improve structural clarity. In addition, we have further clarified the description of each panel in the figure caption to explicitly specify the corresponding ecosystem types and urbanization gradients. The revised caption is as follows: “Panels (a) and (b) show the vertical profiles of SOC burial rates and sediment ages in natural mangroves under high and low urbanization gradients, respectively. Panels (c), (d), and (e) present the corresponding profiles for restored mangroves under high, moderate, and low urbanization levels, respectively.” These revisions improve the clarity of grouping relationships and make the figure more intuitive and easier to interpret.
lines 209-210: Figure 4 description: Please provide more details in your figure description, such as units. This would also be a good place to clarify which of the Inset Figures are natural mangroves vs restored mangroves and high vs moderate vs low-urbanization	Thank you very much for your valuable suggestion. In response, we have revised Figure 4 by adding the units of all relevant variables in the figure caption to improve completeness and standardization. We have also further clarified the correspondence between ecosystem types and urbanization gradients for each panel, thereby enhancing the structural clarity and readability of the figure. The revised description is as follows: “Panels (a) and (b) show the vertical profiles of SOC burial rates and sediment ages in natural mangroves under high and low urbanization gradients, respectively. Panels (c), (d), and (e) present the corresponding profiles for restored mangroves under high, moderate, and low urbanization levels, respectively.” These revisions improve the clarity of grouping relationships and ensure more consistent interpretation of the figure content.
lines 208-210: What locations do Inset Figures A, B, C, D, and E correspond with (ex: Shantou, Zhanjiang, or Huizhou)? Please provide the location names in either the figure or in the figure description.	Thank you very much for your valuable comment. In response, we have further revised the caption of Figure 4 in the manuscript to explicitly clarify the correspondence between ecosystem types and urbanization gradients. The revised description is as follows: “Panels (a) and (b) show the vertical profiles of SOC burial rates and sediment ages in natural mangroves under high and low urbanization gradients, respectively. Panels (c), (d), and (e) present the corresponding profiles for restored mangroves under high, moderate, and low urbanization levels, respectively.” In addition, we have clarified that Huizhou, Shantou, and Zhanjiang represent high, moderate, and low urbanization levels, respectively, as stated in the Methods section. This information has been consistently applied throughout the manuscript to ensure coherence in figure interpretation.
lines 231-232: Figure 5 description: Please include more details in your figure description. What do the different colored boxes for your different depth intervals represent? Do Insets A, B, C, D, and E correspond with the same Insets in Figure 4? If so, it may be helpful to include the locations for each of these either in the figure or in the figure description.	Thank you very much for your valuable suggestion. In response, we have further refined the caption of Figure 5 in the revised manuscript to explicitly clarify the correspondence between ecosystem types and urbanization gradients. In addition, different colored boxes represent distinct soil depth intervals, and the corresponding depth ranges are clearly labeled on the right side of each box in Figure 5. These revisions improve the clarity, readability, and interpretability of the figure. Furthermore, the inset panels A, B, C, D, and E in Figure 5 correspond to the same sampling locations as those presented in Figure 4. Following your suggestion, we have added the relevant location information to the caption of Figure 5 to provide clearer descriptions of the spatial context.
lines 246-247: Figure 6 description: Please provide more details in this figure description. What do you mean by "different letters indicated significant differences in Beta values among urbanization levels within the same ecosystem type..."? Define what the letters are and what each of them means, exactly.	Thank you very much for your valuable comment. Here, we further clarify the meaning of the significance letters in Figure 6. The interpretation of letters should be conducted within each ecosystem type. For example, in restored mangroves, different letters (b and a) corresponding to high and moderate urbanization levels indicate a statistically significant difference in β-value between these two levels. The same interpretation applies to natural mangroves and bare tidal flats, where letters should be interpreted within each ecosystem category rather than across different ecosystem types.
line 270: “Once those limits are crossed, previously buried horizons return to active cycling and carbon stock decline rapidly…” Provide citation.	Thank you very much for your valuable comment. In response to your suggestion, we have added relevant references in the revised manuscript to provide further support for this statement. These additional citations strengthen the scientific basis of the discussion and improve the reliability of the interpretation.

Citation: https://doi.org/10.5194/egusphere-2026-2512-AC1

RC2: 'Comment on egusphere-2026-2512', Anonymous Referee #2, 01 Jun 2026

This MS addresses an important and timely question and is supported by a valuable field dataset. The integration of carbon stocks, sources, and stability into a single framework is interesting, and the finding that shifts in carbon sources and stability may precede carbon stock losses is potentially important. Generally the MS is well written. However, some aspects of the conceptual framing and interpretation could be strengthened, particularly regarding the attribution of observed patterns to urbanization and the consideration of alternative explanations.

1. Abstract

1) The same underlying concepts are introduced using several different expressions (e.g., "how much carbon is stored, who supplies it, and how long it persists", "stocks, sources, and stability", and the "triad framework"). While these terms appear to refer to the same ideas, switching terminology within a short abstract makes the narrative harder to follow. I suggest introducing the preferred terminology early and using it consistently throughout.

2) L22–25 The transition from the main findings to the introduction of the framework feels somewhat abrupt. A brief linking sentence explaining how the findings led to the development of the framework would help improve the logical flow.

3) L25–27 The terms "connectivity" and "accretion" are introduced only in the final sentence of the abstract. Because they have not been mentioned previously, their role in the proposed framework is not immediately clear. Consider introducing them earlier or briefly explaining how they contribute to the diagnostic framework.

2. Introduction

This part does a good job of motivating the study. The authors clearly identify two important gaps: (1) stock, source and stability components of blue carbon have largely been studied separately, and (2) we still know relatively little about how these components vary across habitats along urbanization gradients. However, I think several aspects of the conceptual development could be strengthened.

Specific Comments:

1) One issue is that the Introduction discusses several specific stabilization mechanisms (e.g. MAOM and Fe/Al-bound carbon), but these mechanisms are not clearly linked to the hypotheses. As a result, the hypotheses feel somewhat disconnected from the framework developed in the preceding paragraphs. It would be helpful to add a short mechanistic explanation of why urbanization is expected to affect these stabilization pathways, and how this leads to the predictions being tested.

2) I also found the relationship between the urbanization gradient (low, moderate and high) and the habitat categories (natural, restored and tidal flat) somewhat unclear. These represent different ecological dimensions, and it would be useful to explain more explicitly how they fit together within the conceptual framework of the study.

Related to this, the three study regions are located in different parts of the Guangdong coastline and may differ in important environmental characteristics such as tidal regimes, hydrodynamics and sedimentation processes. A brief discussion of why these locations can be treated as an urbanization gradient, and how potential geographic confounding is addressed, would strengthen the rationale of the study.

3) I found H3 to be relatively broad compared with the mechanistic framework presented earlier in the Introduction. The prediction that lower urban pressure leads to greater stability is intuitively reasonable, but it does not fully exploit the habitat-specific framework developed in Paragraph 4) I think the hypothesis would be more interesting if it made a more specific prediction about how different habitat types are expected to respond to urbanization, particularly whether restored habitats are expected to converge towards natural reference systems under lower levels of urban pressure.

5) L50-55 it’s better to add the references

3. Discussion

1) Throughout Section 4, many paragraphs begin with phrases such as "Our results suggest...", "Our data show...", or "Our results indicate...". While not problematic individually, their repeated use becomes noticeable and can make it difficult to identify which analyses support a given conclusion. Given the diverse evidence presented in this study (SOC stocks, ^210Pb dating, stable isotopes, and turnover metrics), I encourage the authors to refer more directly to the relevant indicators or analytical approaches when discussing specific findings.

2) In Sections 4.1 and 4.2, the discussion links the stronger decline in carbon stocks and the shift towards more labile carbon sources in restored stands and tidal flats largely to urbanization. However, some of the mechanisms invoked, such as young substrates, lower structural complexity, and underdeveloped root systems, are also characteristic of these habitats regardless of urban pressure. The discussion would benefit from more clearly distinguishing habitat maturity effects from urbanization effects.

3) The study design captures a clear urbanization gradient across the three regions. However, these locations also differ in their geographic and geomorphic settings and may vary in hydrodynamics, sediment characteristics, and natural accretion processes. Some observed differences in carbon burial and stability may therefore reflect regional environmental variation in addition to urbanization. A brief discussion of these potentially confounding factors would strengthen the interpretation of the results.

4) Section 4.4 introduces the concepts of "carbon anchors" and "instability fronts" as contrasting system states. While these categories provide a useful synthesis of the observed patterns, their broader applicability may currently be overstated. The study includes multiple habitat types and urbanization levels, and other combinations not represented by these end-members could potentially produce different responses. I suggest presenting these states as scenarios identified within this study system rather than broadly transferable categories.

5) The proposed framework is built around responses across habitat types and urbanization levels. However, the absence of natural mangroves at the moderate urbanization level leaves one part of the conceptual matrix unrepresented. It would be helpful to acknowledge this limitation and discuss how it affects interpretation of the proposed framework and its generality.

Citation: https://doi.org/10.5194/egusphere-2026-2512-RC2

AC2: 'Reply on RC2', Fen Guo, 14 Jul 2026

General Comments

Comments	Responses
This MS addresses an important and timely question and is supported by a valuable field dataset. The integration of carbon stocks, sources, and stability into a single framework is interesting, and the finding that shifts in carbon sources and stability may precede carbon stock losses is potentially important. Generally the MS is well written	We sincerely thank you for your positive evaluation of our manuscript.
However, some aspects of the conceptual framing and interpretation could be strengthened, particularly regarding the attribution of observed patterns to urbanization and the consideration of alternative explanations.	We sincerely thank you for raising this important comment. Following your suggestion, we further refined the overall logic of the manuscript, with particular emphasis on clarifying the conceptual boundaries of our framework and avoiding overinterpretation of causal relationships. Specifically, we clarified the conceptual relationships among the three core dimensions of the triad framework, namely carbon stocks, organic carbon source contributions, and carbon stability, and emphasized that their coordinated variations collectively represent the integrated responses of blue carbon ecosystems along urbanization gradients. In the Discussion, we further considered the potential influence of regional environmental differences, including tidal characteristics, hydrodynamic conditions, and sediment supply, on the interpretation of our observations. Accordingly, we moderated statements implying direct causal effects of urbanization on carbon dynamics and clarified that our conclusions primarily reflect associations identified from a natural gradient comparison design. Furthermore, we clarified that “carbon anchors” and “instability fronts” do not represent universally applicable ecosystem states or classifications, but rather conceptual scenarios summarizing the observed combinations of habitat types and urbanization levels within our study system. These revisions have improved the applicability and interpretative rigor of our conceptual framework.
2. Introduction: This part does a good job of motivating the study. The authors clearly identify two important gaps: (1) stock, source and stability components of blue carbon have largely been studied separately, and (2) we still know relatively little about how these components vary across habitats along urbanization gradients. However, I think several aspects of the conceptual development could be strengthened.	We sincerely appreciate your recognition and positive comments regarding the conceptual framework and underlying motivation of this study. Following your suggestion, we further clarified the conceptual relationships among the three core dimensions of the triad framework, namely carbon stocks, organic carbon source contributions, and carbon stability. We also further clarified the relationship between urbanization gradients and ecosystem types by explicitly stating that urbanization levels represent gradients of anthropogenic disturbance intensity at the regional scale, whereas natural mangroves, restored mangroves, and unvegetated tidal flats represent distinct habitat states. In addition, we refined the study hypotheses to better reflect habitat-specific responses. Specifically, H3 has been revised as follows: “Soil-carbon stability is expected to be higher under lower urbanization pressure, but this effect should be strongest in natural mangroves and weaker or more variable in restored mangroves and tidal flats because of differences in belowground development, geomorphic maturity, and source stability.” These revisions have improved the conceptual clarity of our framework and strengthened the alignment between our hypotheses, and study design. We appreciate your constructive suggestion and hope that the revised manuscript will be suitable for publication.

Specific Comments

Comments	Responses
1. Abstract: The same underlying concepts are introduced using several different expressions (e.g., "how much carbon is stored, who supplies it, and how long it persists", "stocks, sources, and stability", and the "triad framework"). While these terms appear to refer to the same ideas, switching terminology within a short abstract makes the narrative harder to follow. I suggest introducing the preferred terminology early and using it consistently throughout.	We sincerely appreciate the valuable comment. We have carefully reviewed and revised the relevant statements in the Abstract of the revised manuscript. Specifically, we have further clarified the conceptual relationships among the three core dimensions of soil organic carbon characteristics, including carbon stocks, carbon sources, and carbon stability. At the beginning of the Abstract, we introduced consistent terminology definitions to improve the connections among different expressions, avoid potential ambiguity, and enhance the overall logical coherence and readability. Meanwhile, we have retained the term “triad framework” as the name of the conceptual framework to emphasize that it integrates three complementary dimensions of soil organic carbon characteristics: stocks, sources, and stability. This revision improves the clarity and consistency of the conceptual description throughout the Abstract.
1. Abstract: L22–25 The transition from the main findings to the introduction of the framework feels somewhat abrupt. A brief linking sentence explaining how the findings led to the development of the framework would help improve the logical flow.	Thank you very much for your valuable comment. We have revised the Abstract to improve its logical coherence and readability by adding transitional statements between the main findings and the conceptual framework. Specifically, the original sentence “We introduced a triad framework that treats stocks, sources, and stability as coupled state variables along the urbanization gradient and identified two reproducible system states: carbon anchors in low urbanization natural mangroves and instability fronts in restored stands and tidal flats.” has been revised to “These concurrent changes motivated the development of a triad framework that treats stocks, sources, and stability as coupled state variables along the urbanization gradient. Within this framework, we identified two representative ecosystem scenarios: carbon anchors in low urbanization natural mangroves and instability fronts in restored stands and tidal flats.”. These additions will clarify how the results support the proposed framework, thereby strengthening the overall flow and clarity of the Abstract.
1. Abstract: L25–27 The terms "connectivity" and "accretion" are introduced only in the final sentence of the abstract. Because they have not been mentioned previously, their role in the proposed framework is not immediately clear. Consider introducing them earlier or briefly explaining how they contribute to the diagnostic framework.	Thank you very much for your valuable comment. Considering the word limit of the Abstract and that the core focus of this framework is to integrate three dimensions of soil organic carbon characteristics, namely organic carbon stocks, organic carbon sources, and organic carbon stability, we have retained “connectivity” and “accretion” within the framework application section rather than introducing these auxiliary concepts earlier in the Abstract. Following your suggestion, we have revised the related descriptions to improve the logical transitions and overall readability of the Abstract. These modifications help clarify the relationships among the framework components while avoiding unnecessary conceptual complexity, thereby enhancing the accessibility and coherence of the Abstract.
2. Introduction: One issue is that the Introduction discusses several specific stabilization mechanisms (e.g. MAOM and Fe/Al-bound carbon), but these mechanisms are not clearly linked to the hypotheses. As a result, the hypotheses feel somewhat disconnected from the framework developed in the preceding paragraphs. It would be helpful to add a short mechanistic explanation of why urbanization is expected to affect these stabilization pathways, and how this leads to the predictions being tested	We sincerely thank you for raising this important point. We agree that the original manuscript did not sufficiently clarify the link between the discussion of stabilization mechanisms, such as mineral-associated organic matter (MAOM) and Fe/Al-bound carbon, and the hypotheses proposed in this study, which may have weakened the logical connection between the theoretical background and our predictions. Following your suggestion, we have added further explanations of these mechanisms in the revised Introduction. Specifically, human disturbances may influence organic matter–mineral interactions by altering organic matter inputs, sedimentary environmental conditions, and redox states, thereby regulating carbon preservation and turnover processes. These potential stabilization processes may affect SOC turnover dynamics and provide a theoretical basis for interpreting the SOC turnover rate variations assessed using the β parameter in this study.
2. Introduction: I also found the relationship between the urbanization gradient (low, moderate and high) and the habitat categories (natural, restored and tidal flat) somewhat unclear. These represent different ecological dimensions, and it would be useful to explain more explicitly how they fit together within the conceptual framework of the study. Related to this, the three study regions are located in different parts of the Guangdong coastline and may differ in important environmental characteristics such as tidal regimes, hydrodynamics and sedimentation processes. A brief discussion of why these locations can be treated as an urbanization gradient, and how potential geographic confounding is addressed, would strengthen the rationale of the study.	Thank you very much for this important comment. We agree that urbanization intensity and habitat type represent two distinct ecological dimensions and that their relationship required clearer explanation. We have therefore revised the Introduction to clarify that urbanization represents the regional intensity of anthropogenic pressure, whereas natural mangroves, restored mangroves, and unvegetated tidal flats represent contrasting habitat states. Our conceptual framework evaluates how soil-carbon stocks, source composition, and stability vary among these habitat types under different levels of urban pressure, rather than treating habitat type as a component of the urbanization gradient. We have also expanded the rationale for selecting Zhanjiang, Shantou, and Huizhou as representatives of low, moderate, and high urbanization, respectively. Urbanization intensity was quantified using four socioeconomic indicators: the proportion of the urban population, the proportion of the non-agricultural population, per capita disposable income of urban residents, and built-up area. These indicators consistently distinguished the three cities along an increasing urbanization gradient. We nevertheless recognize that the three cities occupy different sections of the Guangdong coastline and may differ in tidal regime, hydrodynamics, sediment supply, and other environmental characteristics. Because each urbanization level is represented by one city, geographic variation cannot be completely separated from urbanization in this observational design. We have now stated this limitation explicitly and have revised the manuscript to avoid interpreting the observed patterns as effects of urbanization independent of all regional environmental differences. Instead, we interpret them as habitat-specific soil-carbon responses associated with contrasting levels of urban development across subtropical coastal settings. Where available, measured local environmental variables were incorporated into the analyses and interpretation to reduce the influence of site-level heterogeneity. These revisions clarify how the two dimensions fit within the study design and provide a more transparent account of the strengths and limitations of treating the three regions as an urbanization gradient.
2. Introduction: I found H3 to be relatively broad compared with the mechanistic framework presented earlier in the Introduction. The prediction that lower urban pressure leads to greater stability is intuitively reasonable, but it does not fully exploit the habitat-specific framework developed in Paragraph 4) I think the hypothesis would be more interesting if it made a more specific prediction about how different habitat types are expected to respond to urbanization, particularly whether restored habitats are expected to converge towards natural reference systems under lower levels of urban pressure.	We sincerely thank you for this valuable suggestion. We agree that the original formulation of H3 was relatively broad and did not sufficiently emphasize the habitat-specific responses implied by our conceptual framework. In response, we have revised H3 to make the prediction more explicit across habitat types. The revised hypothesis now states that soil-carbon stability is expected to be higher under lower urbanization pressure, but that this effect should be strongest in natural mangroves and weaker or more variable in restored mangroves and tidal flats because of differences in belowground development, geomorphic maturity, and source stability. We also carefully considered the suggestion regarding whether restored mangroves converge toward natural reference systems under lower urbanization pressure. We agree that this is an important and interesting question. However, testing convergence would require a sampling design with comparable natural and restored mangroves across all urbanization levels, particularly including natural mangroves under moderate urbanization. Because such sites were not available in the present dataset, we cannot robustly test convergence toward natural reference conditions without exceeding the scope of our study. Accordingly, we have revised the hypothesis and the corresponding text in the Introduction to better reflect habitat-specific expectations, while avoiding an unsupported convergence claim. We have also noted in the Discussion that future studies with balanced natural and restored reference systems across urbanization gradients would be valuable for directly testing restoration convergence.
2. Introduction: L50-55 it’s better to add the references.	References have been added.
3. Discussion: Throughout Section 4, many paragraphs begin with phrases such as "Our results suggest...", "Our data show...", or "Our results indicate...". While not problematic individually, their repeated use becomes noticeable and can make it difficult to identify which analyses support a given conclusion. Given the diverse evidence presented in this study (SOC stocks, ²¹⁰Pb dating, stable isotopes, and turnover metrics), I encourage the authors to refer more directly to the relevant indicators or analytical approaches when discussing specific findings.	Thank you very much for your valuable and constructive comment. We have systematically revised the relevant sections of the Discussion in the manuscript by reducing the repetitive use of generic expressions such as “our results suggest/indicate,” and by explicitly linking each statement to its corresponding evidence to improve clarity and readability. For example, “Our results suggest…” has been revised to “Results of soil organic carbon stocks and burial rates suggest…,” while “Our results indicate…” has been revised to “Soil organic carbon turnover rate results indicate…”. These revisions enhance the precision of interpretation and improve the logical flow between different analytical results.
3. Discussion: In Sections 4.1 and 4.2, the discussion links the stronger decline in carbon stocks and the shift towards more labile carbon sources in restored stands and tidal flats largely to urbanization. However, some of the mechanisms invoked, such as young substrates, lower structural complexity, and underdeveloped root systems, are also characteristic of these habitats regardless of urban pressure. The discussion would benefit from more clearly distinguishing habitat maturity effects from urbanization effects	Thank you very much for your valuable comment. In response, we have revised the Discussion section of the manuscript to further distinguish between the effects of habitat maturity (or successional stage) and urbanization on soil organic carbon stocks and carbon source composition.
3. Discussion: The study design captures a clear urbanization gradient across the three regions. However, these locations also differ in their geographic and geomorphic settings and may vary in hydrodynamics, sediment characteristics, and natural accretion processes. Some observed differences in carbon burial and stability may therefore reflect regional environmental variation in addition to urbanization. A brief discussion of these potentially confounding factors would strengthen the interpretation of the results.	Thank you very much for this important comment. We agree that the three study regions differ not only in urbanization intensity but also in geographic and geomorphic setting, including tidal regime, hydrodynamics, sediment supply, sediment properties, and natural accretion processes. Because each urbanization level is represented by a single region, these regional differences cannot be fully separated from urbanization in the present observational design. We have therefore revised the Discussion to acknowledge this potential geographic confounding and to temper causal interpretations of the observed patterns. Specifically, we now interpret the differences in carbon burial and stability as responses associated with contrasting urbanization contexts rather than as effects attributable exclusively to urbanization. We also discuss how variation in hydrodynamics, sediment characteristics, and natural accretion may have contributed to the differences among regions. This limitation does not negate the consistent changes observed across carbon stocks, sources, and stability, but it constrains the extent to which the relative contribution of urbanization can be isolated from regional environmental variation. Future studies should replicate multiple independent coastal systems within each urbanization category and incorporate direct measurements of tidal dynamics, sediment supply, and geomorphic setting to separate urbanization effects from geographic heterogeneity more robustly.
3. Discussion: Section 4.4 introduces the concepts of "carbon anchors" and "instability fronts" as contrasting system states. While these categories provide a useful synthesis of the observed patterns, their broader applicability may currently be overstated. The study includes multiple habitat types and urbanization levels, and other combinations not represented by these end-members could potentially produce different responses. I suggest presenting these states as scenarios identified within this study system rather than broadly transferable categories.	Thank you very much for your valuable comment. In response, we have revised the Discussion section to clarify that “carbon anchors” and “instability fronts” are defined as typical scenarios identified within the framework of this study, used to summarize the key response patterns observed across different habitat types and urbanization levels, rather than representing universally applicable system states or broadly generalizable classifications. This revision further clarifies the scope of the conceptual framework and avoids overextending its general applicability.
3. Discussion: The proposed framework is built around responses across habitat types and urbanization levels. However, the absence of natural mangroves at the moderate urbanization level leaves one part of the conceptual matrix unrepresented. It would be helpful to acknowledge this limitation and discuss how it affects interpretation of the proposed framework and its generality.	Thank you very much for this important comment. We agree that the absence of natural mangroves at the moderate urbanization level leaves the habitat by urbanization matrix incomplete. We have therefore revised the Discussion to acknowledge that the present design cannot fully characterize the continuous response of natural mangrove soil carbon across the entire urbanization gradient or determine whether the carbon anchor state weakens gradually or shifts abruptly at intermediate levels of urban pressure. This limitation does not invalidate the observed contrasts among the sampled habitat and urbanization combinations, but it restricts our ability to estimate habitat by urbanization interactions and to generalize the proposed framework as a fully resolved response model. We now clarify that the framework should be interpreted as a conceptual diagnostic derived from the available observational evidence, rather than as a complete factorial representation of all habitat states across all urbanization levels. Future studies should adopt a fully crossed design that includes natural and restored mangroves and tidal flats at each urbanization level, ideally with multiple independent regions within each category. Such replication would allow the continuity, thresholds, and generality of the proposed carbon anchor and instability front states to be tested more rigorously.

Citation: https://doi.org/10.5194/egusphere-2026-2512-AC2

RC3: 'Comment on egusphere-2026-2512', Anonymous Referee #3, 03 Jun 2026

This manuscript examines how urbanization affects soil organic carbon stocks, source composition, and stability in natural mangroves, restored mangroves, and tidal flats across three Guangdong cities classified at low, moderate, and high urbanization levels. The topic is timely, and the three-habitat comparison is useful.

The study attributes observed differences in SOC stocks, sources, and stability to urbanization, but the Methods do not specify how the urbanization effect is identified or tested.

The MixSIAR model considers only mangrove leaf tissue and marine phytoplankton but overlooks terrestrial inputs (see the terrestrial end-member from https://doi.org/10.1021/acs.est.3c00012).

The phrasing "composition of organic-carbon sources" (H2, line 75) is awkward. The intended meaning is the relative contribution of different sources to soil organic carbon, not the composition of the sources themselves. Consider replacing with "source composition of soil organic carbon" or simply "organic carbon source contributions," and apply the revised phrasing consistently across the abstract, hypotheses, and Discussion.

The Introduction and Discussion invoke mineral-associated organic matter fractions, Fe/Al-bound carbon, thermal indices, and radiocarbon ages as relevant stability metrics, yet none were measured (see the stabilization mechanisms documented in https://doi.org/10.1111/gcb.70763). The Discussion should more clearly separate what was measured from what is inferred by analogy with the cited literature.

The linear mixed-effects model structure is unreported. The Methods state that LMMs were used but do not specify the fixed-effect structure, the random-effect structure, or the grouping variable.

The phytoplankton sampling (five 500 mL samples per site, line 116) is described, but the processing steps (filtration, drying, isotope measurement) are not. What organism-level end-member values were used in MixSIAR? Were they site-specific or pooled?

Citation: https://doi.org/10.5194/egusphere-2026-2512-RC3

AC3: 'Reply on RC3', Fen Guo, 14 Jul 2026

General Comments

Comments	Responses
This manuscript examines how urbanization affects soil organic carbon stocks, source composition, and stability in natural mangroves, restored mangroves, and tidal flats across three Guangdong cities classified at low, moderate, and high urbanization levels. The topic is timely, and the three-habitat comparison is useful.	We sincerely thank you for your positive evaluation and kind recognition of our study.

Specific Comments

Comments	Responses
The study attributes observed differences in SOC stocks, sources, and stability to urbanization, but the Methods do not specify how the urbanization effect is identified or tested.	Thank you very much for this important comment. We agree that the original Methods did not state sufficiently clearly how urbanization was operationalized as an explanatory variable or how its association with soil-carbon properties was statistically evaluated. We have therefore revised the Methods to distinguish explicitly between the construction of the urbanization gradient and the statistical tests used to assess responses along that gradient. Urbanization intensity was quantified using four socioeconomic indicators, including the proportion of the urban population, the proportion of the non-agricultural population, per capita disposable income of urban residents, and built-up area. These indicators consistently ranked Zhanjiang, Shantou, and Huizhou as representing low, moderate, and high urbanization contexts, respectively. Urbanization level was then included in the analyses as a three-level categorical explanatory variable. Because habitat type and urbanization represent distinct dimensions of the study design, urbanization-related differences were evaluated within each habitat type rather than inferred from pooled comparisons among habitats. Differences in soil organic carbon stocks and stability among urbanization levels were tested using linear mixed-effects models (LMMs), followed by post-hoc multiple comparisons. Differences in source composition were evaluated using MixSIAR. We have now added the corresponding model structures, response variables, fixed effects, random effects where applicable, and post hoc comparisons to the Methods. We also clarify that each urbanization level is represented by one coastal region. The analyses therefore identify associations between soil-carbon characteristics and contrasting urbanization contexts, but they do not isolate a causal urbanization effect independently of all regional environmental variation. We have revised both the Methods and Discussion to make this inferential scope explicit.
The MixSIAR model considers only mangrove leaf tissue and marine phytoplankton but overlooks terrestrial inputs (see the terrestrial end-member from https://doi.org/10.1021/acs.est.3c00012).	Thank you very much for your valuable comment. We fully agree that allochthonous organic matter may represent an important source of soil organic carbon in some coastal ecosystems. In the revised manuscript, the selection of end-members for the MixSIAR model was primarily based on previous studies Huang et al. (2025), Zhang et al. (2024) and Zhang et al. (2022), which consistently highlight the key contributions of mangrove-derived and marine-derived organic matter in blue carbon ecosystems. We do not exclude the potential contribution of terrestrial organic matter; however, because no terrestrial end-member samples were collected in this study, site-specific isotopic signatures for terrestrial inputs were not available. Thank you for raising this important issue. We agree that terrestrial organic matter may contribute to soil organic carbon in coastal wetlands and should not be assumed to be absent from mangrove sediments. However, we do not consider terrestrial organic matter to be a statistically identifiable third end-member in the present MixSIAR model. Mangroves and most terrestrial woody vegetation both use the C3 photosynthetic pathway and therefore show strongly overlapping carbon isotope signatures, commonly centred near minus 28 per mille (Kohn, 2010; Kristensen et al., 2008). Their nitrogen isotope values are also highly variable and may overlap because they are influenced by nitrogen sources, nutrient conditions, and post-depositional processing. Consequently, adding a literature-derived terrestrial end-member to our dual-isotope model would introduce substantial end-member overlap and would not allow reliable separation of terrestrial C3 material from mangrove-derived material. Under such conditions, estimated contributions may become strongly dependent on the specified priors rather than being resolved by the isotope data. Combining isotopically indistinguishable and ecologically related sources is an established approach in stable-isotope mixing models. Previous studies of mangrove sediments have therefore frequently used two broad source groups, namely mangrove or vascular-plant material and estuarine or marine suspended organic matter, to distinguish locally produced plant carbon from externally supplied aquatic carbon (Bouillon et al., 2003; Kristensen et al., 2008; Suello et al., 2022). The terrestrial end-member used by Li et al. (2023) is informative for demonstrating the potential importance of allochthonous inputs, but its isotopic values are site dependent and cannot be transferred directly to our study sites as an independently resolvable source. We have therefore retained the two-source model but revised the interpretation of the plant end-member. It is now defined as C3 vascular plant-derived carbon, represented by locally collected mangrove tissues, rather than as carbon that can be attributed exclusively to mangroves. Accordingly, references to mangrove-derived carbon in the Methods, Results, and Discussion have been revised to mangrove-dominated C3 vascular plant-derived carbon where appropriate. We also clarify that the model primarily distinguishes C3 vascular plant inputs from phytoplankton-derived inputs and cannot exclude a contribution from isotopically similar terrestrial C3 vegetation.
The phrasing "composition of organic-carbon sources" (H2, line 75) is awkward. The intended meaning is the relative contribution of different sources to soil organic carbon, not the composition of the sources themselves. Consider replacing with "source composition of soil organic carbon" or simply "organic carbon source contributions," and apply the revised phrasing consistently across the abstract, hypotheses, and Discussion.	Thank you for this helpful comment. We agree that “composition of organic-carbon sources” could be misinterpreted as referring to the properties of the source materials rather than their relative contributions to soil organic carbon. We have therefore replaced this phrase with “organic carbon source contributions” throughout the Abstract, hypotheses, Results, and Discussion. This revision more accurately describes the quantity estimated by the mixing model and ensures consistent terminology throughout the manuscript.
The Introduction and Discussion invoke mineral-associated organic matter fractions, Fe/Al-bound carbon, thermal indices, and radiocarbon ages as relevant stability metrics, yet none were measured (see the stabilization mechanisms documented in https://doi.org/10.1111/gcb.70763). The Discussion should more clearly separate what was measured from what is inferred by analogy with the cited literature.	Thank you for this important comment. We agree that the original manuscript did not sufficiently distinguish the stability indicator evaluated in this study from stabilization mechanisms documented in the literature. In the present study, soil-carbon stability was assessed only from the fitted β parameter, which we use as a relative proxy for apparent SOC turnover. We did not measure mineral-associated organic matter, Fe- or Al-bound carbon, thermal stability indices, molecular recalcitrance, or radiocarbon ages. We have therefore revised the Introduction, Discussion, and Conclusions to state this distinction explicitly. We have also removed or qualified statements that could be interpreted as direct evidence for changes in mineral protection, thermal recalcitrance, or carbon age. The revised Discussion now interprets declining β values as evidence of faster apparent SOC turnover under increasing urbanization. Mineral association, Fe and Al binding, molecular composition, and microbial processing are discussed only as potential mechanisms identified by previous studies, including Li et al. (2026), rather than as processes demonstrated by our data. We further clarify that the available measurements cannot determine which stabilization mechanism caused the observed turnover patterns.
The linear mixed-effects model structure is unreported. The Methods state that LMMs were used but do not specify the fixed-effect structure, the random-effect structure, or the grouping variable	Thank you very much for your valuable comment. In response, we have revised the Methods section of the manuscript to include a complete description of the linear mixed-effects model (LMM) structure, explicitly specifying the fixed and random effects.
The phytoplankton sampling (five 500 mL samples per site, line 116) is described, but the processing steps (filtration, drying, isotope measurement) are not. What organism-level end-member values were used in MixSIAR? Were they site-specific or pooled?	Thank you for this important comment. We agree that the original Methods did not provide sufficient information on phytoplankton processing or on how the source end-members were parameterized in MixSIAR. We have now added the complete phytoplankton processing procedure. At each site, five 500 mL water samples were processed separately and treated as analytical replicates. Samples were filtered through glass microfiber filters, and the retained material was rinsed to remove dissolved salts. Filters were dried using a freeze dryer for 3–5 days, homogenized, and analysed for carbon and nitrogen stable isotopes using DELTA V Advantage isotope ratio mass spectrometer. We have also clarified the quality-control procedures and isotope-reference standards used during analysis. For MixSIAR, the phytoplankton and plant end-members were not represented by single organism-level measurements. Instead, isotope measurements from multiple samples were used to estimate the mean and standard deviation of each source group. The phytoplankton end-member was δ¹³C = [-18.46 ± 5.27]‰ and δ¹⁵N = [10.66 ± 2.06]‰ based on 39 samples, whereas the C3 vascular plant end-member represented by locally collected mangrove leaves was δ¹³C = [-28.57 ± 1.37]‰ and δ¹⁵N = [7.66 ± 1.38]‰ based on 72 samples. These values and their associated variation are now reported explicitly in the Methods and Supplementary Information. The end-members were pooled across sampling locations rather than specified separately for each site. We adopted this regional parameterization because the regional-scale objective of the study. Importantly, the among-site variation was retained through the source standard deviations entered into MixSIAR, rather than reducing each source to a single fixed value. We have clarified that this approach assumes that the broad isotopic characteristics of the two source groups are regionally comparable and may not capture all local variation in end-member signatures. This limitation has now been acknowledged in the Discussion.

Citation: https://doi.org/10.5194/egusphere-2026-2512-AC3

RC4: 'Comment on egusphere-2026-2512', Anonymous Referee #4, 14 Jun 2026

Huang et al. investigated the coupled shifts in soil-carbon stocks, sources, and stability across natural and restored mangroves and tidal flats under urbanization. This paper demonstrated that urbanization not only reduces carbon stocks and burial rates but also shifts the carbon source from mangrove-dominated to unstable phytoplankton-dominated inputs, while accelerating carbon turnover. This offers value for understanding the degradation mechanisms of blue carbon functions in the context of urbanization and provides new indicators for coastal blue carbon management. I have some comments for the authors.

lines 94–96: The article mentions that restored mangroves cover 2354.46 ha, but does not specify the planting years, species composition, or restoration methods. Such information is essential for understanding differences in carbon burial, especially given the higher SOC stock observed in restored mangroves in low‑urbanization areas. It is suggested to add restoration history data in Methods section 2.2 or as an appendix.

lines 118–120: The study assumes that “under the same urbanization level, the same type of mangroves have comparable sedimentation rates,” yet only one site per type was dated. This assumption may overlook within‑site and between‑site spatial heterogeneity. It is advised to discuss the risks associated with this assumption or provide preliminary evidence to enhance credibility.

lines 120–121: This study did not measure sedimentation rates on tidal flats, so SOC burial rates cannot be calculated, and only stocks are compared. However, in the discussion, the term “burial” is repeatedly used when comparing carbon dynamics between tidal flats and mangroves. It is advised to explicitly distinguish between stocks and burial rates in the title, abstract, and discussion to avoid conceptual confusion.

lines 175–178: The β value is used as a proxy for SOC turnover rate, but the cited literature supporting its validity in mangrove‑tidal flat systems is limited, mainly citing Acton et al., 2013 on a temperate soil study.

lines 259–271: The authors propose that carbon stocks decline nonlinearly along the urbanization gradient, attributing this to connectivity and vertical accretion exceeding “habitat‑specific thresholds.” However, no specific threshold values or statistical evidence are provided. It is suggested to add a threshold detection method or rephrase “nonlinear” as a more qualitative “accelerated decline” to avoid overstatement.

Figure 4 shows vertical profiles of SOC burial rate and sediment age, but age data come from only one site per type, whereas burial rates are averages of multiple sites. This combined presentation may misleadingly suggest that age and burial rate derive from the same spatial scale. It is recommended to clearly state the difference in data sources in the figure caption or to present them separately.

Citation: https://doi.org/10.5194/egusphere-2026-2512-RC4

AC4: 'Reply on RC4', Fen Guo, 14 Jul 2026

General Comments

Comments

Responses

We sincerely thank the referee for the positive evaluation of our manuscript and for recognizing our findings on changes in soil carbon stocks, sources, and turnover under urbanization. We greatly appreciate your encouraging comments and constructive support.

Specific Comments

Comments	Responses
lines 94–96: The article mentions that restored mangroves cover 2354.46 ha, but does not specify the planting years, species composition, or restoration methods. Such information is essential for understanding differences in carbon burial, especially given the higher SOC stock observed in restored mangroves in low‑urbanization areas. It is suggested to add restoration history data in Methods section 2.2 or as an appendix.	Thank you for this important comment. We agree that restoration age, species composition, and restoration method can strongly influence sediment development, belowground biomass accumulation, and soil-carbon burial, and are therefore important for interpreting differences among restored mangrove sites. Before field sampling, we searched publicly available records from local governments, nature reserves, and forestry authorities and consulted local site managers. However, restoration projects in the three regions were implemented at different times and by different agencies, and complete, consistent, and independently verifiable records were not available for all sites. In particular, exact planting years and detailed restoration methods could not be reconstructed reliably for several restored stands. Because no consistent site-level restoration records could be verified, we were unable to provide a comparable restoration-history dataset across the three regions. We have revised the Discussion to acknowledge that unmeasured variation in restoration age, species composition, and restoration method may partly explain differences in soil-carbon stocks and burial among restored sites. We have therefore tempered the interpretation of the relatively high soil-carbon stock observed in restored mangroves under low urbanization and no longer attribute this pattern solely to urbanization. Instead, we interpret it as the combined outcome of urbanization context, site conditions, and potentially restoration history. Future studies should incorporate standardized restoration chronologies and stand-development metrics across sites to separate the effects of urbanization from those of restoration age, species selection, and restoration method.
lines 118–120: The study assumes that “under the same urbanization level, the same type of mangroves have comparable sedimentation rates,” yet only one site per type was dated. This assumption may overlook within‑site and between‑site spatial heterogeneity. It is advised to discuss the risks associated with this assumption or provide preliminary evidence to enhance credibility.	Thank you for raising this important methodological concern. We agree that the use of one ²¹⁰Pb-dated core to represent the sediment accretion rate of a given habitat type within each urbanization region does not capture potential within-site or among-site spatial heterogeneity. We have revised the Methods to clarify that the sediment accretion rate obtained from each dated core was assigned to replicate plots of the same habitat type and urbanization context when calculating soil organic carbon burial rates. This procedure was adopted because the cost and analytical requirements of ²¹⁰Pb dating prevented the establishment of replicated chronologies for every site. We now describe this explicitly as an extrapolation assumption rather than as evidence that sedimentation rates are spatially uniform. Measurements of soil organic carbon concentration, bulk density, physicochemical properties, and stable isotopes were based on replicated plots and cores and therefore capture spatial variability in these soil properties. However, this replication does not resolve spatial variation in sediment accretion. Consequently, the estimated carbon burial rates should be interpreted as regional habitat-level approximations and may be overestimated or underestimated where local accretion rates differ from those of the representative dated core. We have added this uncertainty to the Discussion and tempered statements concerning differences in carbon burial among urbanization levels and habitat types. The results for carbon stocks and source contributions are supported by replicated soil sampling, whereas conclusions based on burial rates are subject to greater uncertainty because sediment accretion was not independently replicated. Previous studies have applied representative-core approaches in regional mangrove assessments, but we agree that such studies do not demonstrate spatial homogeneity in sedimentation within our study regions. Future work should therefore include multiple independently dated cores within each habitat and urbanization category and quantify the spatial variance of accretion rates before extrapolating them across sites.
lines 120–121: This study did not measure sedimentation rates on tidal flats, so SOC burial rates cannot be calculated, and only stocks are compared. However, in the discussion, the term “burial” is repeatedly used when comparing carbon dynamics between tidal flats and mangroves. It is advised to explicitly distinguish between stocks and burial rates in the title, abstract, and discussion to avoid conceptual confusion.	We sincerely thank the reviewer for this valuable comment. We agree that SOC stocks and SOC burial rates represent different aspects of carbon dynamics and should be clearly distinguished. In this study, ²¹⁰Pb-based sediment accretion rates were only determined for natural and restored mangroves, and therefore SOC burial rates were estimated exclusively for these vegetated ecosystems. Following the reviewer’s suggestion, we have carefully revised the title, Abstract, Results, and Discussion sections to avoid using “burial” when referring to comparisons involving tidal flats. Specifically, descriptions of tidal flats have been revised to focus on SOC stocks and carbon source characteristics, whereas the term “SOC burial rate” is now used only when discussing mangrove ecosystems with available ²¹⁰Pb-derived estimates. These revisions improve conceptual consistency and prevent potential confusion between carbon storage and burial processes.
lines 175–178: The β value is used as a proxy for SOC turnover rate, but the cited literature supporting its validity in mangrove‑tidal flat systems is limited, mainly citing Acton et al., 2013 on a temperate soil study.	Thank you very much for your valuable comment. In response, we have added additional references (Huang et al., 2025b; Xia et al., 2021; Zhang et al., 2022; Zhao et al., 2019) in the revised manuscript to further support the use of the β value as a proxy for soil organic carbon turnover characteristics. These newly included studies span mangrove, coastal wetland, and terrestrial forest ecosystems, thereby providing broader and more direct empirical evidence for the application of the β value across different ecosystem types. This strengthens the theoretical foundation for using the β value to characterize soil organic carbon turnover in coastal blue carbon systems within the present study.
lines 259–271: The authors propose that carbon stocks decline nonlinearly along the urbanization gradient, attributing this to connectivity and vertical accretion exceeding “habitat‑specific thresholds.” However, no specific threshold values or statistical evidence are provided. It is suggested to add a threshold detection method or rephrase “nonlinear” as a more qualitative “accelerated decline” to avoid overstatement.	Thank you very much for your valuable comment. We fully understand your concern regarding the use of the term “nonlinear.” In this study, “nonlinear” is not based on formal nonlinear model fitting or threshold detection methods. Instead, it is used to describe the observed response pattern of soil carbon dynamics across the urbanization gradient. Specifically, our results show that carbon burial rates in both natural and restored mangroves do not change proportionally with increasing urbanization intensity; rather, they peak at moderate urbanization levels and decline significantly under high urbanization. In this sense, we consider the term “nonlinear” appropriate to describe the observed trend, without implying that a statistically derived nonlinear function has been explicitly identified. Regarding the term “habitat-specific thresholds,” this interpretation was mainly derived from previously published studies cited in the manuscript and is used here as a mechanistic explanation rather than a threshold directly quantified in this study. To avoid potential misunderstanding, we will further refine and clarify these expressions in the revised manuscript to ensure accurate interpretation of our findings.
Figure 4 shows vertical profiles of SOC burial rate and sediment age, but age data come from only one site per type, whereas burial rates are averages of multiple sites. This combined presentation may misleadingly suggest that age and burial rate derive from the same spatial scale. It is recommended to clearly state the difference in data sources in the figure caption or to present them separately.	Thank you very much for your valuable comment. In response, we have further clarified in the caption of Figure 4 in the revised manuscript the differences in data sources and their spatial representativeness to avoid potential misunderstanding. We retained both datasets within the same figure to facilitate a more intuitive comparison of the relationship between sediment age and soil organic carbon burial characteristics across ecosystem types and urbanization gradients.

Citation: https://doi.org/10.5194/egusphere-2026-2512-AC4

Minde Huang, Fen Guo, Xueqin Gao, Xiaoguang Ouyang, and Yuan Zhang

Viewed

Total article views: 644 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
438	152	54	644	28	31

HTML: 438
PDF: 152
XML: 54
Total: 644
BibTeX: 28
EndNote: 31

Views and downloads (calculated since 13 May 2026)

Month	HTML	PDF	XML	Total
May 2026	288	52	19	359
Jun 2026	58	14	10	82
Jul 2026	35	30	8	73
Aug 2026	16	15	3	34
Sep 2026	41	41	14	96

Cumulative views and downloads (calculated since 13 May 2026)

Month	HTML	PDF	XML	Total
May 2026	288	52	19	359
Jun 2026	58	14	10	82
Jul 2026	35	30	8	73
Aug 2026	16	15	3	34
Sep 2026	41	41	14	96

Viewed (geographical distribution)

Total article views: 604 (including HTML, PDF, and XML) Thereof 604 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 20 Sep 2026

Short summary

We studied how urbanization affects carbon stored in coastal wetlands by comparing natural and restored mangroves and nearby tidal flats across places with different levels of urban development. We found that stronger urban pressure reduces carbon storage, changes where that carbon comes from, and makes it less stable over time. Natural mangroves in less urbanized areas were the most effective carbon stores, highlighting their importance for climate mitigation and coastal protection.


Total:	0
HTML:	0
PDF:	0
XML:	0