the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 11 Oct 2024
Modeling of greenhouse gas emissions from paludiculture in rewetting peatlands is improved by high frequency water table data
Abstract. Rewetting drained peatlands can reduce CO2 emissions but prevents traditional agriculture. Crop production under rewetted conditions may continue with flood-tolerant crops in paludiculture, but its effects on greenhouse gas (GHG) emissions compared to rewetting without further management are largely unknown This study was conducted between 2021 and 2022 on a fen peatland in central Denmark. At the study site, three harvest/fertilization management treatments were implemented on Reed Canary Grass (RCG) established in 2018. Measurements of CO2 and CH4 emissions were conducted biweekly using a transparent manual chamber connected to a gas analyzer and manipulating light intensities with four shrouding levels. Although this was a rather wet peatland (−8 cm mean annual WTD), the site was a CO2 source with a mean net ecosystem C balance (NECB) of 6.5 t C ha−1 yr−1 across treatments. Model simulation with the use of high temporal resolution water table depth (WTD) data was able to better capture ecosystem respiration (Reco) peaks compared to the use of mean annual WTD, which underestimated Reco. Data on pore water chemistry further improved statistical linear models of CO2 fluxes using soil temperature (Ts), WTD, ratio vegetation indices and PAR as explanatory variables. Significant differences in CO2 emissions and water chemistry parameters were found between studied blocks, with higher Reco corresponding to blocks with higher pore water nutrient concentrations. Methane emissions averaged 113 kg of CH4 ha−1 yr−1, equivalent to 11.3 % of the total carbon emission in CO2 equivalents. Because of large heterogeneity among the experimental blocks no significant treatment effect was found, however, the results indicate that biomass harvest reduces GHG emission from productive rewetted peatland areas in comparison with no management, whereas on less productive areas it is beneficial to leave the biomass unmanaged.
- Preprint
(2220 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2024-3030', Anonymous Referee #1, 21 Jan 2025
Suggestion: Major revision or rejection
The manuscript ‘Modeling of greenhouse gas emissions from paludiculture in rewetting peatlands is improved by high frequency water table data’ by Andres F. Rodriguez et al. investigates CO2 emissions, net ecosystem C balance and CH4 emissions, in a fen peatland in central Denmark across three harvest/fertilization management treatments (Paludiculture with Reed Canary Grass (RCG) implemented 3 years ago). The three treatments in the study involved harvesting biomass five times, two times, and not at all, respectively. The authors conducted bi-weekly sampling over the study period of one year to investigate management effects on CO2 and CH4 emissions from managed peatlands and computed net ecosystem C balances. The authors identify the study site as a net CO2 source. When no significant differences in GHG emissions were observed between treatments, the authors shifted their focus to modeling CO2 emissions. They demonstrated that increasing the resolution of soil temperature measurements to an hourly scale improved the model's accuracy, although this improvement was relatively modest, improving the prediction of ecosystem respiration (Reco) by 5%.
Overall, the study addresses the important and timely topic of GHG emissions in managed peatlands, providing important data on CO2 and CH4 fluxes. However, there are several critical issues that limit the impact and clarity of the manuscript:
Structure and Clarity of Major Findings:
The manuscript would benefit from a clear outline of its key findings. Since no significant treatment effects on GHG emissions were detected, the focus shifts to the modeling improvements. However, the practical significance of incorporating hourly soil temperature measurements could be discussed more thoroughly. For instance, exploring intermediate temporal resolutions, such as bi-weekly data (aligned with the sampling interval), might provide a more balanced view of the model's utility and feasibility. Also, this shift in focus during the manuscript often makes it difficult to follow. The manuscript may be clearer if focused on one topic, the modelling or the treatment aspect. One does not exclude the next, but should rather support the other instead of being seemingly separate and only partially connected stories.Sampling Design and Representativeness:
The study's sampling design raises some concerns regarding its ability to address the research questions. With only four plots and three treatments (control and two fertilized/harvested treatments), there are limited replicates (n=4) in what appears to be a highly heterogeneous environment. This complicates statistical analyses. The PCA results suggest significant variation among plots based on nutrient levels and water table depth, possibly impacting the patterns strongly. A seasonal or event-based analysis might help uncover more nuanced patterns and drivers. Instead of focussing on annual budgets drivers of GHG emissions and NEE could be explored throughout the whole sampling period. Additionally, it would be helpful to clarify how well control and fertilized plots were spatially separated to avoid fertilizer spillover effects.Long-Term Context and Ecosystem Carbon Balance:
The manuscript briefly mentions long-term changes in GHG emissions and net ecosystem carbon balance (NECB), but this aspect is not explored in detail. Comparing emissions and NECB to the baseline conditions at the start of the experiment (3 years ago) could add significant insight into GHG emission development. Pursuing this analysis would provide a deeper understanding of how management practices influence peatland carbon dynamics over time. Especially as there may be a delayed effect on carbon balances (as the authors also point out) of rewetting and implementing paludiculture. Alternatively, the authors could focus on the fact that RVI, water table and soil temperature were more important for CO2 emissions that the treatments, correct?Abstract Accuracy:
The abstract states that "biomass harvest reduces GHG emissions from productive rewetted peatland areas in comparison with no management." However, this conclusion does not appear to align with the results, as no significant treatment effects on GHG emissions were observed. For example, line 447 notes that there was “similar NECB for all treatments.” Revising this statement for consistency with the findings would improve the abstract's accuracy and alignment with the study's outcomes. In general, often marginal significances or non-significant trends are stated and discussed rather than the consequences of non- significant differences between treatments.Statistical Details and Manuscript Quality:
Including sample sizes and test statistics throughout the manuscript would enhance clarity. The text size in all figures would need to be increased and some of the captions clarified. Additionally, improving the precision and clarity of the text would help communicate the findings more effectively. There are many small errors and unclear statements. A careful and thorough revision would be needed to address these issues and improve the overall readability (see specific comments).Specific comments
L19: What were these treatments? A short explanation would help to better follow the results
L22: Please explain abbreviation when used for the first time
L25 – 26: Yes this makes sense, but is probably not a major finding of the study major finding – please revise to say something more to the effect of “ The use of high temporal resolution water table depth facilitated capture ecosystem respiration (Reco) peaks”
L28 -30 : Please highlight the general direction of the differences between the 2 blocks – particularly for CO2 emissions. Was this motioned in the results section?
L32 – 33: This is likely due to a low sample number that limited ability to detect significant differences. Also very different or highly variable annual climate can cause this. High levels of Natural heterogeneity suggest that there is indeed no significant difference to be expected – is this the message ? Else, please suggest reasons why “expected” differences were not found (eg. low sample N, or need for longer measurement time possibly precluded your ability to detect a difference).
L33 -35 : To me, this is your main finding, and a relevant one ! Please consider using this as a major finding as little literature specifically focuses on this aspect. Also please support this better in the results and discussion if possible (see comment above)?
L48-49 : What about drainage effects on N2O (especially since you mention leaching of N in the same sentence)?
L50 – 54: What about N2O emissions ? And more importantly, what about increased CH4 emissions upon increasing the water table ?
L56 – 57: Peat degradation as a result of draining causes this, not the act of draining itself. Please provide a bit more details and specificity, or state there are relationships between degree/length of drainage and higher bulk density etc
L58-59: I would expect a degradation of the humic compounds (due to microbial degradation of the organic matter) upon increasing oxygen due to drainage. Can you cite this?
L65 -67 – consider moving up this statement or moving down the previous (L50-54)
L 73-75: Can the clarity of this sentence be improved?
L76: Missing comma after “peatlands”
L77: Do you mean Paludiculture with ‘it’?
L80-81: Which materials? Can you cite this?
L81 -82: Maybe state why – because of constant crop removal whereby the nutrients are incorporated into the biomass that is removed
L83-85: This statement is true, but not well connected with the previous text. What would you like to point out with this?
L87: Missing comma after “level”.
L89: Isn’t uncertainty by definition “unexplained” ? Please consider removing “unexplained”.
L96” Comma missing before (2)
L98: Missing comma after (WTD)
L113-114: Mean minimum relative to what time period ? (the study period ?), please state in the text. Also, what is the mean based on ? Average of all annual minima 2019 -2022 ? Please be clear.
L114- 116 : Please be clear the time period considered
L117: Missing comma after 2m
Figure 1a : Scale is not legible, no North indication – I would suggest not using Google earth for such images
Figure 1b – Please use white writing – red is hard read on the green background, also increase font size to increase legibility
Figure 1 : Why is the order of sub-figures shifting from left to right and then right to left?
L129 : please be consistent with full month names and abbreviations
Table 1: remove units under C:N ratio
L139-142: Maybe the experimental design could be described in more detail? Were the ‘0-cut, 2-cut, 5-cut’ treatments replicated in all plot. The authors mention other treatments- Were these also present in the investigated plots. In total 4 blocks were investigated, each containing 3 treatments, correct?
L145: I assume 0-cut, 2-cut and 5-cut stands for how often the sites were harvested. A short statement (eg ‘, referring to number of harvests’ indicating this at this point or further up could make it easier for the reader to follow.
L149 – none à non
L152: continuous at what resolution?
L171: sec à seconds or “s”
L172: stored “using”
L185-186 : Was this actually checked / confirmed before successive (increasingly) shaded measurements ? or was it assumed ?
L195 – 198: How were ebullitions (if any) treated ?
L198: Could there have been a potentially significant temperature effect between the first and last (100% dark) measurement that could have affected fluxes ? What does air/soil temperature show ?
L204-205: Is there something missing in this sentence? The ratio between the near-infrared to red light reflectance and what?
L217: delete ‘and it was calculated’
L236: Comma missing before RVI
L263-265: What kind of pump was used and at what flow rate ?
L 278: Were the statistics based on measured or modelled values?
L286-288: This sounds like forward selection was used, yes?
L311-312: The last sentence repeats what the previous one stated
L314: Doesn’t GPP usually increase with PAR an soil temperature ?
Figure 4: here seems to be a relatively strong effect of the blocks? How does this look when the treatments are color coded? Why does GPP decrease with increasing PAR? NEE decreases with increasing Reco, however I would expect that if respiration increases at the same GPP the NEE reduces (considering Reco-GPP=NEE)?
Figure 5: Panel A and B are not described. It would the reader to include abbreviations in the figure caption.
L340: Please mention the test, and n. Was this tested using the modelled GPP or the measured GPP?
Figure 6: Please increase text size in plots
Figure 7: Please increase text size
Line 392-396: This is hard to follow, consider simplifying and starting with: Including hourly Ts improved the model by X percent…..
Line 394: Which table does ‘table A4’ refer to?
Line 398: Which R eco values? Compared to which other values? And why are there only two values for NECB when there are many more treatments?
Figure 8: Please increase text size
Table 5: Here a table set up similar to table 3 would help to understand the patterns better. It seems that there were very clear trends in the blocks, complicated by the fertilization per treatment. Please add the number of samples
Table 6: Does ‘N’ mean ‘not significant ‘ns’? What are the sample sizes?
Line 456: please add standard deviations
Line 458: Should NECB not be given in units C and not CO2-C, since this includes yields and is not based only on emissions? I am not sure what is compared here : ‘our NECB is larger than emissions’ is NECB compared directly with CO2 emissions?
468-469: What is the considered time frame here?
L 473: Please insert ‘and ‘ before ‘drove’
L480: What happened 3 years ago?
L495: Was this reduction significant?
L516: How are these manged? How do the CH4 emissions compare to not manages peatlands?
Line 525: delete ‘some’
L532 which study site? Please indicate if the same one that the current study is focused on is meant?
L 549: Use percent to make this result more generalizable. Is this significant? How could this impact budgets that were made with simpler models?
L550: Model 4 also included Ts as explanatory variable
588-590: Please also indicate how big this improvement was
622-625: Please rephrase these two sentences to make them clearer. In the same sentence it is stated that paludiculture provides an opportunity for emission reduction but also that harvested biomass makes out a considerable amount of GHG emissions from cultivated RCG in fen peatlands. Please clarify.
L628: If I understand correctly biogas production would then still lead to an overall GHG emission from cultivated RCG in fen peatlands?
L637-639: were these measurements compared and tested?
L641-644: Is this based on a non-significant trend or was this significant. If this is only a trend it should probably not be highlighted in the conclusion.
Citation: https://doi.org/10.5194/egusphere-2024-3030-RC1 -
AC1: 'Reply on RC1', Andres Rodriguez Grisales, 05 Mar 2025
We would like to thank referee # 1 for taking the time to revise our manuscript. We appreciate the feedback, input, and thoughtful comments. We worked to address each comment and implement the suggestions. We are currently working on a revised version of the manuscript. The text below includes the reviewer’s comments in italics followed by our answer in plain text.
Structure and Clarity of Major Findings:
The manuscript would benefit from a clear outline of its key findings. Since no significant treatment effects on GHG emissions were detected, the focus shifts to the modeling improvements. However, the practical significance of incorporating hourly soil temperature measurements could be discussed more thoroughly. For instance, exploring intermediate temporal resolutions, such as bi-weekly data (aligned with the sampling interval), might provide a more balanced view of the model's utility and feasibility. Also, this shift in focus during the manuscript often makes it difficult to follow. The manuscript may be clearer if focused on one topic, the modelling or the treatment aspect. One does not exclude the next, but should rather support the other instead of being seemingly separate and only partially connected stories.Answer: The results section of the manuscript will be edited to follow a more clear structure presenting the findings which aligns with the objectives and the discussion. In the structure of the revised manuscript, results from the C budget connect to the heterogeneity of the rewetted peatland under paludiculture which is addressed by using the data on water table dynamics and pore water chemistry though the modeling approaches. This will help to bring up the finding on the potential of paludiculture applied in nutrient rich areas, which both reviewers agree should be raised as a major finding in the manuscript. We have explored two additional temporal resolutions (seasonal WTD and monthly WTD), these results can now be seen in table A4 (in the appendix). These results will also be mentioned and explored in the revised manuscript. We would like to point out that including WTD data at increasingly higher temporal resolution improves the accuracy of the models, however, the most relevant comparison of models performed in this study is between the hourly data and the mean annual data because current emission factors for rewetted peatlands are established based on mean annual WTD, which according to our findings can underestimate Reco in these peatlands.
Sampling Design and Representativeness:
The study's sampling design raises some concerns regarding its ability to address the research questions. With only four plots and three treatments (control and two fertilized/harvested treatments), there are limited replicates (n=4) in what appears to be a highly heterogeneous environment. This complicates statistical analyses. The PCA results suggest significant variation among plots based on nutrient levels and water table depth, possibly impacting the patterns strongly. A seasonal or event-based analysis might help uncover more nuanced patterns and drivers. Instead of focussing on annual budgets drivers of GHG emissions and NEE could be explored throughout the whole sampling period. Additionally, it would be helpful to clarify how well control and fertilized plots were spatially separated to avoid fertilizer spillover effects.Answer: We did not expect heterogeneity among the replicate blocks because the fields seemed uniform and without differences in topography. We consider that for the study area, four 288 m2 blocks were a sufficient number of replicates to conduct the study, however, we agree with the reviewer that more replicate blocks would have strengthened the statistical analyses, we unfortunately did not have enough personnel and equipment to include more replicate blocks in the study. The PCA, correlations, and the linear mixed model analyses were conducted as event-based analyzes because in them we included the Reco and GPP obtained at each measurement campaign and we relate them to their drivers (WTD, RVI, PAR, and soil temperature), as well as the pore water nutrients obtained at each measurement campaign. In these analyses we did not use annual averages. Following the recommendation of the reviewer we have included the number of samples used in these analyses, which will clarify this in the manuscript. We conducted statistical comparisons for the treatments and blocks based on annual budgets because emissions are regularly evaluated and quantified in terms of annual budgets, and this makes the results presented in this manuscript more relatable to other studies. Spillovers were avoided by performing measurements within 55 x 55 cm collars located in the middle of the 3 m wide treatment plots.
Long-Term Context and Ecosystem Carbon Balance:
The manuscript briefly mentions long-term changes in GHG emissions and net ecosystem carbon balance (NECB), but this aspect is not explored in detail. Comparing emissions and NECB to the baseline conditions at the start of the experiment (3 years ago) could add significant insight into GHG emission development. Pursuing this analysis would provide a deeper understanding of how management practices influence peatland carbon dynamics over time. Especially as there may be a delayed effect on carbon balances (as the authors also point out) of rewetting and implementing paludiculture. Alternatively, the authors could focus on the fact that RVI, water table and soil temperature were more important for CO2 emissions that the treatments, correct?Answer: A comparison to baseline conditions would be useful, however, those baseline measurements were not taken, and therefore we are not able to perform that comparison. We refer to trends and differences in NECB after three years of reed canary grass establishment at the study site as rewetting progresses. Even though we could not compare our results to baseline conditions, we compared them to results from 2020 to 2021 presented in Nielsen et al (2024), and in doing so we analyzed trends in C emissions as rewetting progressed at the study site. We explored the importance of WTD, RVI, soil temperature, and PAR in Reco and GPP in figure 5 (figure 7 in revised manuscript), in which we see the variability of fitted parameters corresponding to WTD, RVI, soil temperature, and PAR as they fit to the Reco and GPP models of each of the plots.
Abstract Accuracy:
The abstract states that "biomass harvest reduces GHG emissions from productive rewetted peatland areas in comparison with no management." However, this conclusion does not appear to align with the results, as no significant treatment effects on GHG emissions were observed. For example, line 447 notes that there was “similar NECB for all treatments.” Revising this statement for consistency with the findings would improve the abstract's accuracy and alignment with the study's outcomes. In general, often marginal significances or non-significant trends are stated and discussed rather than the consequences of non- significant differences between treatments.Answer: Abstract was revised and edited for accuracy. Even though no significant differences between management strategies were found on NECB, we found marginally significant differences in GPP, as well as statistical differences in CO2 emissions and pore water nutrients between blocks. Based on these differences, we state that “biomass harvest might potentially reduce GHG emissions from productive rewetted peatland areas in comparison with no management, whereas on less productive areas it might be beneficial to leave the biomass unmanaged” on the revised abstract.
Statistical Details and Manuscript Quality:
Including sample sizes and test statistics throughout the manuscript would enhance clarity. The text size in all figures would need to be increased and some of the captions clarified. Additionally, improving the precision and clarity of the text would help communicate the findings more effectively. There are many small errors and unclear statements. A careful and thorough revision would be needed to address these issues and improve the overall readability (see specific comments).Answer: Manuscript will be revised and edited accordingly. Specific comments regarding statistical information to be added and figures were addressed and incorporated in the revised version we are currently working on.
Specific comments
L19: What were these treatments? A short explanation would help to better follow the results
Answer: a short explanation was included.
L22: Please explain abbreviation when used for the first time
Answer: change implemented.
L25 – 26: Yes this makes sense, but is probably not a major finding of the study major finding – please revise to say something more to the effect of “ The use of high temporal resolution water table depth facilitated capture ecosystem respiration (Reco) peaks”
Answer: change was implemented.
L28 -30 : Please highlight the general direction of the differences between the 2 blocks – particularly for CO2 emissions. Was this motioned in the results section?
Answer: further explanation was included in the abstract. Differences between blocks in CO2 emissions are mentioned in the results section 3.2 (section 3.1 of the revised manuscript) in the second paragraph, while differences between blocks in pore water nutrients are mentioned in section 3.4 in the second paragraph.
L32 – 33: This is likely due to a low sample number that limited ability to detect significant differences. Also very different or highly variable annual climate can cause this. High levels of Natural heterogeneity suggest that there is indeed no significant difference to be expected – is this the message ? Else, please suggest reasons why “expected” differences were not found (eg. low sample N, or need for longer measurement time possibly precluded your ability to detect a difference).
Answer: We agree with the reviewer in the sense that more block replicates would most likely have resulted in a more significant treatment effect. However, as stated in the answer to the general comments above, we decided to use four block replicates based on the observation of the uniformity and similar topography of the block replicates. A previous study by Nielsen et al. (2024), which is also referenced in our manuscript, also found no significant differences in NECB between the same three management strategies at the study site for the 2020 to 2021 period, which possibly suggest that a longer study period would still indicate no differences between management strategies.
L33 -35 : To me, this is your main finding, and a relevant one ! Please consider using this as a major finding as little literature specifically focuses on this aspect. Also please support this better in the results and discussion if possible (see comment above)?
Answer: We consider this to be one of the major findings of this manuscript. In the revised manuscript it will be highlighted as the major finding. Further edits in the results and discussion on revised manuscript will highlight this as the major finding.
L48-49 : What about drainage effects on N2O (especially since you mention leaching of N in the same sentence)?
Answer: Peat drainage can also increase N2O emissions by releasing N bound to C in peat soils. In these sentences in the introduction though, we try to keep the focus on what we measure in our study such as CO2, DOC, DON, and thereby be concise in the introduction section. However, in response to a comment from reviewer 2, we have included a discussion on the implications of N2O emissions, which were not measured in this study, in section 4.5.
L50 – 54: What about N2O emissions ? And more importantly, what about increased CH4 emissions upon increasing the water table ?
Answer: In this sentence we want to relate peatland drainage to emissions. We then focus on the relation between drainage, WTD, and CO2. Since we focus on CO2 emissions in this study, we would like to stay within the topic. We agree that N2O emissions can be significant in rewetted peatlands, but this is not the focus of this study. Since we also report CH4 emissions, we included an explanation of the trade-off between CO2, CH4, and WTD in the introduction.
L56 – 57: Peat degradation as a result of draining causes this, not the act of draining itself. Please provide a bit more details and specificity, or state there are relationships between degree/length of drainage and higher bulk density etc
Answer: We agree, sentences were edited to state that peat degradation is the cause of these changes and not drainage itself.
L58-59: I would expect a degradation of the humic compounds (due to microbial degradation of the organic matter) upon increasing oxygen due to drainage. Can you cite this?
Answer: The reviewer is correct. We have corrected these sentences regarding a decrease in humic substances and polyphenols upon peat degradation. Sentences were then linked to the next sentence to add references.
L65 -67 – consider moving up this statement or moving down the previous (L50-54)
Answer: Sentences were reorganized as suggested: Sentences on the trade-off between CO2 and CH4 were moved to the previous paragraph, and additionally, sentences on the interest on peatland restoration were moved to the next paragraph.
L 73-75: Can the clarity of this sentence be improved?
Answer: Sentence was rephased for clarity.
L76: Missing comma after “peatlands”
Answer: comma included.
L77: Do you mean Paludiculture with ‘it’?
Answer: That is correct. We tried not to repeat paludiculture again because we already used the word in the previous sentence.
L80-81: Which materials? Can you cite this?
Answer: Insulation materials. A reference was included.
L81 -82: Maybe state why – because of constant crop removal whereby the nutrients are incorporated into the biomass that is removed
Answer: A sentence was added to the text to explain this effect.
L83-85: This statement is true, but not well connected with the previous text. What would you like to point out with this?
Answer: This sentence is part of the justification, therefore it was relocated to the beginning of the next paragraph, which includes the justification of the study.
L87: Missing comma after “level”.
Answer: comma added.
L89: Isn’t uncertainty by definition “unexplained” ? Please consider removing “unexplained”.
Answer: The reviewer is correct. “Unexplained” was removed.
L96” Comma missing before (2)
Answer: comma added.
L98: Missing comma after (WTD)
Answer: comma added.
L113-114: Mean minimum relative to what time period ? (the study period ?), please state in the text. Also, what is the mean based on ? Average of all annual minima 2019 -2022 ? Please be clear.
Answer: Information was included to make clear that the means were for the study period and that the minimum and maximum correspond to means among experimental blocks.
L114- 116: Please be clear the time period considered
Answer: Information was included.
L117: Missing comma after 2m
Answer: comma was included.
Figure 1a : Scale is not legible, no North indication – I would suggest not using Google earth for such images
Answer: A new map was included replacing the Google Earth image.
Figure 1b – Please use white writing – red is hard read on the green background, also increase font size to increase legibility
Answer: White writing is now used. Font size has been increased.
Figure 1: Why is the order of sub-figures shifting from left to right and then right to left?
Answer: Figure was rearranged.
L129 : please be consistent with full month names and abbreviations
Answer: Full month names were included.
Table 1: remove units under C:N ratio
Answer: Units were removed.
L139-142: Maybe the experimental design could be described in more detail? Were the ‘0-cut, 2-cut, 5-cut’ treatments replicated in all plot. The authors mention other treatments- Were these also present in the investigated plots. In total 4 blocks were investigated, each containing 3 treatments, correct?
Answer: Correct, there were four blocks each containing three treatments. Text was edited and sentences were added to make clarify the experimental design used.
L145: I assume 0-cut, 2-cut and 5-cut stands for how often the sites were harvested. A short statement (eg ‘, referring to number of harvests’ indicating this at this point or further up could make it easier for the reader to follow.
Answer: statement “referring to the number of harvest events applied” was included in the first mention of the three management strategies in the description of the experimental design.
L149 – none à non
Answer: changed to “non”
L152: continuous at what resolution?
Answer: soil temperature and water table depth were measured continuously at hourly intervals. This was added in the manuscript for clarification.
L171: sec à seconds or “s”
Answer: changed to s
L172: stored “using”
Answer: “using” was included.
L185-186 : Was this actually checked / confirmed before successive (increasingly) shaded measurements ? or was it assumed ?
Answer: The re-establishment of atmospheric concentrations was confirmed by measuring the concentrations in the chamber while the chamber was aerated before each successive measurement because the analyzer remained connected to the chamber between measurements.
L195 – 198: How were ebullitions (if any) treated ?
Answer: The “fluxx” function of the Flux package used to calculate these fluxes finds linear conditions within the data, thereby excluding rapid fluctuations in concentrations. This way ebullitions were excluded for CH4 fluxes calculations. This explanation was included for clarification.
L198: Could there have been a potentially significant temperature effect between the first and last (100% dark) measurement that could have affected fluxes ? What does air/soil temperature show ?
Answer: A potential temperature increases during deployment were controlled with a temperature sensor, a cooling ice bag, and a fan that were part of the chamber design. When the temperature inside increased more than 1°C compared to outside, the fan blew cold air into the chamber until temperatures were equalized. This system is described in Detail in Elsgaard et al. (2012), which is referenced in the chamber’s description at the beginning of section 2.3.
L204-205: Is there something missing in this sentence? The ratio between the near-infrared to red light reflectance and what?
Answer: The RVI was calculated as the ratio between the near infrared reflectance and the red light reflectance. The sentence was edited for clarification.
L217: delete ‘and it was calculated’
Answer: change accepted. Sentence was edited for clarification.
L236: Comma missing before RVI
Answer: comma included.
L263-265: What kind of pump was used and at what flow rate ?
Answer: No pump was used, instead, a syringe attached to the tube with the aquarium air stone was used to extract the water samples. This text was included for clarification.
L 278: Were the statistics based on measured or modelled values?
Answer: Statistical comparisons to test the effect of harvest treatment and block on Reco, GPP, NEE, and NECB were performed using modelled values taken to annual budgets, whereas statistical comparisons on correlations, PCA, and linear mixed models were performed using the measured Reco and NEE.
L286-288: This sounds like forward selection was used, yes?
Answer: Each water chemistry parameter was added independently to the base Reco and GPP models. For example, TOC was added to model 5 (base Reco model) and its performance was compared to the base model’s performance, afterwards, DOC was added to model 5 (without TOC) and its performance was compared to the base model, and so forth with all tested water chemistry parameters.
L311-312: The last sentence repeats what the previous one stated
Answer: The second sentence was deleted.
L314: Doesn’t GPP usually increase with PAR an soil temperature ?
Answer: Correct. GPP increases when PAR and soil temperature increase due to more vigorous growth of the vegetation under higher temperatures and warmer conditions. In the manuscript, we state that GPP was negatively correlated to Ts and PAR because, as stated at the beginning of the 2.5 section, we consider (from an atmospheric perspective) Reco as a positive CO2 flux and GPP as a negative CO2 flux. Additional explanation on increases of GPP as larger CO2 uptake were included following recommendation of reviewer 2.
Figure 4: here seems to be a relatively strong effect of the blocks? How does this look when the treatments are color coded? Why does GPP decrease with increasing PAR? NEE decreases with increasing Reco, however I would expect that if respiration increases at the same GPP the NEE reduces (considering Reco-GPP=NEE)?
Answer: Yes. This figure highlights the differences between the blocks and the heterogeneity of the peatland. We did previously a plot color-coding with the treatments but they did not show a clear effect and they were superimposing each other so we decided not to add that figure. The GPP decrease with increasing PAR because we used a negative sign for GPP (see previous comment). The negative correlation seen between Reco and NEE is due to GPP being more dominant than Reco in determining the magnitude of the NEE. NEE can be positive or negative depending on the balance between GPP and Reco. For the PCA and correlation analyses we used the measured Reco and NEE (pore water sampling was done at the same time as flux measurement campaigns). For the NEE, we used the measurement under fully transparent conditions (0% PAR blocked), which was of a larger magnitude compared to the Reco generally, and this resulted in the NEE being negatively correlated to Reco and positively correlated to GPP.
Figure 5: Panel A and B are not described. It would the reader to include abbreviations in the figure caption.
Answer: Figure and caption were edited.
L340: Please mention the test, and n. Was this tested using the modelled GPP or the measured GPP?
Answer: The test and n was included. This was tested using the annual budgets, therefore we used the modelled data taken to annual budgets to perform these comparisons.
Figure 6: Please increase text size in plots
Answer: text size was increased and figure was edited as recommended by reviewer 2.
Figure 7: Please increase text size
Answer: text size was increased and figure was edited as recommended by reviewer 2.
Line 392-396: This is hard to follow, consider simplifying and starting with: Including hourly Ts improved the model by X percent…..
Answer: Paragraph was edited by simplifying it and improving its readability.
Line 394: Which table does ‘table A4’ refer to?
Answer: table A4 is located in the appendix.
Line 398: Which R eco values? Compared to which other values? And why are there only two values for NECB when there are many more treatments?
Answer: The sentence was edited for clarification. The Reco values refer to the Reco calculated using annual means of WTD and soil temperature as described in the same paragraph in the manuscript. There are two values because these are mean NECB values among all plots.
Figure 8: Please increase text size
Answer: Text was increased and edited as recommended by reviewer 2.
Table 5: Here a table set up similar to table 3 would help to understand the patterns better. It seems that there were very clear trends in the blocks, complicated by the fertilization per treatment. Please add the number of samples
Answer: In this table the blocks are averaged by treatments and the treatments are averaged by block, therefore, a setup similar to table 3 could not be implemented. With a setup similar to table 3 we would not be able to show the statistical comparisons performed between blocks and harvest treatments which is important for the discussion on peatland heterogeneity. Therefore, we consider that Table 5 should remain in its current structure. N values were included in the caption.
Table 6: Does ‘N’ mean ‘not significant ‘ns’? What are the sample sizes?
Answer: correct N meant not significant. Table has been edited, now “Ns” as “not significant” has been included. Sample sizes have been included.
Line 456: please add standard deviations
Answer: standard deviations were included.
Line 458: Should NECB not be given in units C and not CO2-C, since this includes yields and is not based only on emissions? I am not sure what is compared here : ‘our NECB is larger than emissions’ is NECB compared directly with CO2 emissions?
Answer: In this paragraph we compare our NECB to studies that reported NECB for Danish, German, and UK drained and rewetted peatlands. The referenced studies reported NECB based on CO2 emissions in units of t C ha-1 yr-1.Units were changed to t C ha-1 yr-1in the whole paragraph for clarification. To make clear that we report NECB of CO2 and following recommendation of reviewer 2, this was included when NECB is mentioned in abstract, objectives, methods, and in caption of tables.
468-469: What is the considered time frame here?
Answer: In these sentences, time frame is the period between 2020 and 2022 in which there were changes in management and a progressive increase in the WTD at the study site.
L 473: Please insert ‘and ‘ before ‘drove’
Answer: change was not implemented because it would have change the meaning of the sentence. Our sentence: “This naturally stimulated CO2 production in the peat and together with plant respiration drove the high annual Reco” is intended to state that lower WTD along with higher soil temp. naturally stimulated CO2 production and this, together with plant respiration drove higher Reco.
L480: What happened 3 years ago?
Answer: Three years before the study was conducted, reed canary grass was established at the study site. This is explained in section 2.2. Sentence was modified for clarification.
L495: Was this reduction significant?
Answer: Differences in NEE and NECB between harvest treatments were not significant due to variability among blocks. However, in this case, we are referring to the difference that could be seen in one particular block which represents a high porewater nutrients and high CO2 emissions area (block D), since we refer to one block, we call this a “potential reduction”.
L516: How are these manged? How do the CH4 emissions compare to not manages peatlands?
Answer: The referenced studies compiled CH4 emissions from both pristine and rewetted peatlands in Denmark and Germany. Text was included for clarification in revised manuscript. Pristine peatlands have generally lower emissions than rewetted sites, however, CH4 emissions vary considerably as explained in the introduction. Since the focus of this manuscript is CO2 emissions, we included the information in this paragraph to compare our CH4 emissions to values reported in the literature but did not elaborate in the processes controlling them in order to not divert from the manuscript’s focus.
Line 525: delete ‘some’
Answer: “some” was deleted.
L532 which study site? Please indicate if the same one that the current study is focused on is meant?
Answer: Correct, this is the same study site. The sentence was edited.
L 549: Use percent to make this result more generalizable. Is this significant? How could this impact budgets that were made with simpler models?
Answer: The text was edited and the percent values were included. In this sentence, we compared the total mean NECB that would be obtained with models 2 and 3 vs. the one obtained with model 4, therefore, the values were compared descriptively, and no statistical significance was tested. The budget results we report in this study were all obtained with our best model (model 4). Budgets that were calculated with models 2 and 3 were only to be compared to model 4 results.
L550: Model 4 also included Ts as explanatory variable
Answer: correct. Text was edited for clarification.
588-590: Please also indicate how big this improvement was
Answer: the text was edited and magnitude of improvement was included.
622-625: Please rephrase these two sentences to make them clearer. In the same sentence it is stated that paludiculture provides an opportunity for emission reduction but also that harvested biomass makes out a considerable amount of GHG emissions from cultivated RCG in fen peatlands. Please clarify.
Answer: Sentences were rephrased and an explanatory sentence was included.
L628: If I understand correctly biogas production would then still lead to an overall GHG emission from cultivated RCG in fen peatlands?
Answer: That is correct because the biogas would still be carbon being lost to the atmosphere, however if biogas from paludiculture is used it would replace the use of fossil fuels from other sources, therefore we suggest this as a possible use of the harvested biomass.
L637-639: were these measurements compared and tested?
Answer: The NECB from 2020 to 2021 at the study site was previously reported by Nielsen et al (2024). As part of the discussion, in section 4.1, this previously reported NECB is compared with our reported NECB. The relation between the NECB and the ongoing rewetting process is also discussed. The text was edited in the conclusion for clarification.
L641-644: Is this based on a non-significant trend or was this significant. If this is only a trend it should probably not be highlighted in the conclusion.
Answer: These statements are based on significant differences found in emissions and pore water nutrients between blocks which are discussed in sections 4.2 and 4.4. Sentence was edited to indicate that a potential reduction in C fluxes could be achieved in nutrient rich areas.
Citation: https://doi.org/10.5194/egusphere-2024-3030-AC1
-
AC1: 'Reply on RC1', Andres Rodriguez Grisales, 05 Mar 2025
-
RC2: 'Comment on egusphere-2024-3030', Anonymous Referee #2, 07 Feb 2025
General comments:
The aim of the study is to investigate the net ecosystem carbon balance (NECB) of reed canary grass production at shallow water level depths and its relationship with pore water chemistry, and to compare modelling approaches for gapfilling for obtaining annual balances. In a riparian temperate fen four blocks with three harvest/fertilization treatments were established. The authors present data on one year of manual closed chamber measurements of carbon dioxide and methane fluxes and accompanying measurements of pore water parameters.
Information on greenhouse gas emissions of wet utilization of peatlands (paludiculture) is still scarce while the interest in this land use remains high, so the presented data is a valuable contribution to the field. A further strong point of the presented study is the thorough methodological approach leading to high-quality data. The manuscript addresses topics relevant to the scope of SOIL and that are multidisciplinary. It is written clearly and concisely.
However, the manuscript could use some improvement regarding the presentation of the results: The present narrative closely follows the comparison of different modelling approaches and, in my view, undersells the strong points of the data itself. While it is good to consider methodological/modelling questions, I think the results are not particularly groundbreaking (isn’t using more data always better?). It therefore remains unclear to me, why the authors regard this information as important enough to justify it being the main point of the manuscript’s title. However, if the methodological questions are going to be kept as the main selling point of the manuscript, the presentation of the results should be focused more on supporting their claims, e.g. by showing figures of measured vs. modelled values or comparisons of calibration/validation (at least in the appendix). So far, it seems like the models have only been calibrated, not yet validated.
For these reasons, I recommend to re-consider the manuscript for publication following Major Revisions.
Specific Comments:
Figures: In my view all of them need to be reworked somewhat to make their message clearer and their presentation more efficient. See Technical Comments for details.
L25/26: What is the novelty of the finding, that more data is always more accurate? Also, how do you know the additional Reco peaks you modelled actually happened?
L67-71: To make the overall pattern of the CO2/CH4-tradeoff as observed in the literature clearer to readers, a statement highlighting the general benefits of peatland rewetting for the climate (even when considering CH4) should be added here.
L163: “depending on meteorological conditions” à What does this mean, exactly?
L196: “if a possible leakage was identified” à How was this done?
L197-198: Why did you only consider CH4 from opaque measurements?
L264-265: “placed 20 cm below the water table in each piezometer” à so the absolute depth of the measurements changes over the study period? How did this affect your results?
L307-308: “In general, the 0-cut plots provided the best model performances” à Any explanation?
L356-357: There seems to be a hydrological gradient in your study area that could explain the observed patterns. I think the influence of water level on the different results for the blocks needs to be discussed more.
L368: Supports the explanation of water levels having an influence, see previous comment.
Line 376: NECB technically also needs to include carbon from CH4 emissions (although it won’t make much of a difference).
Figure 8: Adding measured values would be helpful, in order to assess which model is more accurate.
L454-457: Please add information whether the values for your comparison also came from fertilized peatlands.
L539-541: Why is this a less important finding than the sensitivity in regard to frequency of water level measurements?
L584-585: Could it not also be a result of fertilization?
L593-594: Could also be a result of less peat mineralization?
L606: This might be a good place to discuss the implications of water level differences, as mentioned above.
L634 The implications for your conclusions of not knowing about nitrous oxide emissions should be discussed somewhere.
Technical Comments:
L38 “global C cycle” à explain abbreviation?
L44 add comma after “conditions” to make sentence more readable
L47 “sp.” After genus name should not be italicized
L84 add “with or without paludiculture” or similar after “peatlands”
Figure 1: Generally: Panels should be arranged in rows as abcd, white space should be removed. (A) Why is why is this map oval? remove unreadable text (google earth etc.), use 2-dimensional marker. (B) red on green is hard to read, explain letters/blocks, text needs to be larger.
Figure 2: This figure needs to be reworked to improve the space/infomation-ratio and quick readability. Both panels should be combined.
Figure 3: I think this would work better as a table.
L261: “electroconductivity” with lower case “e”
L314: “GPP was negatively correlated to Ts…” This kind of wording is always very confusing when talking about negative values. Add “(larger CO2 uptake)” or similar.
Figure 4: What is your main point for this figure? If it is the direction of arrows for fluxes and environmental parameters, the colours should be chosen so that these arrows stand out more, and the labels should be larger. Right now, the feature that stands out most is the oval marking the measurements of block A. I assume, this is not what you want to lead the readers to see first and what is the most interesting about this figure?
Figure 5: x-axis labels are far too small.
L358: “higher”à see comment above, add “(more positive) or similar to make the direction clearer
Figure 6: Text in the figure is very hard to read and should be larger. The colours (blue and green) are too similar and hard to distinguish
Figure 7: Text too small, lines too thin, red/green contrast should be avoided to make figure readable for people with colour-blindness.
Figure 8: See Figure 7.
L594: “.” Or “;” instead of “,”?
Figure A1: In its present form I don't know how useful this figure is, at least I cannot read almost any of the variable names?
Citation: https://doi.org/10.5194/egusphere-2024-3030-RC2 -
AC2: 'Reply on RC2', Andres Rodriguez Grisales, 05 Mar 2025
We thank referee # 2 for revising our manuscript, for the thoughtful comments and the valuable feedback on how to improve the manuscript. We are currently working on a revised version which includes the feedback provided. The text below addresses each of the comments and questions raised. The text in italics are the reviewer’s original comments followed by our answer in plain text.
General comment:
The manuscript could use some improvement regarding the presentation of the results: The present narrative closely follows the comparison of different modelling approaches and, in my view, undersells the strong points of the data itself. While it is good to consider methodological/modelling questions, I think the results are not particularly groundbreaking (isn’t using more data always better?). It therefore remains unclear to me, why the authors regard this information as important enough to justify it being the main point of the manuscript’s title. However, if the methodological questions are going to be kept as the main selling point of the manuscript, the presentation of the results should be focused more on supporting their claims, e.g. by showing figures of measured vs. modelled values or comparisons of calibration/validation (at least in the appendix). So far, it seems like the models have only been calibrated, not yet validated
Answer: The results section and the title will be edited to reflect a more clear structure which also aligns with the objectives and the discussion. In the structure of the revised manuscript, results from the C budget open the door to the heterogeneity findings, then results from the modeling approaches are presented followed by the pore water chemistry results. This, along with further discussion on heterogeneity and the hydrological gradient (as recommended in reviewer 2 specific comments) will help to bring up the finding on the potential of paludiculture applied in nutrient rich areas, which both reviewers agree should be raised as a major finding in the manuscript. Regarding the modeling work, additional temporal resolutions were included in the appendix table A4 (as recommended by reviewer 1) and this will be briefly mentioned in the discussion; furthermore, plots of modelled vs. measured Reco will be included in the appendix to support model performance findings. In the structure of the revised manuscript, the modeling work is used to emphasize the importance of higher temporal resolution data in heterogeneous rewetting peatlands. The novelty of this modeling work, as it is also explained in the specific comments below, resides in the fact that current emission factors of rewetted peatlands are established based on mean annual water table data, which, our findings indicate, can lead to an underestimation of Reco .
Specific Comments:
Figures: In my view all of them need to be reworked somewhat to make their message clearer and their presentation more efficient. See Technical Comments for details.
Answer: All figures were edited following recommendations of both reviewers.
L25/26: What is the novelty of the finding, that more data is always more accurate? Also, how do you know the additional Reco peaks you modelled actually happened?
Answer: The text was edited. The words “better” and “peaks” were replaced. This was done following recommendations from reviewer 1. The edited text indicates that variability in Reco is facilitated by high temporal resolution WTD data. The modelled Reco peaks were the result of implementing high temporal resolution data and constitute the most detailed results obtained in this manuscript. The reviewer is correct, since these peaks are modelled, we cannot be certain that they occurred. However, results from the model evaluation seen in table 2 (table 4 in revised manuscript) show model performances for model 4 with R2 values between 0.74 and 0.97, additionally, plots of modelled vs. measured Reco will be included in the appendix to support the high model performances obtained. Our approach was to compare the results obtained using hourly input data to results obtained using larger temporal resolution input data. The novelty of the finding is that, using mean annual WTD data to establish emission factors, as it is currently done, can lead to an underestimation of CO2 emissions in rewetting peatlands.
L67-71: To make the overall pattern of the CO2/CH4-tradeoff as observed in the literature clearer to readers, a statement highlighting the general benefits of peatland rewetting for the climate (even when considering CH4) should be added here.
Answer: A statement regarding the long term reestablishment of the C sink function of rewetted peatlands was included supported by references.
L163: “depending on meteorological conditions” à What does this mean, exactly?
Answer: Measurements were conducted on days with predominantly clear sky conditions. The text was edited for clarification. This was done to obtain the gradient in light intensities by using the shrouds in the chamber.
L196: “if a possible leakage was identified” à How was this done?
Answer: Leakages were identified when negative or non-linear Reco fluxes were measured. We expected clear linear behavior in CO2 fluxes under dark conditions, when this was not observed it was an indication of possible leakage. Text was included for clarification.
L197-198: Why did you only consider CH4 from opaque measurements?
Answer: We did this because most studies with manual chambers measure CH4 fluxes under opaque conditions and we wanted to obtain CH4 fluxes under comparable conditions. We did a preliminary comparison (not included in the manuscript) between the CH4 fluxes of the transparent and opaque measurements and did not see clear differences so we decide to only use the opaque ones.
L264-265: “placed 20 cm below the water table in each piezometer” à so the absolute depth of the measurements changes over the study period? How did this affect your results?
Answer: Correct. The depth is the soil profile at which samples were collected varied depending on changes in the WTD. However, since all plots and blocks experienced similar variations in WTD throughout the study (Fig. 2), comparisons between them were not affected. Pore water samples were collected 20 cm below the water table because we considered that at that depth (closer to the groundwater/soil interface), they reflected better the biogeochemical conditions in which GHG emissions were taking place.
L307-308: “In general, the 0-cut plots provided the best model performances” à Any explanation?
Answer: That is correct. Harvesting events were more difficult to model even though pre- and post-harvesting RVI measurements were included in models. In terms of the models, harvesting is an abrupt change in the biomass with minor changes in soil temperature and WTD, therefore it is expected that modeling of harvest events would be challenging.
L356-357: There seems to be a hydrological gradient in your study area that could explain the observed patterns. I think the influence of water level on the different results for the blocks needs to be discussed more.
Answer: We agree, further discussion on the influence of the hydrological gradient will be included in the revised manuscript.
L368: Supports the explanation of water levels having an influence, see previous comment.
Answer: Further discussion will be included in the revised manuscript.
Line 376: NECB technically also needs to include carbon from CH4 emissions (although it won’t make much of a difference).
Answer: Correct. To make clear that we are reporting NECB of CO2, this was included in the abstract, objectives, methods, and in the caption of tables. Additionally, in section 4.1 we calculated CH4 contribution to the NECB in terms of CO2e as 11.3% but this is not further discussed to keep the focus on the CO2.
Figure 8: Adding measured values would be helpful, in order to assess which model is more accurate.
Answer: Prior to this figure, a model evaluation was performed (table 2) in which, based on measured values, the best performing Reco model was selected. The best performing model (model 4) was then used to test the effects of the different temporal resolutions of the input data in the sensitivity analysis, and these results are shown in Figure 8. We consider that figure 8 should be focused on the sensitivity analysis. Including the measured Reco values could alter the flow of the manuscript and also saturate the figure therefore we consider that measured values should not be included in Figure 8. Even though measured values will not be included in this figure, additional plots of measured vs. modelled Reco will be included in the appendix to support model performance findings.
L454-457: Please add information whether the values for your comparison also came from fertilized peatlands.
Answer: The dataset used by Koch et al (2023), which is used to compare our measurements, included fertilized peatlands. This information was included in the edited manuscript for clarification.
L539-541: Why is this a less important finding than the sensitivity in regard to frequency of water level measurements?
Answer: We consider the findings on heterogeneity for accurate estimation of emissions in rewetted peatlands to be an important finding of this manuscript. It is also related to findings on the importance of high frequency WTD data because they both aim at a more complete understanding of peatland heterogeneity and variability. Manuscript will be edited to highlight this findings on heterogeneity’s importance in emission estimations from rewetted peatlands as the major finding.
L584-585: Could it not also be a result of fertilization?
Answer: Correct. Fertilization can also contribute to these nutrient concentrations. Sentence was edited to include this.
L593-594: Could also be a result of less peat mineralization?
Answer: Yes, higher WTD would reduce peat mineralization, however, for the particular case of the study site, yearly WTD fluctuations were conducive to peat mineralization taking place when WTD decreased. Further discussion on the relation between WTD fluctuations and nutrients will be included in the revised manuscript.
L606: This might be a good place to discuss the implications of water level differences, as mentioned above.
Answer: Further discussion will be included in revised manuscript.
L634 The implications for your conclusions of not knowing about nitrous oxide emissions should be discussed somewhere.
Answer: A discussion regarding N2O emissions has been included in section 4.5.
Technical Comments:
L38 “global C cycle” à explain abbreviation?
Answer: abbreviation was explained.
L44 add comma after “conditions” to make sentence more readable
Answer: change implemented.
L47 “sp.” After genus name should not be italicized
Answer: change implemented.
L84 add “with or without paludiculture” or similar after “peatlands”
Answer: “including paludiculture” was included in the sentence.
Figure 1: Generally: Panels should be arranged in rows as abcd, white space should be removed. (A) Why is why is this map oval? remove unreadable text (google earth etc.), use 2-dimensional marker. (B) red on green is hard to read, explain letters/blocks, text needs to be larger.
Answer: The figure was edited by adopting all recommended changes. The map was also changed as recommended by Reviewer 1.
Figure 2: This figure needs to be reworked to improve the space/infomation-ratio and quick readability. Both panels should be combined.
Answer: The proportions were adjusted in figure to make a better use of space. We consider that panels A and B should remain independent because they convey important information that can be presented more clearly this way. Panel A shows the changes in the WTD as the peatland was rewetted (from 2019 to 2022), while panel B shows differences in WTD between the blocks in this study (2021 to 2022). A rectangle was included to make clear that panel B zooms in a section of panel A figure. Caption was also edited.
Figure 3: I think this would work better as a table.
Answer: We did consider using a table for this information previously, however, since it is rather a limited amount of information and we already have several tables in the manuscript, we decided to include this information as a figure. We consider this information is better as a figure because it also gives a visual break to the reader.
L261: “electroconductivity” with lower case “e”
Answer: Change implemented.
L314: “GPP was negatively correlated to Ts…” This kind of wording is always very confusing when talking about negative values. Add “(larger CO2 uptake)” or similar.
Answer: An explanatory sentence was included.
Figure 4: What is your main point for this figure? If it is the direction of arrows for fluxes and environmental parameters, the colours should be chosen so that these arrows stand out more, and the labels should be larger. Right now, the feature that stands out most is the oval marking the measurements of block A. I assume, this is not what you want to lead the readers to see first and what is the most interesting about this figure?
Answer: The main point of this figure is to show the existing correlations in the data using the PCA plots but also to show the heterogeneity of the blocks. Therefore, both the arrows and the colored ovals are important in the figure. The figure has been edited to highlight the arrows, the color scheme of the blocks has been changed and the labels have been enlarged.
Figure 5: x-axis labels are far too small.
Answer: the figure and captions were edited.
L358: “higher”à see comment above, add “(more positive) or similar to make the direction clearer
Answer: Change was implemented. Additionally, wording on GPP was revised and edited accordingly in the manuscript.
Figure 6: Text in the figure is very hard to read and should be larger. The colours (blue and green) are too similar and hard to distinguish
Answer: Figure was edited by increasing text size and changing the color scheme.
Figure 7: Text too small, lines too thin, red/green contrast should be avoided to make figure readable for people with colour-blindness.
Answer: Figure was edited by increasing text, changing color scheme, and making lines thicker.
Figure 8: See Figure 7.
Answer: Figure was edited by increasing text and changing color scheme.
L594: “.” Or “;” instead of “,”?
Answer: Change implemented.
Figure A1: In its present form I don't know how useful this figure is, at least I cannot read almost any of the variable names?
Answer: This plot with results from correlations supplements the results seen in PCA plots. It provides further information on individual correlations. We included this plot as an appendix to give the reader the possibility to explore these individual correlations. Since Appendix information is accessed online, we expect readability to be better, however the text in the labels was increased and the resolution of the plot was also improved.
Citation: https://doi.org/10.5194/egusphere-2024-3030-AC2
-
AC2: 'Reply on RC2', Andres Rodriguez Grisales, 05 Mar 2025
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 333 | 95 | 188 | 616 | 11 | 12 |
- HTML: 333
- PDF: 95
- XML: 188
- Total: 616
- BibTeX: 11
- EndNote: 12
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 132 | 21 |
| Denmark | 2 | 85 | 13 |
| Germany | 3 | 58 | 9 |
| Netherlands | 4 | 36 | 5 |
| China | 5 | 31 | 5 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 132