We would like to thank the reviewer for their detailed feedback on this manuscript.

One major suggestion I have is to improve the EC-based analysis of the contribution of fluxes from ditches. The authors mention that what makes their study unique is the inclusion of ditches at active harvest sites. Throughout the manuscript, they refer to a chamber-based study that showed high fluxes from ditches, compared to the field. But for methane, the total CH4 fluxes from the tower and the combined ditch and field fluxes from the chambers are off by an order of magnitude. The authors do not mention how much of the tower footprints are comprised of ditch area which I think is important to know in this context. Furthermore, the authors should consider disaggregating the ditch and field flux components as is being increasingly done in many heterogeneous footprint analyses (Ludwig et al. 2024 https://doi.org/10.5194/bg-21-1301-2024; and Pirk et al. 2024 https://doi.org/10.1029/2024GL109283 as two examples). An unmixing approach has the potential to clarify some of the more confusing aspects in the manuscript related to ditch contribution and mismatch with chamber measurements. This approach could also strengthen some of the relationships between temperature and wtd for the field fluxes. If ditches are a major component of the tower flux as suggested by the authors, then the soil temperature and water table relationships with the field fluxes would be dampened by the over-powering signal from the ditches, which are likely to have a different relationship with env. variables as is common for aquatic ecosystems. 

Thanks for these comments. You make a good point that the chamber-based ditch measurements from Hunter et al., (2024) are higher than the field and ditch integrated fluxes we measured in this study. We agree that we should provide the relative contribution of ditches and fields to the overall flux footprint to give context to our measured fluxes. We therefore propose incorporating analysis of the relative proportion of ditches to the overall footprint in the revisions of this manuscript.

The ratio of fields to ditches is about 30:1. Since the ditches cover such a small area of the tower footprint, we believe that disaggregation will be difficult. There are unlikely to be periods where the ditches make up the majority of the tower footprint.

Line 41. Can the authors provide some context for how significant 2.1 Mt Co2 equivalents is? How does this compare to other sources? Some sort of context will help the reader interpret the potential significance of harvested peat site C emissions, as mentioned in the preceding sentence.

That is a good idea, thanks. In consultation with our industry partners, we were made aware of an updated value of 1.6 Mt CO₂ equivalent in the March 2025 publication of the Environment and Climate Change Canada National Inventory Report. We will update this number and compare it to estimates of other wetland disturbances in Canada.

Line 48. I am confused what the 15-40 year range is referring to. How long the method has been used in Canada? How long the sites are extracted for? Later on in the paper the authors use a different year range for how long sites are extracted.

Thank you for this comment. Yes, the 15-40 years refers to how long the sites are extracted for. We will clarify this, and make sure it is consistent throughout the paper.

Lines 66-67. Here the authors should provide some more details on Canada’s Tier emissions system. Those outside of Canada might not be familiar. How are the Tiers used? Is Tier 2 high or low compared to other Tiers? This topic comes up again in the discussion so I think it would be very helpful to provide more information here.

Thanks for this suggestion. We will add in more context about the Tier emission system in the revised manuscript.

The Intergovernmental Panel on Climate Change (IPCC) introduced a three-tier system for emission factors to convey the type of data used, and the accuracy of the estimates. Canada currently uses a Tier 2 approach for greenhouse gas emissions from extracted peatlands, but with data largely collected from post-extraction peatlands. An improved Tier 2 emission factor, which would include Canadian specific national measurements of CO₂ and CH₄ emissions from actively extracted sites, would improve the accuracy of our reported emissions. A Tier 3 approach often includes the use of process-based models for emission estimates. This study, along with other recent work by Clark et al., (2023), He et al., (2024) and Sharma et al., (2025) can support Canada in an improved Tier 2 approach or development of a Tier 3 approach.

Line 67. The sentence on methane suggests the authors will present annual fluxes in this paper. Some indication of summertime methane fluxes from other studies could be of value here.

 This is a good idea, thanks. We will add in a range of summertime chamber methane fluxes from Clark et al., (2023), Hunter et al., (2024) and Sundh et al. (2000).

Line 73. Many folks present WTD positive and negative signs differently. Perhaps clarify here what it means to have a “positive relationship”

Thank you for this comment. We know that conventions vary about using positive and negative values for WTD in peatland papers. We will clarify this sentence. By positive relationship, we mean that as water table becomes further from the surface, CO₂ emissions also increase.

Line 75. “was less than 50 cm below the surface”  (this will help clarify water table direction)

Thanks for pointing this out. In line with your previous comment, we will make sure to define that a positive WTD is when the water table is below the surface.

Line 123. Can the authors provide a comparison with available air temperature data too?

Thank you, we will add that in the revised manuscript.

Line 128. Please describe how the instruments were placed in the peat

Thanks for this comment. The current sentence is “These instruments were placed at the center of a field in a harvesting exclusion zone near each tower.”  We will replace it with the following sentence. “These instruments were installed in a 2 m deep PVC well located at the center of a field in a harvesting exclusion zone near each tower.”

Line 153. Please describe what a u* threshold is for those who might not be as familiar with EC methods :)

Good idea! We will add in a statement about how the u* threshold parameter is used to identify periods of low turbulence which do not meet the assumptions of the eddy covariance technique (Burba, 2013).

Burba, G. (2013) Eddy Covariance Method for Scientific, Industrial, Agricultural and Regulatory Applications: A Field Book on Measuring Ecosystem Gas Exchange and Areal Emission Rates. Li-Cor Biosciences, Lincoln. DOI:10.13140/RG.2.1.4247.8561

Line 158-159. This would be a good section to add more information about the average percentage of ditch area in the footprints.

 Thanks for this suggestion.  See our comments at the start of this document for how we suggest we provide more context to the readers on the relative contribution of field and ditch fluxes to the overall flux.

Lines 164-174. Either here or elsewhere in the manuscript, the authors should describe the limitations of not having nighttime data. Do daytime fluxes over estimate total daily fluxes? Underestimate? Do any of the European sites include night-time data that could help provide rational for applying the daytime fluxes across the 24-hour period as done here (line 185) ?

Thanks for this comment. Due to low turbulence conditions at night, the majority of our nighttime data did not meet quality control standards. Due to the absence of vegetation at the site, the biogeochemical processes responsible for CO₂ exchange should be the same during the day and night. By using only daytime data, we are likely overestimating the total CO₂ flux due to cooler temperatures at night. We will add a sentence to the methods, and to the discussion to make the implications of this methodological choice clearer.

Lines 183-184. It is a bit unclear why the authors chose to combine the data, especially since later on in their results they state that site was a significant factor for CO2 fluxes.

 Thanks for this comment. For national reporting in Canada, emission factors are heavily reliant on data from post extracted unrestored peatlands. These sites are not representative of actively extracted peatlands, as they may have partial revegetation, and may no longer be actively drained. We wanted to use our data to improve annual emission factors to be applied at a national level. By combining data from multiple years, and from two sites operated by different companies, we aim to arrive at a CO₂ and CH₄ emissions factor value that is representative of Canadian extracted peatlands.

Line 189-190. A sensitivity analysis of the non-growing season month contribution could be beneficial here. How much does the total annual budget change as this value is adjusted? Some studies from natural peatlands suggest upwards of 50% of the annual flux could be from the non-growing season (Treat et al. 2018 GCB).

We agree that we should have a clearer discussion of the sensitivity of our non-growing season assumptions to our annual CO₂ and CH₄ budget. We attempted to do this by using three different methods to estimate the CO₂ flux contribution from the non-growing season. We propose that in our discussion section, for each annual CO₂ and CH₄ estimate, we can include an error term that accounts for a 15% increase and decrease of the non-growing season flux estimate as a sensitivity analysis of the assumptions used to compute non-growing season emissions.

Line 339. Does this mean there were no significant relationships at 10cm for AB and 20 cm for QC?

Yes, it does. We will edit this sentence to improve its clarity.

Line 343-344. It would be helpful to add sampling sizes here for each water table group

Thanks for this suggestion. We will add the sampling size to both the figure caption, and the main body text to improve the clarity of the results.

Lines 381-384. What is the field:ditch ratio in AB? Some of the mismatches between EC and chamber methods mentioned here could be resolved by unmixing the EC fluxes, as mentioned previously.

The field to ditch ratio is 30:1. See our comments at the start of this document for how we suggest we provide more context to the readers on the relative contribution of field and ditch fluxes to the overall flux.

Lines 398-399. I am not quite sure I am convinced by the authors' suggestion that their findings conflict with previous results as this section is currently written. The authors go on to say that the WTDs for the other studies are much shallower (line 411). Do the results actually conflict, or is it a matter of differences in relative changes in WTD between the sites? Providing the WTD ranges for the other studies mentioned here could help clarify the magnitude of “conflicting” results.

Thank you for this feedback. The goal of this section was to justify our finding of decreased heterotrophic respiration during very dry periods, when many studies report the opposite. You are right that our results are not conflicting with other studies. Instead, they help to expand our understanding of CO₂-WTD relationships because our measurements were conducted at heavily drained sites. We agree that it was a poor choice to use the word conflicting. We will edit this discussion section to better capture our intended message.

Lines 443-444. Many natural peatland studies also show deeper soil temperature is a better predictor in general. See Heffernan et al. 2024 Global Change Biology as an example.

 Thank you. We will add references to relevant literature, such as the paper you suggested, to support our finding.

Lines 483-485. What is the rational for including negative fluxes when ecologically speaking, there should be no uptake? How do the authors know this is not an instrument or analysis error? It is unclear what the reasons are for including negative fluxes if there is no biological explanation.

 The eddy covariance methods, along with data processing steps required to convert the measured signal into a meaningful flux measurement are expected to produce on average an accurate estimate of flux, but with noise around that signal. This noise will result in both positive and negative values around the mean resulting from this noise. These negative fluxes remained after testing a range of data processing steps and finally selecting the approach we determined to be most correct given the site conditions. In order to avoid biasing the data, we chose to present the fluxes as computed. Further, there is recent research that identifies the potential for CO₂ uptake by phototrophic microbes in peat soils, presenting a possible biological explanation for the measured fluxes (e.g., Hamard et al., 2021). This small uptake would be unlikely to be noted in other systems as it would be attributed to vegetation, but could be indicative of real soil CO₂ uptake that we are able to see here due to lack of vegetation (although we did not set out to measure this at our site). We will add this reference to the discussion as a possible biological explanation, noting that more research is needed to confirm this finding.

Hamard et al. 2021. Contribution of microbial photosynthesis to peatland carbon uptake along a latitudinal gradient. Journal of Ecology, 109, 3424-3441.

Lines 490-493. I found this section confusing to follow. Perhaps it could be clarified by also including what the ditch-only chamber flux values were. Also refer to my main comment about separating fluxes for the ditch and field components.

Thanks for this feedback. We will edit these sentences to provide more context for the site integrated fluxes were report from Hunter et al., 2024. In that study, the authors used a field to ditch ratio of 30:1 to combine their chamber field and ditch measurements into a sitewide flux value. We agree that it would be helpful to provide their ditch only chamber flux values in this section.

Line 532-533. Just flagging this sentence as a reminder to provide more information about the Canadian Tier emission factors and their relevance/importance to this study, as mentioned in the introduction earlier.

Thank you for flagging this. See our above response about adding in context for Canadian Tier emission factors in the revised manuscript.

Line 547. Does "fluxes" refer to CO2 fluxes here or CO2 and CH4 fluxes?

Thanks for catching this. We will add the word “CO₂” to make it clear that we are referring to CO₂ fluxes and not CH₄ fluxes

Line 544. Can the authors provide references for existing process-based models that use C-temperature relationships? That would be a good addition here.

This is a good idea thanks! We will add in a reference to He et al., (2023) in Ecosystems, which modelled CO₂ emissions from actively extracted peatlands in Quebec, Canada.

We would like to thank the reviewer for their detailed feedback on this manuscript.

However, I have major concerns about the data analysis and results. Particularly, they used the Gaussian function to describe the relationship between CO2 flux and WTD. Based on the relationship, one can easily conclude that drained peatlands emit a lot of CO2, simply because the drainage is not deep enough. This kind of message could be misleading. There are a few problems that I can see in the analysis of concluding the relationship. Firstly, data from the two sites were pooled together to get the relationship while the declining CO2 flux at the high WTD end is mostly driven by data from QC; however, the QC data alone don’t really show a significant decreasing pattern. The two sites showed contrasting WTD throughout the seasons, that could make the two sites fundamentally different (e.g., soil properties) in the long-term even though the managements are similar, and thus the fluxes and their relationships with environmental factors. So I feel they should be analyzed separately. Secondly, the authors claimed that CO2 uptake is absent (line 155) in the ecosystem. I believe the moss photosynthesis can’t be ignored in such a system. Plus, figures 5 and 6 clearly showed a few negative points. Therefore, I suggest considering light conditions in the analysis. Thirdly, while weekly average may show you a clear pattern of the relationship, readers are interested in seeing hourly/30-min data plotted in Figure 5.

Thanks for these comments. Overall, we believe that the data we have presented supports the finding that heterotrophic respiration can be reduced under dry conditions. We combined data from both sites in Figure 5 to show how CO₂ fluxes varied across the whole range of WTDs. This allowed us to capture the range of precipitation and management conditions and to illustrate that the results were consistent with the broader literature on soil respiration that shows this type of unimodal response to soil moisture (for which we chose a Gaussian response).

During our review of the relevant literature, we found that there are a limited number of studies at drained peatlands with WTDs greater than 50 cm. We thus believe that our finding that CO₂ respiration can decrease under heavily drained conditions is an important contribution to understanding the mechanistic processes of decomposition. We agree that we don’t want to create a misleading narrative about the impact of drainage and vegetation removal on CO₂ fluxes. We will add in a sentence in the discussion that explicitly states that any drainage and vegetation removal during peat extraction shifts these systems from net sinks to net sources of CO₂.

We realized that we were not clear enough in the background and methods section. This is an unvegetated actively extracted site. In preparation for extraction, the site managers removed all surface vegetation. This vegetation remains absent for the duration of extraction.  We will clarify this in second paragraph of the introduction and in section 2.1. Due to the absence of vegetation, we assumed that all CO₂ fluxes were due to heterotrophic respiration, and thus did not conduct analysis under varying light levels.

We like your idea of showing the hourly data CO₂ -WTD data. We suggest that we keep the weekly averages data in Figure 5, and then add a plot of the weekly averaged data to the supplemental file.

L25: “small sources”, how small? Report the numbers here too

Good idea. We will include the mean May to August CH₄ flux of 7.22 mg C m^-2 d^-1 in the updated abstract.

L170: only daytime data were used. How does this affect the results and budget.

Due to low turbulence conditions at night, the majority of our nighttime data did not meet quality control standards. Due to the absence of vegetation at the site, the biogeochemical processes responsible for CO₂ exchange should be the same during the day and night. By using only daytime data, we are likely overestimating the total CO₂ flux due to cooler temperatures at night. We will add a sentence to the methods, and to the discussion to make the implications of this methodological choice clearer.

L184: apparently, fluxes of the two sites are not completely comparable. Combining data and averaging everything is a very crude strategy. Plus, the authors used only daytime data. How much uncertainties/biases are in the final budgets?

Thanks for this comment. For national reporting in Canada, emission factors are heavily reliant on data from post extracted unrestored peatlands. These sites are not representative of actively extracted peatlands, as they may have partial revegetation, and may no longer be actively drained. We wanted to use our data to improve annual emission factors at a national scale. By combining data from multiple years, and from two sites operated by different companies, we aim to arrive at a CO₂ and CH₄ emissions factor value that is representative of Canadian extracted peatlands.

L190: NECB should also count the harvested biomass.

We agree that the quantity of peat that is removed during extraction, and the fate of this peat, should be included in a full net ecosystem carbon balance (NECB) of the whole extraction process. For this study, our scope was to focus on the NECB of the extracted peatland site. In national emission factor reporting, they report an emission factor of the removed peat, and an emission factor of the extracted peatland site. We designed our research methods to align with this reporting practice. We will add in a sentence in the methods and discussion sections of the paper to give the readers this context and direct them to recent work on the fate of extracted peat, such as Sharma et al., (2025).

Sharma B, He H, Roulet NT. (2025). CO₂ emitted from peat use in horticulture supports a lower emission factor. Carbon management, 16(1). doi:10.1080/17583004.2025.2468476

L200: since continuous measurements are also a form of repeated measures, I suggest using a mixed model with dates/time as random effect.

This is a good suggestion, thanks. We will consider adding in month as both a fixed and random factor in our linear model as we revise this manuscript.

L207: How about using a GAM to gain more flexibility? Also, using bootstraps could give you an idea of the uncertainties of the relationship as well as the range of the optimal WTD.

 We like your idea of using bootstraps to determine the uncertainty with the modeled parameters. We will incorporate this into our revised analysis. We chose to use a Gaussian model to provide flexibility for future applications of the data. Future modelling work can incorporate this relationship into their model.

Fig. 6: the interaction between WTD and Ts is the key to this study. I would rather be more interested to see how Ts affects the CO2 vs WTD relationships.

You make a good point that when there is a significant two-way interaction (in this case the interaction between WTD and Ts), you can’t know the direction of the interaction. We chose to divide the data up by WTD since there was a larger range of observed WTs than soil temperatures. We propose that in the main text we add in analysis of the CO₂-WTD relationship when the data is binned by soil temperature. Depending on the results, we can either place the newly created figure in the main text or in the supplemental file.