Interactive comment on “ Carbon balance of a restored and cutover raised bog : Comparison to global trends ”

Reviewer 1 General Comments: What’s the paper about: In this ms Swenson et al attempt to estimate C-balance of five managed raised bogs in Ireland. The five sites differ in their hydro-physical and ecological characteristics, as defined by the National Parks and Wildlife Servies of Ireland. The strength of their study is in their attempt to estimate total C-balance of each site based on measurements of various aspects of ecosystem carbon balance during the same season, including: ecosystem CO2-gas flux, ecosystem CH4-gas flux, aquatic fluxes of dissolved organic and inorganic carbon and CO2 gas efflux from open water in drainage ditches. Furthermore, they provide a literature compilation and review of studies that have measured ecosystem CO2 and methane flux in boreal and temperate, managed and intact, peatlands. They use the

tion is included in the text in line with Reviewer 1's comments.This makes the use of the term significant is clearer.
Manuscript changes: Two instances of "tons" changed to "tonnes" in the manuscript.
R2 COMMENT: Fig. 10 Given that MAWT is used as a predictor variable in the models, these observations are not independent (especially as the collars were lumped together for modelling) and I am far from convinced as to their value in this manuscript.
Response: Hourly water table (as opposed to MAWT) data was used as a parameter in modelling NEE, but not CH4 flux.The NEE was modelled using collar specific empirical models fit to field data, and water table had a minor impact on the over modelled results.Thus, the changes in hourly water table help explain the variability in NEE but do not strongly control the modelled annual NEE.I argue that these are independent variables and the trends in these plots are useful results for comparing with other studies.These plots are valuable because they show the interaction between ecotype, genus percent coverage, MAWT and NEE, CH4 flux, GWP.
Manuscript changes: The plots in figure 10 were changed to include different symbols for the five ecotypes (as per comment by Reviewer 1), which improves the value of this figure in the manuscript.

R2 COMMENT: L490 five decades
Response: OK Manuscript changes: "5 decades" changed to "five decades" R2 COMMENT: Section 4.2 I am not sure of the value of this section.The manuscript is already quite long and this seems superfluous (especially given the extensive data set in the Supplementary).If it really must be kept, then it should be moved to the Results section and then discussed here.
Response: After re-reading the manuscript, I agree that much of this discussion is unnecessary and a little off topic for the main findings of the paper.However, the literature comparisons in fig 11 and 12 are valuable as are the extensive data sets behind them included in the supplementary tables.This information puts the study in a broader context.Manuscript changes: Figure 11 and Figure 12 have been moved to the results section, and the discussion in section 4.2 has been substantially shortened. C12 with each parameter estimate.Given the large number of parameters in the models, I would suspect that the SE will be very high and would invalidate your approach.
Response: Including the SE of the model parameters was also suggested by Reviewer 1.That will be included in these tables.
Yes, there are a large number of fitting parameters.The same GPP and ER models were used for all 29 collars, but the model fit parameters were determined empirically for each of the collars individually.For the GPP modeling, there was sufficient field data to justify a model with 5 empirical fitting For the majority of the collars, all of the GPP model fit parameters are significant to 95% confidence.This model was designed in such a way that the effect of modelled parameters reduces to zero as the explanatory value of the additional variables decreases such that insignificant model parameters have a minor impact on the modelled results.For the ER modeling, the sample size is smaller and the point that Reviewer 2 is making is quite valid.Previously, there had been some debate among the authors as to which of two models to use for ER, so much so, that information was included on both of these models in the supplemental section in the original manuscript draft.Thus, the simpler ER model (with 3 fitting parameters compared to 5 fitting parameters) has been used to calculate ER in the updated manuscript.This model was taken directly from Wilson et al., 2016b and developed from the same type of data, collected in Ireland.
Manuscript changes: Additional statistical information has been included in the table S1 and S2 of the model fit parameters.The ER has been calculated by a different and simpler model in the revised manuscript, which had been previously described in the supplemental section but not used in the manuscript and was taken directly from Wilson et al., 2016b.The change in ER model had a minor effect on the overall conclusions of the paper.Also, the text of this supplemental section has been substantially revised to clarify confusion expressed by both Reviewers.
Please also note the supplement to this comment: C16 R2 COMMENT: L436 Consider Nugent et al (2018) Global Change Biology, https://doi.org/10.1111/gcb.14449C11 Response: This study was not published at the time of initial submission, but is very relevant.Manuscript changes: This suggested publication has been included and cited in section 4.1.R2 COMMENT: L477 1960s Response: OK Manuscript changes: "1960's" changes to "1960s" as per previous comment R2 COMMENT: L484/485 join the sentences Response: Yes, this should be one sentence.Manuscript changes: Lines 484/485 have been joined as one sentence.