Dear Dr Bai,

Many thanks for responding to the referee comments. Whilst both referees consider the study to be of interest to readers, there are some areas of uncertainty associated with your method where I would ask you to revise your manuscript. Referee 1 registers some significant concern regarding the use of global data bases to extract site specific information. I think it is important to validate the data you obtain on climate and total C stocks against values published. I appreciate that this will not be the case for all studies, but where these are available, you should indicate whether your assumed values (from data bases) match those reported from sites.

I agree with both referees that this is an interesting and potentially informative approach that can illustrate the connection between soil carbon turnover and climatic conditions, allowing it to be potentially extrapolated more widely. For this reason, it is extremely important that the method is robust and information underlying the approach is validated meaningfully.

I hope that you can agree with this assessment, and I look forward to seeing a revised manuscript and detailed response to the points raised by referees soon. Please note that given the substantive concerns, I will consult with referees again before reaching a final decision on this manuscript.

Best regards,
Jens-Arne Subke

Dear Dr Bai,

Many thanks for submitting your response and a revised manuscript. I think that you incorporate the concerns of referees well, and also think that the suggestions by Dr Balesdent (referee 2) to incorporate further data from an existing data base, but also remove ecosystems where past occurrence of C4 vegetation impacts on the distribution of 13C in the soil profile have made this version stronger.

Concerns by referee 1 regarding the methodology of deriving soil C stock and respiration values are still valid, and it’s good that you now acknowledge these as uncertainties. I can think of further modification to the method (e.g. using not total precipitation, but a parameter that also includes potential evapotranspiration, as this ultimately determines the moisture status if a soil), but I’m happy for this study to be published to establish a method for constraining SOM turnover estimates based on isotopic soil profiles. Subsequent improvements will then hopefully mean a more robust methodology can be developed, based on your initial approach. I do, however, have a number of further requests for edits on your manuscript, which are needed before the manuscript can be finally accepted. Please see the points below, which I hope you are able to incorporate.

Best regards,
Jens-Arne Subke


76: Supplemental table S1 has 177 entries, which matches your statement in the abstract, but not what is being said here. Further, entry 196 in the supplemental table lacks results, so is seemingly not included? (I.e. should total number of profiles be 176?)

103-105 Replace “We applied […] millennial time-scales” by: “In order to calculate the SOM decomposition rate constant, we assumed that input of organic matter to soil and decomposition are in equilibrium with no change in SOM stocks over time (steady state assumption). We acknowledge that this assumption is not valid in most disturbed environments, including agricultural systems, but it provides a reasonable approximation in natural ecosystems, where SOC turnover has equilibrated on century to millennial time-scales.” (This is a suggestion from me, please feel free to rephrase. I think that simply stating “steady state assumption” was not clear enough and that a slightly more elaborate justification was needed.)

228-234: This section is important, but it is poorly integrated into the overall discussion. These points reflect the concerns of referee 1 well, but the reader is left not knowing how significant these uncertainties are for your findings. I suggest moving this section up into the main part of the discussion and qualifying whether you think that the patterns you observe in your study are robust despite these limitations. I’m not asking for a lengthy elaboration of implications, buy a clearer indication of whether, in your judgement, these limitations are a major limitation to your conclusions.

Figure 4: In panel B, the axes are seemingly labelled incorrectly. To match data in panel A, k seems to be scaled on the y-axis, and ln(-beta) on the x-axis, but your axis labelling suggests the opposite.

Dear Dr Bai,

Number of soil profiles included is stated as "149" in line 76, contradicting other occasions in the manuscript and the author's response. Please amend to read "176".

Otherwise this manuscript is ready for publication.

Jens-Arne Subke