The present research tried to tackle three relevant questions about global warming and denitrification at once. Not all was answered here in spite of the interesting experimental design but the authors did manage to bring up some relevant new questions. Particularly intriguing to see that mineral soil would be able to reduce N2O emitted from the organic soil layer, in spite of lower nosZ abundances.

The introduction sets the stage well and makes a case for the presented work. I agree with the authors in that the three hypotheses are sufficiently well underbuilt and no further elaboration is needed. In response to comments raised by referees some further methodological clarifications are now presented as well. The results were already efficiently presented. The discussion equally reads well, but I have a few comments:

While the methodology is relatively well thought through but it is somewhat regretful that incubations were carried out in a completely O2-free atmosphere, which really renders the results less representative. This needs to be brought up in the discussion near L434: N2O reduction was probably strongly promoted in these experiments with nearly no O2 present. In a field situation it may well be that N2O-reduction in the quite porous mineral layer is a much lesser likely process and so would then the ‘soil horizon interactions’ be a lot smaller.

I partially do agree with referee 3 that generally observations on N2O emission and reduction are not very much discussed mechanistically, viz. with respect to drivers of these processes. But it is equally understandable that the authors chose not to do so as these experiments were run in O2 free air. One occasion where I hope the authors could still complement is L450: just why in your view would N2O reduction be particularly favoured in the mineral topsoil vs. organic layer? Perhaps because with a smaller porosity and probably also finer pore size distribution vs. the organic layer hot-spots for complete denitrification have always been more common in the mineral soil. This would then have led to a microbial community more fit (efficient) to reduce N2O? Or perhaps, N2O reduction is mostly simply uncommon in the organic layer as it is too porous and more directly in contact with the atmosphere so that N2O residence times are just too short to allow further reduction. I did find the hypothesis on L138-142 plausible but again, at present it is not discussed why the opposite result was found. It is for referees or readers not possible to reflect a bit on this with very little information provided on the sampled soil horizons. To the least also soil texture and SOC concentration should be added to table 1, but perhaps the authors also still see some room to further contemplate on just why mineral soil would harbour a microbial community more fit for N2O reduction.

I recommend publication after final minor revision.

Some smaller comment:
L59 I suggest you omit “the emergent”
L70-79 should be shortened a bit. The message is clear but lengthy
L131 replace “predicted” by “expect” or “hypothesize”
L779 at the end of a 60 h incubation +
Figure captions: in my view no need to repeat the three names of the sites once more, “coolest” etc. suffices
L794 C instead of (B); In tab A the orange dot at 25°C is invisible, is that correct?
L400 I wonder if “our predictions” is really the best term here. You are mainly bringing up an expectation/hypothesis, no?
L425 ‘increased net N2O production to temperature’ sounds awkward
L426 ‘did not contradict’ seems a bit overly careful, should the authors want to then this phrasing could be omitted.
L442-444 is not that readable; the phrasing “our horizon interactions” should be reworded. Probably this sentence can do without “although these results do not contradict the possibility of mineral soil reduction”

The present research tried to tackle three relevant questions about global warming and denitrification at once. Not all was answered here in spite of the interesting experimental design but the authors did manage to bring up some relevant new questions. Particularly intriguing to see that mineral soil would be able to reduce N2O emitted from the organic soil layer, in spite of lower nosZ abundances. The introduction sets the stage well and makes a case for the presented work. I agree with the authors in that the three hypotheses are sufficiently well under-built and no further elaboration is needed. In response to comments raised by referees some further methodological clarifications are now presented as well. The results were already efficiently presented.

The discussion equally reads well, but I have a few comments:
While the methodology is relatively well thought through but it is somewhat regretful that incubations were carried out in a completely O2-free atmosphere, which really renders the results less representative. This needs to be brought up in the discussion near L434: N2O reduction was probably strongly promoted in these experiments with nearly no O2 present. In a field situation it may well be that N2O-reduction in the quite porous mineral layer is a much lesser likely process and so would then the ‘soil horizon interactions’ be a lot smaller.

I partially do agree with referee 3 that generally observations on N2O emission and reduction are not very much discussed mechanistically, viz. with respect to drivers of these processes. But it is equally understandable that the authors chose not to do so as these experiments were run in O2 free air. One occasion where I hope the authors could still complement is L450: just why in your view would N2O reduction be particularly favoured in the mineral topsoil vs. organic layer? Perhaps because with a smaller porosity and probably also finer pore size distribution vs. the organic layer hot-spots for complete denitrification have always been more common in the mineral soil. This would then have led to a microbial community more fit (efficient) to reduce N2O? Or perhaps, N2O reduction is mostly simply uncommon in the organic layer as it is too porous and more directly in contact with the atmosphere so that N2O residence times are just too short to allow further reduction. I did find the hypothesis on L138-142 plausible but again, at present it is not discussed why the opposite result was found. It is for referees or readers not possible to reflect a bit on this with very little information provided on the sampled soil horizons. To the least also soil texture and SOC concentration should be added to table 1, but perhaps the authors also still see some room to further contemplate on just why mineral soil would harbour a microbial community more fit for N2O reduction.
I recommend publication after final minor revision.

Some smaller comment:
L59 I suggest you omit “the emergent”
L70-79 should be shortened a bit. The message is clear but lengthy
L131 replace “predicted” by “expect” or “hypothesize”
L779 at the end of a 60 h incubation +
Figure captions: in my view no need to repeat the three names of the sites once more, “coolest” etc. suffices
L794 C instead of (B); In tab A the orange dot at 25°C is invisible, is that correct?
L400 I wonder if “our predictions” is really the best term here. You are mainly bringing up an expectation/hypothesis, no?
L425 ‘increased net N2O production to temperature’ sounds awkward
L426 ‘did not contradict’ seems a bit overly careful, should the authors want to then this phrasing could be omitted.
L442-444 is not that readable; the phrasing “our horizon interactions” should be reworded. Probably this sentence can do without “although these results do not contradict the possibility of mineral soil reduction”

All previously criticism raised has been appropriately addressed. The manuscript is now nearly ready for publication in SOIL. Only a two minor technical issues remain to be resolved.

1° All equations should be numbered sequentially with Arabic numerals in parentheses on the right-hand side, e.g. (1), (2). If too long, split them accordingly. When using Word, the equation editor and not the graphic mode should be used under all circumstances. They should also be referred to in the text by the abbreviation "Eq." and the respective number in parentheses, e.g. "Eq. (14)". However, when the reference comes at the beginning of a sentence, the unabbreviated word "Equation" should be used, e.g.: "Equation (14) is very important for the results; however, Eq. (15) makes it clear that..."

2° En dashes (–) are used to indicate, among other things, ranges (e.g. 12–20 months). -> check L172, 176, 230,

Dear Authors,

First of all, thank you for chosing SOIL as an outlet for your work! Also, thanks for working with the reviewers and the topical editor to improve the manuscript. It is with great pleasure that I can accept the manuscript after some further minor revisions as suggested by the Topical editor:

1° All equations should be numbered sequentially with Arabic numerals in parentheses on the right-hand side, e.g. (1), (2). If too long, split them accordingly. When using Word, the equation editor and not the graphic mode should be used under all circumstances. They should also be referred to in the text by the abbreviation "Eq." and the respective number in parentheses, e.g. "Eq. (14)". However, when the reference comes at the beginning of a sentence, the unabbreviated word "Equation" should be used, e.g.: "Equation (14) is very important for the results; however, Eq. (15) makes it clear that..."

2° En dashes (–) are used to indicate, among other things, ranges (e.g. 12–20 months). -> check L172, 176, 230,

Best regards,
Jo