Dear authors,

Many thanks for your revised manuscript. After consulting the reviewer comments and your responses, as well as my own reading of the revised manuscript I think that major revisions are required before this manuscript may become publishable in Biogeosciences. My main criticism is that it does not become clear, what the new understanding gained from this new model is. I believe that this is mainly a question of the structure of the manuscript, and I make some suggestions for making the story more transparent below.

I also think that the revised manuscript should contain some of the responses to the reviewer that you provided (see below) in the discussion section.

When revising your manuscript, please provide a point-point explanation of what has been changed.

For somebody, who is not familiar with the Kirkby model, it is hard to see what your new developments are compared to the original model formulation. Try to emphasise more your new parts relative to the original model. I would suggest to move the equations of the original model into a second appendix, summarising only the key points in Section 2, which are required to get what the model is doing. I would revise Figure 1 to illustrate which parts of the model you introduced (maybe as dashed versus solid arrows), such that the reader can better appreciate your contribution. Figure 2 can be safely moved to this new Appendix, or removed.

The methods, results and discussion section as very interwoven, which may seem logical when you are familiar with the topic, but which makes it hard for a reader to follow, because sometimes it remains unclear whether you state a model assumption, explain a model result or refer to something the model should be doing. I would recommend to more closely follow the standard structure of a journal article (methods, results, discussion).

In particular, I would move the text in Section 4 prior to Section 4.1 in a new Section 3.1, which describes the model set-up and runs

Similarly for Section 5, the text before p25 line 4 (goes to Section 3.2). The point here is that you not only introduce a new set of site conditions and runs, but a lot of other information, which distracts from your model results. It also introduces a new model equation, which should not be presented here, but where you describe the model.

Move the discussion texts in page 19 l 2 ff into a separate discussion section. At this point in the text it is sufficient to write “As discussed in Section X (the discussion section),the weathering sequence of basic oxides displays similar properties than in other studies.” or something along the lines. The rest of the text distracts from comparing this result to the other model formulations. It is unclear to me, why you discuss Section 4.1 in depth, but do not provide a similar assessment for the other terms?

Similarly the discussion of K and P in the chronosequence p 25 l 12 to p 26 l 4 would be best located in a separate discussion.

I’m stressing the need for a separate discussion, because my impression is that currently the strengths and weaknesses of the model are buried in the results section. Having them altogether is a separate section would make it easier to appreciate the quality of the model (and its limits).

3) The text following page 28 l 15 in the Conclusion section aren’t conclusions, but a discussion of the results - they should go into the discussion section. Please limit the conclusion section to things you have learned from your study, or needs for further research that you’ve identified with your simulations.

Abstract: mentioning “vegetation interactions” implies that the vegetation plays an active role in your model (otherwise it would not be interacting). After reading your model description I think that this is not correct, because vegetation production and water-use are prescribed, and the only thing that varies is litter stoichiometry, which however is not actively controlled by the plants, but simply follows soil nutrient availability. I would therefore rephrase this to be “effect of vegetation in terms of ....”. Only then the meaning last two sentences of the abstract become clear.

Minor comments.

Section 2.6. The model seems to be partitioning the Np into different pools, but how this is done remains unclear. Please specify. You added in response to the reviewer comments at the end of this section that the litter stoichiometry is flexible. However, it would make more sense to state this directly in page 16 line 11, because the less diligent reader will likely miss this. Please also clarify what happens to the other nutrients if nutrient i is deficient - does this then imply that all other nutrients would be taken up less as well, or do you then break the stoichiometry?

Section 4.4 should not be labelled “Vegetation interactions”, because the model does not simulate any interactive vegetation, since Np is prescribed, and so is root respiration. In terms of nutrients, because the vegetation does not actually influence litter production or soil moisture or soil C interactively, the vegetation simply acts as a buffer mechanism. I would suggests “Effects of vegetation” or something along the lines.

I agree with reviewer #1 that the inclusion of a comparison to Manaus soil data is odd (and not helpful - see reviewer #2’s comment). I appreciate that there is a need to evaluate the soil C profile and that the Uni of Leeds has good data from Manaus, but I cannot see the relevance of a unique soil profile in the middle of the Amazon for modelling the chronosequence of soil C in Hawaii - this is simply too random a site selection to be useful. I would recommend to remove Figure 7, and shorten the text referring to it to state that the soil profile agrees in general of what you would expect from old tropical soils (refer to the Manaus data here), but in the absence of good data from Hawaii not much more can be said. I would also caution a too firm statement regarding the need for a depth dependent k, because your model might have other biases (missing/wrong bioturbation, depth-distribution of litter input), which could contribute to the misfit.

I don’t see the need for Figure 9 and the associated sensitivity study, specifically as no data are presented to evaluate the profiles. Aren’t these results trivial (and thus don’t really need a figure to illustrate? What do we learn from this that we need to interpret the results in Section 5? Given that the model does not assume any vegetation feedback and the Np is prescribed, the soil CO2 efflux in equilibrium must by definition independent of the CO2 profile.

In Section 6: “importance of vegetation”. As far as I understand it, this is your main conclusion and novel finding. It merits somewhat better explanation (what is it that matters and why). To my understanding, the major effect resides in the dependence of soil pH on root respiration? Somewhere then there needs to be discussion as to whether it’s realistic to assume that changes in pH only occur due to biological activity, or whether the low pH of precipitation or acidity resulting from weathering processes should not also play a role.

Figure 3-4: The text does not make any reference to the 10k results, and they are also not strikingly different from the 20k. I would recommend to merge Figure 3 and 4 so as to represent only the 20k results. This also avoid plotting the pH results from 20k twice.

Figure 10: I’m surprised that your soil CO2 efflux is only 2.5 kg. Assuming steady-state and a constant Np of 1 kg C I yield 1/12*44 = 3.6 kg CO2 efflux. What am I missing?

Answers to reviewer comments:

Your answers to reviewer #1 1-5 comments should be reflected in the text as a discussion of the current model caveats. You can rightly say that you will be addressing this in a future version, but I think that it is still important to acknowledge that the current version is very limited in the way “vegetation” is represented.

Your answers to reviewer #2’s comment on Hawaii as an erosional land scape should likewise be included in the revised manuscript.

page 22 line 11: check Figure reference
page 24 line 8: change “interact” to “affects” or something a long the lines. There is no effect of soil processes on vegetation growth or water-use.
Table 1&2,4: Change wt% to something generally understandable (mass fraction?)