General comments

The authors examine an important topic: what is the size and age and variability of soil organic carbon (SOC) in deeper portions of the soil profile (30 -100 cm in depth e.g. subsoil), and how do these quantities vary across space?  The authors description of SOC and soil microbe composition changes along the soil profile will inform ongoing scientific discussions of soils and their ability to store carbon (C). This question and the findings of this study are also relevant for the growing number of projects (public and private) hoping to mitigate climate change through increasing SOC pools via soil amendments or ecological restoration.

The methods used by the authors are appropriate for investigating this topic. I am also pleased by their use of multiple estimates of SOC age - which gives confidence to their findings. I found the manuscript to be well written, maintaining a healthy balance of thoroughness and interest throughout the text.

Response: We thank Reviewer 1 for providing their time and expertise to review our manuscript and provide constructive feedback. We found the comments helpful and presented in a kind and balanced manner. We believe that taking on the recommendations you have made will improve the quality of our manuscript. We have responded to Reviewer 1’s specific comments below. For most comments, we are happy to revise the manuscript according to the reviewer’s recommendation.

Specific comments

To understand and generalize from these dynamics, I (and I assume readers) would be interested to know the average and range of pH in the soils of this planted forest. I would add this information to the description of this site (which is otherwise fairly comprehensive).

Response: Thank you for taking interest in the finer details of our field site. Puruki Forest has an average soil pH of 5.2 (Beets & Brownlie, 1984). Further research by Beets et al. (2004) identified the soil pH of Puruki Forest to increase with soil depth, reaching pH 5.62 at soil depths of 2 m. We are happy to add this information to the manuscript under the reviewer’s recommendation.

The authors need to pay attention to the notation that they use and keep this consistent throughout the text and figures. At times they switch between percent and permille for 14C (Fig. 2). Figure axis titles should also be the appropriate symbol, and not %o for permille.

Response: Apologies for these errors, we are happy to correct these in a revised manuscript as per the reviewer's recommendation.

CRA is not defined in the text. Please do so before introducing this as a measurement.

Response: We apologize for not defining CRA properly. We are happy to add the following sentence to the revised manuscript to clarify our terminology- “Conventional Radiocarbon Age (CRA) and ¹⁴C are reported as defined by Stuiver and Polach (1977), with CRA presented as years before present (yBP) and ¹⁴C presented as per thousand [per mille, ‰].”

My understanding of Fig. 5 is that C and D are replots of A and B, but just now with the environmental vectors overlaid. These replots with the environmental vectors don't allow me to interpret the shifts in your data points (the data is too scrunched near 0,0). I suggest either removing C and D or re-scaling the vectors (divided by 10 maybe) and replot A and B with these rescaled vectors so that readers can see how these environmental variables are affecting your estimated microbial compositions.

Response: Reviewer 1 is correct; subplots C and D are re-plots of A and B with fitted vectors. We agree that these plots are not easy to visually interpret- which makes their purpose potentially redundant. The information visualized in subplots C and D was detailed in-text in Results section 3.4 (lines 251 to 257). Considering this, we are happy to remove subplots C and D from Figure 5 in the revised manuscript.

6 has some bizarre mispellings of names (Sebciunkles instead of Sebacinales) and I would standardize the names to remove the trailing '_ unk' artifacts of taxonomic clustering

Response: Such errors are well noticed by Reviewer 1, apologies for the misspellings. We fully support the recommendation to standardize the names and remove ‘_unk’ from Figure 6 in a revised manuscript.

Line 175 - McMurdie not Mcmurdie

Response: We are happy to correct this error in a revised manuscript.

It's not clear to me why you exclude negative interactions from the network analysis. If you can justify this briefly, do so in the text.

Response: we excluded negative interactions as we were primarily interested in observing positive species co-occurrences. That is, we focused on which species are commonly present and defined the ‘core microbiome’ of the topsoil and subsoil. However, we understand including negative interactions is an insightful piece of information to present. If deemed appropriate, we are happy to re-run the analyses to include negative interactions in a revised manuscript.

Line 414/415- you didn't examine microbial biomass though, you quantified DNA and you've stated that you did not partition between viable and nonviable cells/hyphae. While DNA abundances can be used as a proxy for biomass in controlled systems of relatively short age I don't agree with claiming this as biomass here, as much of the DNA you sampled from these lower depths may actually be relictual. Best refer to it as something neutral like 'DNA abundance'

Response: Reviewer 1 provides a valid argument that we support. We are happy to rephrase the term microbial biomass to microbial DNA abundance in the revised manuscript (or words to the same effect if this does not grammatically fit the sentence in question i.e., microbial 16S rRNA gene abundance).

I found it interesting that Bray P did not correlate with other factors (Fig. 4). In the text you mention that you also measured total soil P, however it looks like this wasn't included in analyses. Was this estimate just not variable? I noticed that the soils are very young (from a recent volcanic eruption even), so I'm assuming that the microbes and vegetation are more N limited while P is abundantly available? If it's not too much trouble I'd add general P and N abundance or availability at this site in the site description (plant-microbe people love this). Total P should atleast show up in the supplemental materials along with total C and N(e.g. Table A2)

Response: We measured many different soil properties. As the primary focus of our research was carbon, to keep our analyses focused we selected a few non-carbon metrics. However we agree with Reviewer 2, information relating to soil phosphorus is interesting regardless of whether we included it in our analysis. Thus, we are happy to revise Table A2 to add the data for total P, as well as inorganic P and organic P. Or as an alternative, we would also be interested in submitting a Data in Brief article to accompany this manuscript to provide the full datasets for all the soil chemical properties we measured. This may be more valuable to the wider scientific community. Whatever the Reviewer or Editor feels is more appropriate and useful.

As background, Puruki Forest was formerly pasture that was converted to Pinus radiata forest in 1973 (Beets & Brownlie et al., 1984; Meder et al., 2007). During the establishment of Puruki into pasture in 1957, superphosphate treatments were applied and therefore the site has a high soil fertility including soil N and P – see Beets & Brownlie (1984).

• Meder, R. O. G. E. R., Beets, P. N., & Oliver, G. R. (2007). Multivariate analysis of IR, NIR, and NMR spectra of soil samples from different land use conversions: Native Forest, pasture, and plantation forest. New Zealand Journal of Forestry Science, 37(2), 289.
• Beets, P. N., & Brownlie, R. K. (1987). Puruki experimental catchment: site, climate, forest management, and research. New Zealand Journal of Forestry Science, 17(2/3), 137-160.

This study explored the changes in SOC dynamics and soil microbes in the Puluki Experimental Forest (New Zealand) down to a soil depth of 1 m. ITS and 16S rRNA sequencing and quantitative real-time PCR were used to measure changes in soil microbial diversity, composition, and abundance. Stable (δ13C) and radioactive (14C) C analyses were performed to assess depth-driven changes in SOC stability and age. It has to be said that it is very important to use these methods to explore the dynamic changes of deep SOC, but it is not the most innovative. The sentences are smooth and comfortable to read. To my surprise, the total C stocks of deep soil accounted for only 35%, which was not more than 50% as previously reported. The Introduction is like science fiction, lacking clear questions and hypotheses. Moreover, few links were made between changes in soil microbes and soil carbon stocks in the Discussion. Other suggestions are as follows:

Response: As with Reviewer 1, we would like to thank Reviewer 2 for contributing their time and expertise to review our manuscript. We value your comments on our work and will take them on board when revising the manuscript. Regarding the comment that “The Introduction is like science fiction”. Nothing we have written is fictional and is entirely based on current scientific knowledge.

Reviewer 2 was surprised to read that 35% of total C stocks were allocated to the subsoil considering we previously reported the value of 50%. This 50% value was mentioned in the Introduction and was a reference to a different study (lines 36 to 39). At no point did we suggest this was a finding of our own. Furthermore, deep C allocation will vary depending on a range of factors such as soil type, soil age, vegetation, climate, and management history. Thus, deviations in our findings from previous research are expected and build on the limited knowledge available in the scientific literature.

Additionally, Reviewer 2 commented that few links were made between changes in soil microbes and soil carbon stocks in our Discussion. We agree; our balanced our discussion against the methods we undertook and took care in not drawing too many direct conclusions on soil-microbial relations as we examined microbial DNA only. No need to present the readers with fiction. Whilst DNA-based NGS targeting SSU rRNA genes provide a wealth of valuable information, they provide limited information on the functional role or activities of soil microorganisms.

Title: Microbial communities are not suitable to be described as big and old.

Response: “Large, old pools…” was in reference to soil carbon, not microbial communities.

L16: Why must it be an incremental change?

Response: we used the term ‘incremental change’ as we analyzed changes in SOC and the microbiome of the soil cores at 10cm increments down the soil profile. Many previous studies have analyzed depth-driven changes in soil carbon at broader horizons i.e. 0 to 10, 10 to 50, and 50 to 100cm. For our study, divided soil core into smaller increments to identify at which point large shifts in SOC and the microbiome take place.

L22: Does soil carbon refer to carbon storage or carbon concentration or stability or others?

Response: Upon review of our manuscript, we understand how this sentence is confusing to the reader. We were referring to soil carbon concentration (or calculated soil carbon stock). We are happy to clarify this ambiguity in the revised manuscript. Thank you to Reviewer 2 for identifying this.

L23-L25: This study is only a sample study, why do you say “These research findings highlight the importance of quantifying subsoil C stocks for accurate systems-level global and local C budgets and modeling”? Moreover, this study does not address climate change.

Response: We added this sentence in to address the wider implications of our research findings- that subsoil C can contribute to total forest soil C stocks. Consequently, accurate quantification of subsoil C should be considered more by the wider research community. However, we understand including this sentence in the Abstract may be misleading as our study did not directly research climate change. This sentence may be better suited to the Discussion/Conclusion where we can fully expand upon its intent and meaning. Thus, we are happy to remove this sentence from the Abstract of the revised manuscript and add it to a more appropriate section in the Discussion and ask the Editor for clarification as to if this is necessary. Thank you.

L28-31: These three sentences all emphasize the importance of forest soil carbon, which can be simplified and combined with the next paragraph.

Response: Thank you to Reviewer 2 for identifying this. Whilst lines 28 to 31 provide valuable information, we agree the 3 sentences are unnecessarily repetitive. We are happy to make this section of the manuscript more concise for the reader.

L41-42: What is the meaning of “having a physical and chemical nature”?

Response: the physical and chemical properties of the soil environment i.e. reactive mineral surfaces, soil redox state, and access by soil microorganisms. This will be clear to most readers.

L44-46: Although these are all factors affecting the stability of SOC, they are not addressed in this study, and the preamble should introduce more advances in microbes, isotopes, etc.

Although we appreciate Reviewer 2’s comment, we think this is an important section of information to include in the introduction of the manuscript. Providing the reader with information on environmental factors governing SOC sequestration is relevant when we are quantifying subsoil carbon stocks. Furthermore, the paragraph following this section (lines 50 to 62) details the importance of studying soil microbes. Lines 67 to 72 outline why we performed isotopic measurements.

L46: delete “, thereby,”

Response: this will be deleted in the revised manuscript

L51: delete “fundamental”

Response: this will be deleted in the revised manuscript

L64: How to understand the meaning of the word " fundamental"?

Response: In this context, we were referring to baseline or foundational. We understand how this term may be confusing and are happy to remove or replace fundamental with a more suitable term (or just remove it from the sentence).

L65: How to understand “at a highly refined spatial scale”?

Response: Here we were referring to the fact that we measured depth-driven changes at 10cm increments. We can see how “at a highly refined spatial scale” is non-specific and not helpful for the reader's understanding. We are happy to rephrase the sentence to “we aimed to examine the depth-driven variability in soil C dynamics in 10cm increments down a soil profile” (or words to such effect).

L87: delete the second “(”.

Response: this will be deleted in the revised manuscript

L100: To avoid DNA degradation, soil samples are usually stored at -20 or -80 â before DNA determination. How long will this study complete DNA determination after sampling?

Response: Soils were sampled from Puruki Forest located in the North Island of New Zealand (NZ). After sampling soils were immediately stored at 4^oC for a brief period (~1 week). Soil DNA extractions were performed in a laboratory in Christchurch (South Island of NZ). Consequently, after sampling soils and storage at 4°C, soils required for DNA extraction were transported down to the South Island based laboratory. Once at the laboratory soils were then placed in the –20 freezer prior to DNA extraction. Apologies, this information should have been added to the manuscript. It will be included in the revised manuscript as “Once transported to the laboratory, soils required for DNA extraction were stored at –20°C".

L103: What are the “Mehlich 3 extractable elements”? How are fractions equal to 2 mm treated?

Response: The Mehlich 3 extractable elements were Potassium (K), Calcium (Ca), Magnesium (Mg), and Sodium (Na). The second question is not clear. Is Reviewer 2 asking how soil fractions that were exactly 2 mm in size were treated? If so, I cannot answer this question as to the best of my knowledge such soils were not found. Soils that passed through a 2 mm sieve were regarded as < 2 mm soils, and those which did not were classified as > 2 mm. My apologies if I have misinterpreted the question. We have detailed in Section 2.2 exactly what forms of analyses were tested for each soil fraction.

L150: How to calculate soil carbon storage, it is recommended to list a formula.

Response: Soil carbon stock was calculated using the formula ‘SOC_stock = BD x SOC_conc x D’ (Gattinger et al., 2012; Jones et al., 2008). Here BD is bulk density (g/cm³), SOC_conc is Total C %, and D is thickness of the soil layer (cm). Following this, soil carbon stocks were slope corrected by multiplying SOC_stock by the slope ratio. We are happy to add this information to the revised manuscript and thank Reviewer 2 for the suggestion.

• Gattinger, A., Muller, A., Haeni, M., Skinner, C., Fliessbach, A., Buchmann, N., Mäder, P., Stolze, M., Smith, P., Scialabba, N.E.H. and Niggli, U., 2012. Enhanced topsoil carbon stocks under organic farming. Proceedings of the National Academy of Sciences, 109(44), pp.18226-18231.
• Jones, H. S., Garrett, L. G., Beets, P. N., Kimberley, M. O., & Oliver, G. R. (2008). Impacts of harvest residue management on soil carbon stocks in a plantation forest. Soil Science Society of America Journal, 72(6), 1621-1627.

L153: What are the “Bray P sequential 1”? What does slope corrected mean?

Response: Bray P was measured using the Bray 2 extraction method. Three sequential extractions were performed (sequential extraction 1, 2, and 3) to measure the amount of P extracted from the soil. We used the value of the 1^st sequential extraction in our data analysis. We will revise the manuscript and instead write “Bray P (sequential extraction 1)” to be clearer.

‘Slope corrected’ refers to the slope factor applied to values for total C & N stocks, as well as Bray P1 and exchangeable cations. Our data was slope corrected because the sampling transect from which our soils were sampled was along a 12^o slope. This detail should have been included in the Methods section; it will be added to the revised version of the manuscript. Apologies for this omission.

L209: It is recommended to merge 3.1 and 3.2.

Response: Thank you for the suggestion, we are happy to merge these two sections in the revised manuscript.

L213: Is the result reliable if the variation of SOC stocks in the coarse soil fraction is so large? Some similar results can be described together to avoid redundancy. For example, the results of most indicators decrease with the soil layer. Where is the Table?

The variability in SOC stocks between soil cores is consistent with previous research also conducted at Puruki Forest. This is something we discussed in detail in the Discussion, lines 306 to 318. Apologies, the Table referred two in line 215 needs to be corrected to Table A3 (note: an appendix, not located in the main text). Any incorrect table numbers referred to in-text will be corrected in the revised manuscript.

L220: delete “this trend was variable. In particular,”.

Response: this will be deleted in the revised manuscript

L225: It is suggested to delete some results in 3.4 and 3.5. Not all results need to be described, but they should be targeted.

Thank you for the suggestion, we agree these too sections are detailed unnecessarily, and we are happy to make these sections more concise in the revised manuscript or create a supplementary results file for secondary findings.

L251-257: Generally, |R|<0.3 is considered a weak correlation, where R is generally lower than 0.3. I doubt the reliability of the results

Response: The reliability of the correction was tested using a formal statistical test. The results should be interpreted as ‘there is a statistical association between the response of the two variables, however the association strength is moderate (or weak, depending on your view).  i.e., there is an association between the variables, the ‘explain’ some the variation in each other’s response, but a lot remains unexplained.   We agree that an element of subjectivity comes into if particular R values are weak, moderate, strong etc, but this in no way affect the reliability of calculating that particular R value. 

L264: Microbial classification names need to be italicized.

Response: We followed the convention of not italicising classifications above genus level which is why the names in this section are not in italics.

L333: neatly to? Or nearly to.

Response: here neatly referred to corresponded well or closely. We are happy to change this to a less ambiguous term in the revised manuscript.

L350: What does "quantity" mean here?

Response: the soil carbon concentration (Total C %). We will reword this sentence in the revised manuscript to “…correlated with the concentration (Total C %) and radiocarbon age…”

L365: How to calculate the microbial density?

Response: Here we were referring to microbial biomass, or DNA abundance. Microbial DNA abundance (16S and ITS rRNA gene abundance) were outlined in the Methods section 2.5. We will reword this in the revised manuscript as ‘density’ may be confusing and is inconsistent with the terminology we have otherwise used.

L367-368: The abundance of bacteria significantly declined with soil depth. Is it contradictory to say that the relatively large community abundance?

Response: Reviewer 2 is correct; bacteria did decline with depth. However, what we were stating in this sentence is that bacterial communities had larger abundances relative to fungal communities- which is true. Our results show the diversity and taxonomic coverage of subsoil bacterial communities was greater than that of fungal communities. However, we understand the point Reviewer 2 to making here, so we are happy to remove or reword this sentence to be less suggestive.

L408: How to understand that the abundances of Myxomycetes, Thelephorales, and Tremells were related to soil C quantity and age?

Response: The methods for this were outlined in lines 192 to 194- “log adjusted abundances of microbial taxa obtained from ANCOM-BC analysis were correlated to soil chemical properties using pairwise Spearman’s rank correlation tests…”. The results of these tests were presented in the Results section (lines 284 to 288). NOTE: the manuscript refers the reader to Table A15 for more detailed supplementary results however this is incorrect, it will be updated to Table A14 in the revised manuscript.

The red line in Figure 1 is not obvious, so it is recommended to consider other colors.

Response: We thank Reviewer 2 for the suggestion. We are happy to change it to a more contrasting colour in the revised manuscript.