General comments

This study investigates controls on the erosion and stability of pyrogenic across eroding hillslopes in the Brittany region of France with the aim of identifying controls on PyC accumulation across this landscape. By the use of combined methods for quantifying PyC across landscapes, they show increased PyC stocks at deeper soil depths and at some, but not all downslope landform positions.

The authors should consider that they have not measured rates of erosion or slope directly along the transect. This may confound some of their results, as slope can directly affect erosion rates. This manuscript could be improved by including some discussion and/or analysis of data that includes the slope of their landform positions.

Specific comments

Line 81-2. This study (Abney et al 2017) does not give full evidence that the PyC was further eroded, although this is a possibility. It also may have been buried by subsequent erosion events, leached down the soil profile, and/or decomposed.

Paragraph starting line 93. This paragraph is a little bit hard to follow. It would be improved by starting with the overall aims of the study, then the hypotheses and methodological approaches.

Section 2.2. This section would benefit from some more details – are the same number of samples taken from each transect? How many samples total and how many per transect?

For hypothesis 2b, I think this could be more specific, as we would hypothesize that all soil carbon at depth should be older than surface carbon. Perhaps the authors would hypothesize that PyC at depth is older than bulk soil C at depth (which is supported later in results)?

Figure 3 – This addresses hypothesis 2a, but perhaps it would be illustrative to also conduct this analysis with slope. As some upslope landform positions (i.e. far from the toeslope) could be relatively flat and have low erosion rates, or even be depositional environments along the slope where one might hypothesize PyC would remain in place or accumulate compared with more erosion prone (i.e., steep) slopes.

Line 434. It would be better to say that this hypothesis is refuted for this specific site. I don’t think there is sufficient statistical power (n=3 transects) to make this broad statement even for this general region, as there are many soil classifications along the transects.

Technical comments

Line 34-5, Bird et al 2015 review supports this statement. Reference: Bird, M. I., Wynn, J. G., Saiz, G., Wurster, C. M., & McBeath, A. (2015). The pyrogenic carbon cycle. Annual Review of Earth and Planetary Sciences, 43(1), 273-298.

Line 111 – There are several abbreviations here that are not spelled out previously (ORE AgrHys and OZCAR).

Line 112 – “Slopes are at most 5%.” What is the range of slope, or average? This description could benefit from more details, if available.

Line 114 – It would help to identify the time frame of the last glacial period.

Figure 1. The transects in inset A should be labelled in either the legend or caption for this figure.  

Line 127. It is not clear, but I assume the last 10,000 years is the most recent unglaciated period? I recommend editing for clarity.

GENERAL COMMENTS

This is a very thorough and interesting study that assess PyC distribution through the soil profile and in different topographic situations. The literature review in the Introduction is very detailed and tprovides a very good picture of the current state of the studied topic. Interestingly, it uses two different methods for PyC quantification (Hypy and CTO375) and determines PyC and SOC age, what is a great and valuable addition. The discussion also does a good job setting the study results in the wider literature context.

The main conclusion that slope position does not control PyC stocks and age needs to be sent with caution. The studied soils have a very gentle slope (less than 5%), this obviously limits the role of soil erosion (lateral movement across the landscape) hugely and makes this system very different to many landscapes where PyC production is high (any fire-prone ecosystem in mountainous regions), this is already metioned in the discussion but should be better recognized.

Another important point is that PyC contributions in the last 150yrs are assumed negligible due to the no presence of wildfires but, what about atmospheric deposition? Especially if these inputs would come from fossil fuel combustion they would not give a young PyC and would be detected by the CTO method (soot like PyC). 

SPECIFIC COMMENTS

• Abstract: please include what soil depths the calculated stocks go to. Also, it would be interesting to include even if briefly how CTO and HyPy methods correlate (or not) when quantifying PyC in this study.
• Introduction: in this section, lateral erosion is referred to as “lateral movement”, “erosion” and even horizontal displacement (L67) or downhill redistribution, this can be confusing for readers who do not much about this process. Please define, pick a term and use it across the ms.
• Introduction: not clear how can PyC distribution be affected by the gradient of soil evolution along the slope (L99), please expand a bit on this.
• Methods: please explain a bit the land-use changes that are assumed to have happened in the study area over the last millenia (do not only refer to the Appendix). Also, the justification of “the last 10,000 yrs” (L127) needs to be expanded.
• Methods: please explain which samples and how where they selected by the HypY analyses.
• Discussion: please elaborate on the effect of tillage (L355) as it can substantially modify the soil profile. It may be worth mentioning it in the abstract too.
• Discussion: not super clear the differences between hypotheses 2 and 3 in lines 370-375, please describe.
• Discussion: in the paragraph starting at L475, please explain a bit more in detail the differences between the PyC detected by the two methods and how this can contribute to the differences seen between them.
• Discussion: section 4.3.2. is too speculative, stronger proof of “recent” erosion should be provided, are there any measured erosion data available?
• Discussion: please give some numerical data to support the estatement about the comparison with the wooded area (L524).

TECHNICAL CORRECTIONS

• L2: please rephrase, not all PyC is condensed
• L3: no clear whether the variation depends to the method used for determining residence time or PyC.
• L22: give age of PyChypy at 0-10cm.
• L47: not clear how those “losses” relate to the previous info in the paragraph, please rephrase.
• L56: favourable means preferential? Please clarify
• L62: altered bedrock or/and groundwaters, right?
• L89: previous studies are not all in the US, please rephrase.
• Please avoid paragraphs starting with “Figure X…”.
• L319: weaker than?
• L520: please rephrase, “disaggregation of soil during erosion” is not a “erosion modality”