See attached. 

Queiroz Alves and colleagues present a record of organic biomarkers and carbon isotopes from a sediment core from the Bay of Biscay, which provides observational evidence of post-glacial release of terrestrial carbon. Using source-fingerprinting of different biomarker indices and a dual-carbon isotope-based carbon mixing model, the authors deduce carbon remobilization from ancient and frozen peat during deglacial climate warming. Because permafrost/peat thaw processes may have released a larger amount of organic carbon at the time, the authors suggest that in-situ remineralization of the liberated organic matter may have contributed to the onset of the rise of atmospheric CO₂ concentrations around 17.5 kyr ago as seen in Antarctic ice cores.

The paper is overall well-written and structured logically. The topic is timely and important as climate change scientists seek to better understand the functioning and potential climate impact permafrost thaw and carbon release may have. The methods used in this study are sound and I am convinced that the work was carried out in a careful and rigorous manner. However, I see some weaknesses in the writing of the paper and its data interpretation, which I believe can be addressed during major revisions.

Overarching comments

Methods: Some of the methods are insufficiently explained in the main manuscript, and I would like the authors to include more information about the functioning and limitations of the different proxies used in their work. For instance, a big part of their story rests on a ¹³C/¹⁴C mixing model, which involves quite large uncertainties and assumptions and thus needs to be explained in detail in the main manuscript. Also, it is not really clear from the methods which data were generated in the present work and which originate from earlier work.

Age model: This is a central part of this study, and much of the results and interpretation rest on this. From my understanding, the age model of the GeoB23302-2 is not published elsewhere, so I would ask the authors to provide more detail about the ¹⁴C measurements of the planktic foraminifera, including a data table of conventional ¹⁴C ages and resulting calibrated ages. I also have concerns about the choice of the Marine20 calibration curve and the potential impact of an additional marine reservoir effect in planktonic foraminifera. Particularly during the last glacial, there may have been a large difference in terms of ventilation between the ocean surface and the interior. Hence, I wonder if the reservoir is not over-estimated when applied to planktonic ¹⁴C ages. I would like the authors to expand on this topic in more detail and include a concise statement in the methods section of the paper.

Carbon sources: There is much talk about petrogenic carbon throughout the paper, although the authors do not seem to find petrogenic carbon (but ancient peat) to be an important source to the sediment C. If petrogenic carbon is not important, I wonder why a petrogenic (¹⁴C-free) carbon source is even quantified here (your dual isotope model suggests petrogenic carbon contributions up to 60%), which contradicts your discussion and conclusion. I also wonder about the use of D¹⁴C values of n-alkanoic acids as a terr-OC end member, as they may follow different transport and aging dynamics than the bulk material. Furthermore, n-alkanoic acids are more degradation resistant than other organic compounds and may thus represent a pre-aged terrestrial carbon pool. Have you considered to distinguish between a contemporary (¹⁴C-young) terrestrial OM source (e.g., soil OM) and strongly pre-aged (¹⁴C-depleted, but not free of ¹⁴C) terrestrial sources that represents LGM (or pre-LGM) peat OM?

Minor points

Abstract

Line 9: What kind of processes? Carbon release, degradation and CO₂ emissions?

Line 12-13: It would be helpful to clarify add how thawing permafrost affects the global carbon cycle.

Line 17: Rising atmospheric levels?

Line 43-45: This sentence reads a bit awkward. What is the “strong response from this system” specifically?

Line 50: The continental shelf and what about the deeper ocean?

Methods

Line 57-60: Please explain briefly how these elemental ratios work; i.e. why Zr/Rb is thought to mirror grain size and Fe/Ca is a proxy for terrigenous sediments.

Line 68: Similar here; how does the CPI indicate OM degradation?

Line 69: Also the BIT and fββ indices need to be explained a bit more in the main manuscript.

Results

Line 80-81: Isn’t the age model based on foraminiferal ¹⁴C ages? What is the OxCal time series agreement index?

Line 81-83: Please add more information about these reconstructions. Clarify that the NGRIP and CO₂ reconstructed are based on ice cores, and explain what the sea surface temperature record based on.

Line 95: What are “relatively high contributions”? A dominating fraction?

Discussion

Overall: Most of the data interpretation rests on concentrations and relative contributions of terrestrial compounds and OC fractions. Have the authors considered to estimate fluxes of the terrestrial OC fractions and compounds (e.g., of alkanes or the terrigenous OM fraction in g per sqm per year), which may provide an additional perspective to the paleo-variability of terrestrial carbon release. If not, I encourage the authors to do so.

Line 104-106: Please explain how the Zr/Rb ratio or coarse-grained sediments are consistent or indicative of terrestrial material.

Line 107: How are these two periods different? Why not say 20.5-16.5 ka BP?

Line 111: Heinrich events are not introduced or explained.

Line 127: Isn’t the CPI by definition >1 or 1? It’s not surprising the CPI values remain above 1 throughout the core in such a near-coastal environment.

Line 144-146: A few lines above you provide a discussion about different processes (transport vs. petrogenic C) that may lead to high pre-depositional ages. Which of those processes/end members were included in the mixing model, and will the resulting C fractions not just be dependent on the end members chosen? Did you account for possible C aging during transport, which would probably affect the age of peat material?

Line 169: Do the authors mean that peat from older periods (e.g. the last interglacial) was frozen during the glacial? Perhaps add something like “OM from older periods (e.g., frozen peat OM)” to clarify this.

Line 195: Could post-glacial sea level rise and erosion of terrestrial deposits also contributed to thermo-erosion of frozen deposits?

Line 225-229: It may have contributed to the post-glacial carbon cycle perturbations but it could be good to mention that the loss of terrestrial carbon reservoirs was likely compensated by expansion of soil carbon stocks and vegetation growth at the time or thereafter (e.g., see Lindgren et al., 2018, https://www.nature.com/articles/s41586-018-0371-0).

Line 233-235: What kind of leads and lags do we expect during permafrost carbon feedback? I would think that permafrost thawing leads to CO₂ emissions rather instantaneously, while the lateral transport of permafrost organic matter to rivers and the ocean may cause a lag, such that the 'peak signal' is seen delayed in the offshore sediment archive. So this would actually point to the opposite; that the post-glacial release of old peat/permafrost organic matter in central Europe did not affect atmospheric CO₂. So again, how certain are the authors about the correctness of their age-depth model (see my main comment)?

Line 236: How do the authors explain the sudden drop of terrestrial C discharge (as shown by the BIT index, Paq and the age of n-alkanoic acids). Could the river re-routing be related to that? It is not very clear from reading this paragraph.

Line 254-255: Why is there so much weight on petrogenic C in the paper if it is not one of the main sources?

Line 261-263: This is a quite complicated way to say that terrestrial/permafrost carbon release likely continued also after 17 kyr, although this was not recorded in your core. I suggest to write this more clearly.

Line 263: In the light of the findings and the offset between the terrestrial carbon peak and the period of CO₂ rise I would not call this ‘likely’. I think that ‘it is possible’ reflects this and the underlying certainty better.

Line 267: And what are these important consequences? Please be more precise with these implications.

See review in attached PDF.