This manuscript addresses an important topic and presents a potentially valuable dataset combining seismo-acoustic mapping, vibrocores, radiocarbon dating, sediment characterization, and organic carbon measurements from a Mediterranean rhodolith deposit in the Menorca Channel. The study has clear potential, particularly because millennial-scale sedimentary development of rhodolith systems remains poorly constrained, and because carbon storage in such calcifying habitats is still under-documented. However, in its present form, the manuscript overstates several conclusions relative to the data presented. The central problem is a mismatch between the scale of the claims and the scale of the measurements. The paper repeatedly frames the system as a millennial-scale organic carbon reservoir, yet organic carbon was only measured in the upper 50 cm of sediment, while chronology is available for only four of seven cores and includes reworked material in at least some basal horizons. As a result, several conclusions about long-term preservation, sequestration, and carbon-reservoir function are currently more interpretive than demonstrative. A second major issue is conceptual framing. The manuscript uses blue-carbon language for a calcifying system, but only later acknowledges the unresolved balance between sedimentary organic carbon storage and calcification-related CO₂ release. That caveat is fundamental, not secondary, and needs to be built into the framing from the introduction onward. At present, the manuscript reads as more certain about climate-mitigation significance than the evidence supports. In addition, the manuscript would benefit from more rigorous treatment of representativeness, uncertainty, and inference. The geophysical mapping demonstrates strong spatial heterogeneity in deposit thickness, yet the stock estimates are based on a small number of targeted cores collected from thicker deposits. That does not invalidate the study, but it does mean that statements about the broader Menorca Channel rhodolith bed need to be more tightly bounded.

For these reasons, I think the manuscript has merit, but it requires substantial revision before it is ready for publication.

Main comments

Claims exceed direct evidence

The manuscript’s strongest direct results are: highly heterogeneous Holocene biogenic deposits, plausible onset of persistent dense rhodolith–maërl facies after sea-level stabilization, and measured organic carbon stocks in the upper 50 cm. Those are worthwhile findings. But several broader claims—especially around millennial-scale carbon preservation and long-term reservoir function—go beyond what was directly measured. The authors need to distinguish much more carefully between what is measured, what is inferred, and what remains hypothetical.

Blue-carbon framing is conceptually incomplete

The introduction foregrounds organic carbon burial in coastal ecosystems, then extends that framing to rhodolith beds, but the manuscript does not adequately distinguish carbon storage from net carbon sequestration in a calcifying carbonate-producing habitat. This issue is acknowledged only near the end of the discussion. That is too late. The manuscript should explicitly state early on that the study quantifies sedimentary organic carbon stocks and depositional history, but does not resolve net ecosystem carbon balance.

Representativeness and extrapolation need stronger caution

Seven cores were retained from twelve collected, and core sites were selected on the basis of maximum sediment thickness. This is a sensible strategy for reconstructing deposit history, but it also biases the dataset toward thicker accumulations. That sampling design needs to be discussed explicitly when presenting carbon stock estimates and any implication for the wider bed.

Chronostratigraphic interpretation is plausible but needs more restraint

The manuscript uses reworked basal dates to interpret pre-Holocene material and then infers onset and consolidation of rhodolith-rich facies during the early to mid-Holocene. This is reasonable, but the text sometimes presents the sequence rather too smoothly, given the limited number of dated horizons and evidence of reworking. The depositional history should be framed as a best-supported reconstruction rather than as a tightly resolved chronology.

The manuscript needs major language revision

There are several grammatical, syntactic, and stylistic problems that they materially affect readability and precision. Several sentences are difficult to understand or ungrammatical, and some interpretive passages become less convincing because of this. The paper would benefit from careful editing by a fluent English speaker before resubmission.

Specific comments

Abstract / lines 24–27

The abstract currently presents the results as showing that Mediterranean rhodolith beds “act as long-term organic carbon stores.” This should be toned down. The study demonstrates organic carbon stocks in the upper 50 cm within a Holocene depositional framework; it does not directly measure long-term preservation throughout the full deposit.

Introduction / line 52

There are so many papers that give definitions of rhodoliths, so no need to use this in-review citation (Cabrito et al under review) from one of the co-authors. Check articles by Basso, Bosence, etc.

Introduction / lines 60–67

This paragraph needs a clearer conceptual distinction between organic carbon storage in sediment and net climate mitigation in calcifying systems. Without that distinction, the framing is incomplete. For organic carbon, it is also important to refer to floridean starch commonly produced by many coralline algae, which might play an important role in the blue carbon realm (see Pyko et al 2025 in Aquatic Conservation: Marine and Freshwater Ecosystems, doi:10.1002/aqc.70045)

Introduction / lines 93–98

The sentence ending “reconstruct the depositional history through in relation with the last glacio-eustatic oscillation” is ungrammatical and should be rewritten. More importantly, this section should state explicitly that the OC measurements are limited to the upper 50 cm.

Methods / lines 136–139

The manuscript should report uncertainty or resolution limits for the thickness estimates derived from the acoustic interpretation. This is especially important because the manuscript emphasizes fine-scale patchiness.

Methods / lines 141–146

Why were 5 of the 12 collected cores excluded? This should be stated clearly. At present, the omission is insufficiently explained.

Methods / lines 165–169

Please clarify whether age reversals or other stratigraphic inconsistencies were evaluated formally, particularly given the presence of reworked older material.

Methods / lines 178–196

The OC methods need more detail. For highly carbonate-rich sediments, the acidification protocol, mass correction, and QA/QC steps matter. The present description is too brief for full reproducibility.

Results / lines 222–230

The mapping result is important, but the manuscript should explicitly connect this heterogeneity to the limits of stock extrapolation. The current text highlights heterogeneity descriptively but does not fully integrate its implications.

Results / lines 253–254

The figure is extremely difficult to read.

Results / lines 259–277

The age interpretation would be clearer if the authors separated reworked/pre-Holocene ages from depositional ages more explicitly in both the text and Table 2.

Results / lines 288–293

The derivation of annual carbon accumulation from the 50 cm stock and median accretion rate needs to be shown more transparently. At present, this estimate appears precise, but the underlying assumptions are not fully explained.

Results / lines 306–315

The mixed-effects model is potentially useful, but the interpretation is stronger than warranted by the limited sample size. With seven cores and upper 50 cm subsamples only, the modelling results should be presented more cautiously.

Discussion / lines 337–338

“when sea-level lowered down its rise almost reaching its actual level” should be rewritten entirely.

Discussion / lines 362–369

The interpretation of no depth trend as evidence for mixing, burial, and textural control is plausible, but still speculative. The authors should label it as such.

Discussion / lines 377–384

The comparison to Scottish rhodolith beds is potentially useful, but it needs more caveats. Differences in depth interval, age structure, hydrodynamics, sediment composition, and analytical design make direct comparison difficult.

Discussion / lines 398–401

This is the conceptual caveat the manuscript most needs, yet it appears only near the end. It should be moved earlier into the framing and discussion.

Conclusions / lines 403–413

The conclusions are currently overstated. “Actively storing carbon over millennia” and broader references to global stability are too expansive relative to the evidence presented here. The authors should conclude more narrowly: this deposit preserves measurable organic carbon in the upper sediment and sits within a dated Holocene depositional framework consistent with long-term accumulation.

This is an interesting paper on the role of rhodolith beds as short-term carbon storage in the Menorca channel. Authors combine different methodological approaches to present robust results. Furthermore, the results strengthen the general view that rhodolith beds are vulnerable ecosystems that should be protected due to their relevance as ecological services, biodiversification centers, and because they contribute to regulate global climate due to carbon storage.

Despite the results and conclusions are robust, the paper could be benefited by broadening the discussion section. Several parts of the discussion are actually repetitions of the results (i.e., lines: 320-326; 331-338; 354-359;370-376). I think that the discussion could be enlarged, including and discussing the information provided by other authors. In particular, regarding the carbon storage, which is the main aim of the paper, authors only consider the paper by Mao et al. (2020). They should also take into consideration the papers by van der Heijden and Kamenos, 2015; Macreadie et al., 2019; Mao et al., 2024 [Biol. Lett. 20: 20230598 (doi: 10.1098/rsbl.2023.059]; James et al., 2024 [Glob. Chang. Biol., 30, e17261 (doi: 10.1111/gcb.17261]; Schubert et al., 2024) and references therein. In addition, authors can check Bosence and Wilson (2003; Aquat. Conserv., 13, S21-S31) and Ehrhold et al. (2021; Palaeo3, 577, 110525, doi: 10.1016/j.palaeo.2021.110525) for further comparison of thick buried rhodolith beds.

Apart from this issue, there are a few minor suggestions/comments:

            Line 26. I suggest to delete “Mediterranean”

            Line 49. The reference Aguirre et al. should be 2017.

            Line 52. I think that it is not appropriate to refer to a paper under review. Delete the reference to Cabrito et al.

            Line 54. In addition to Basso et al., consider adding also Aguirre et al. (2012, 2017).

            Line 182. “24-48 hours”? Indicate this.

            Line 209. What is AR(1)?

            Line 212. What is AICc?

            Line 341. Consider adding Aguirre et al., 2017 to the references.

            Line 405. Delete one “with”.

            Figures. Letters, numbers and symbols in Figs. 2 and 3 are very small. I suggest to increase them.

                        Symbols in Fig. 4 can be depicted with different colors to highlight them.