General Comments

The present study adds to the knowledge of CO₂ cycling in the critical zone. It uses Grotte de Milandre as case study to investigate the unsaturated zone. This is attained by an extensive monitoring of the pCO2, stable C isotope ratios and 14C02 from shallow soil (under meadows and forest), to epikarst to cave atmosphere upstream (topographically high) and downstream (topographically low).

There is a wealth of information in the present research manuscript that is commendable.  Yet, there are aspects of the systems that are not fully explored, which yield, in the end, to a conclusion that the deep root zone respiration is more important than we think.

I am no expert in rhizosphere, and in microbial processes. I am just a humble mineralogist. But, by exploring how crystals grow in both tropical and polar environments, I am aware that sources of CO₂ are multiple and complex. It is well known to those who work on Antarctic subglacial environments that microbes thrive whenever there is water. Temperature is not the critical factor, but the absence of liquid H₂O…on Earth.

In the dark cave environment, microbes and fungi thrive, and their respiration must influence the cave air pCO₂. Unique chemoautotrophic communities have been described for Kartchner cave, with potentially multiple CO₂ fixation pathways (Ortiz et al., 2014). Cave methanotrophic and heterotrophic bacteria are producers of bioactive compounds acting as rapid CO₂ sources and sinks (Martin-Pozas et al., 2020). Cave sediments appear to play an active role in carbon cycle (Martin-Pozas et al., 2022). And a role for cave sediments is not really discussed in the present research paper.  Yet, I do not know if there are sediments in Grotte de Milandre, as I could not find any references related to that specific topic. Perhaps there are no sediments. So, this should be stated somewhere in the text.

The importance of deep rhizosphere has been already pointed out in karst studies (Iversen, 2010; Chen, 2019; Wen et al., 2021).  In contrast, the importance of the cave autochthonous microbial assemblages remains poorly known.

In particular, cave sediments and speleothems (and soils above caves) may host methanotrophs, which may change the ¹³C values of both soil and cave atmosphere (Waring et al., 2017). As it has been demonstrated that caves host thriving microbial (including fungi) communities, the discussion should tackle this aspect of the cave dynamics. Not just ventilation, but in-situ microbial processes that occur in both oxidising and reducing “local” settings.

In the framework of understanding C cycling, it is important to advance our understanding of C release and sinks in the subterranean environment (Zheng et al., 2024). One of the C-cycle contributors in caves is methane. Thus, more of the discussion should be focused on ruling out methane…as this gas can also be evolved from ancient pore fluid waters trapped in the Oxfordian marls. Old methane can be ruled out by F ⁴C, but in-cave produced C from methane cannot.

An old isotope study on stable C and O isotope ratios in karst spring waters, where catchments were under forest and meadow soils in karst systems (Stichler et al., 1997) should be discussed to see how the Grotte de Milandre dataset compares with the 1997 one from Slovenia.

Overall, this reviewer believes that more insight should be given to the complexity of the rhizosphere by reporting more results from soil research. It should also provide more insight in the discussion about the contribution of cave microbial communities. If (as I would seem) no research has been carried out on fingerprinting the cave microbial communities as it was done elsewhere (see for example Tomczyk-Żak and Zielenkiewicz, 2017; Kosznik-Kwa´snicka et al., 2022; Lange-Enyedi et al., 2022; Lange-Enyedi et al., 2023; Gogoleva et al., 2024) then the “in-cave” contribution…which would be associated with the “ventilation”  process (as in-situ microbes are unaccounted for) would go…zilch.

Also, no insight is given about the possible contribution to the C cycling of fungi (see detailed comments).

 Finally, it is difficult at times to read and understand the text. Perhaps there are also issues with figures (see specific comments). I have highlighted some of these difficulties arising from long sentences and graphics in the detailed comments. I believe the manuscript needs another “layer” of editing to improve readability.

I am not 100% sure about a general significance of the study, apart from the Grotte de Milandre context, because the study did not really discuss what the role of AM-vegetation in the soil and cave microbes (including fungi) could be. Yet, it is an honest study, with commendable efforts, which deserves to be published.

Specific Comments

line 75  Perhaps strike-slip fault is easier to understand for all.

Also, same line, add: Late Jurassic (Middle-Oxfordian)  St-Ursanne Fm which overlies Early to Middle Oxfordian marls (Bischof et al., 2018).

Lines 80-8110 to 20 m deep or wide?

Fig. 1 Perhaps superimpose the name of the village on the aerial photo? It would also be useful to add the cross-section of the cave (see Vuilleumier et al., 2019 and 2021) within its geologic cross-section and plot the sampling point in the cross section. It would be easier to follow/understand the upstream-downstream datasets and the “cross-trip”.  This would make it easier to interpret figures such as Fig. 8, especially if the sample sites number were reported both on the cross section and the figure(s).

Line 114. ….at cool temperature…or in cool boxes …in the refrigerator….

Line 125 I suggest to change into: …and in single lines from a depth of 5 m (…).

Line 170. I suggest to state somewhere here that F¹⁴C is the fraction modern (post-bomb peak ¹⁴C).

Fig. 2. I did not see the theoretical carbon reservoirs for the C4 (green). I see a grey bar.

Line 225 10’700 and all the following concentrations. I do not understand the ^, why 10’000 and let’s say 6100?  Should be 10,700; 10,000; 6,100; 3,000…

Line 355. I suggest to  change in “…shallow boreholes gas samples, which vary with depth, vegetation cover and seasons, suggest…….”

Line 364.  I suggest: due to: i) slow, but persisting metabolic microbial activity at below freezing temperatures as low as….; ii) reduced, but persisting rhizosphere …; reduced transportation….

There should be a role for soil lichens and hyphae here as well. They seem to adapt to temperature changes (Lange, Green, 2005; Atkin et al., 2009). The role of lichen and fungi is still poorly explored in cave-based research. Yet, the ubiquitarian presence of fungi in the soil zone and, in particular, in the root zone, is probably important. Any idea if you have AM-plants (associated with Arbuscular mycorrhizal fungi)? AM are present in the majority of terrestrial ecosystems on earth. For example, Glomus intraradices has been found in Swiss alpine meadows (Sykorova et al. 2007). Would you be able to comment on a potential role of fungi and soil lichens rather that only on that of  “microbes”?

Line 366-6  “Similar seasonal trends in CO2 concentrations have been reported in other studies (Billings et al; 1998; Pumpanen et al., 2003; Zhang et al., 2023). The CO2 concentrations in the Shallow 2 meadow boreholes do not show seasonality as  pronounced as… due to consistent CO₂ accumulation year-round”.

Line 376 “with the meadow soils deeper, more… higher…” than?

Line 380-5.  Perhaps the work of Muhr et al., 2009 (Global Change Biology) on soil frost effects on soil respiration and its radiocarbon signature may shed some light here. Contrasting behaviour in moist and dry soils has also been observed for AM, which is an aspect that is not much investigated in karst science (see Lekberg, Y. and Koide, R.T., 2008. Effect of soil moisture and temperature during fallow on survival of contrasting isolates of arbuscular mycorrhizal fungi. Botany, 86(10), pp.1117-1124).

Lines 397-401. Sentence is too long. One needs to breathe in-between.

Line 404. Upward movement of cave air….and downward ….

Line 450 dilution with atmospheric air

474- Dissolution of the host rock …two extreme cases: completely closed and completely open systems. In a completely closed...

1. To be consistent, replace fully with completely.

1. Remove solid, as a rock is, by definition…solid. Perhaps replace with carbonate ?

511-513. The sentence is a bit difficult to read. Perhaps stop at highest 𝐹14𝐶𝑐𝑎𝑣𝑒 .  Then: When 𝐹14𝐶𝐷𝐼𝐶 = 0.88 it results in an highly unlikely scenario on the basis….

520…though this hypothesis is highly unlikely (?).

529 (we considered or took a range…). There are several sentences mixing tenses. Please try to be consistent. Commonly, the past tense is used:…we took…we assumed…But perhaps this is not so important.

1. I would omit “fresh”.

1. …fed through fracture flow. (fresh seems here to be associated with fracture…fresh fracture-flow?)

595 One possibility: could gas trapped in formation water from the marls contribute CO2 (and methane)? (see Cailteau, 2008).

607 “The topography likely reflects influences of the secondary porosity in the bedrock, which is possibly more highly fractured in the meadow doline formation compared to the forest leading to a greater effect of ventilation in the meadow area and an undisturbed modern 14C signature of forest soils. As the doline structure likely has a higher secondary porosity and is more highly fractured than the higher bedrock, ventilation effects may be more important here than at the other locations”.

I did not understand the full meaning of the sentences.  Does it mean that the formation of dolines is guided by fractures? But what is the “higher bedrock”? Also, given that the Formation(s) into which the cave develops is not completely pure limestone, the potential effects of the presence of marls should somewhat be considered.

After this, I gave up in the reviewing of sentences, because I feel it is not really the role of a reviewer.None

This paper investigates the subsurface CO2 dynamics in the Milandre cave karst system in northern Switzerland, focusing on seasonal and spatial variations. ​ The study spans several years and involves analyzing CO2 concentrations, stable carbon isotopic ratios (δ13C), and radiocarbon (14C) compositions from various sources, including the outside atmosphere, boreholes, and cave air. ​

Key findings include i) that CO2 concentrations in shallow boreholes are higher in summer due to increased respiration rates and lower in winter, ii) cave CO2 concentrations and isotopic compositions are influenced by temperature-driven seasonal ventilation, iii) the cave air CO2 levels were anti-correlated depending on where exactly in the cave the measurements were obtained, and iv) that the CO2 in the cave is primarily sourced from modern soil respiration and aged organic material in the epikarst. ​

The research highlights the complexity of CO2 dynamics in karst systems and the significant impact of seasonal and ecological factors on carbon cycling within the terrestrial Critical Zone. ​It is sound research reflecting the results of a long and intense monitoring campaign, and I recommend publication following the minor revisions suggested below.

There is a substantial amount of text devoted to describing the data, but a bit more describing the characteristics of the site with respect to carbon would be useful. For example, is there organic sediment within the cave site? How deep are the deep roots in the forest? Related to the above, I could not find the depth to the cave anywhere. It could well be here (apologies if so), but could it be included in the site description section? There is the sentence ‘The unsaturated zone ranges between 40 and 80 m depth, with a saturated zone of ~ 20 m’ but the cave could be developed anywhere within the unsaturated zone, so this does not let the reader know where the cave is relative to the surface. This is really critical for the interpretations presented here; for example, do the tree roots reach the cave? Also, why is the saturated zone only 20 meters? What is below the saturated zone?

Line 48: Suggest rephrasing ‘exclusively aged CO2 reservoirs’ – a bit awkwardly phrased

Line 60 (and throughout): the comma in the 1,000s separator in the numbers should be on the bottom rather than the top.

L69: ‘provenance’ not plural

L112: Elaborate – if it was built in house, why is there a brand name? Also spell out ‘nondispersive infrared (NDIR)’ for readers who may not be familiar with acronym

L138 and elsewhere: ‘through trip’ rather than ‘cross trip’

L201-202: Technically this is all correct, but consider using either ‘r’ instead of the letter ‘rho’ or spell out ‘rho’ instead of using the letter. It looks a lot like the ‘p’ in p-value, and this would avoid some confusion. But up to the authors.

L207: Use ‘average’ or ‘mean’ but be consistent. On this line both words are used in the same context. I’d recommend ‘mean’ when talking about a statistical quantity and ‘average’ when talking about something in general terms.

Figure 2: Consider writing ‘C3’ and ‘C4’ and putting the text inside the filled boxes. Or connect the labels to the boxes with a connector line. The issue is that where they are now makes it seem like where the actual field is, and some readers might not notice the boxes. The same is very true for the blue atmosphere box – it is almost invisible. I definitely recommend drawing a line or an arrow between the text and the blue box in that case.

Figure 2: Use ‘Upstream (Cave air)’ and ‘Downstream (Cave air)’ in the legend, this might make it clear that these are in-cave measurements. It is clear from the text, but someone scanning the paper would be forgiven for thinking you had somehow taken these measurements from a stream.  

Figure 3: It might be useful to show atmospheric CO2 concentration and chemistry here as well?

Figure 3 and elsewhere: Why are there a maximum and minimum numbers for each timeslice? I assume that this is max and min of the monthly recorded values? I feel that this is probably obvious - but i can't find it. So, if it is in the manuscript, it should be made more apparent, including here in the figure caption.

L263: Is this truly referring to the epikarst (the relatively shallow zone of fractured rock directly beneath the soil), or is this bedrock within the unsaturated zone? Earlier, you use the term unsaturated zone, and the depths of the boreholes seem deeper than the epikarst, so I think you mean might mean unsaturated zone. The two have different meanings, and not interchangeable. Similarly, ‘karst’ refers to all the karstified limestone bedrock, regardless of whether it is saturated or unsaturated – it is not just the rock above the cave. Please change to appropriate terminology if these terms are used incorrectly. If they are used correctly, you might want to consider defining the term ‘epikarst’ since it is so often used incorrectly. This way readers will know for sure that what you are referring to as the epikarst is the irregular surface between the karst and the soil.

L418: How does borehole Deep 1 influence the correlation? I think needs to be reworded.