The paper by Schüttler et al. offer some interesting data on BVOC exchanges between the soil and atmosphere and extends our understanding of these fluxes in tropical systems.  As a Biologist, not a Chemist, I found the methods descriptions commendably clear, logical, and easy to follow.  I do NOT know enough to comment critically on these methods, but the descriptions made perfect sense to me.

The importance of BVOC fluxes in these systems has been known for over thirty years, but we still have very few data sets of soil fluxes, per se; this study, even though it is based on few chambers in one site, offers some tantalizing insights.  The authors need to bear in mind in their statistical models when their samples are truly independent, when they are time series, and when they are pseudo-replicated.  I suspect they will want to either re-structure their models or, at the very least, acknowledge when their data violate assumptions of independence.  I do not see this as a major issue because the results are so striking. 

A few results stood to me, with respect to the underlying Biology.  I hope the authors find these useful as they revise.

1. the lack of isoprene uptake in the litter layer suggests that litter microbes have not adapted to this carbon source. Given the earlier described role of litter microbes in methanol and terpene consumption, this result is worth highlighting.
2. the difference in enantiomeric variability between the monoterpenes and the sesquiterpenes is congruent with earlier ideas that the sesquis come from the MVA pathway while the monoterps are made by the DOX-P pathway.
3. the idea of terpene emission from leaf litter, per se, rather than microbial activity during decomposition, could be developed a bit. The reason for pursuing this distinction is that a fair bit is known about the distribution of different terpenes in different plant taxa, so the direct leaf-contribution to BVOC fluxes should be predictable if the plant species composition is known.
4. For over 30 years, thanks to Carleton White, we’ve been aware of the importance of soil microbes as consumers of BVOCs. When the flux is net upward, as Schüttler et al. found here for the 10 and 15 C terpenes, then the consumption processes are smaller than the production ones. The results of this manuscript suggest that it is time for some more process-based studies of soil BVOC fluxes so that we can begin to estimate gross fluxes also. 
5. I applaud the authors for their wet & dry season measurements, but I think the link to El Nino is rather tenuous, and, imo, they should de-emphasize this point. But I leave this to the editor’s discretion.
6. the authors make the important point that soil fluxes could result from microbes and/or roots. There have been similar observations on volatile sulfur fluxes in tropical systems. Distinguishing plant from microbial sources of production and consumption is crucial for developing predictive modeling.  If the authors choose to discuss this, they should bear in mind McDonald and Falls’s older work showing microbial consumption on the leaf surface, which could look like leaf consumption. 

In short, this is a fine contribution and one that I look forward to citing when it appears in the literature. 

General Comments

This study examines fluxes (emissions and uptakes) of VOCs from soil in the Amazon rainforest. The authors present seasonal measurements spanning multiple wet and dry periods and investigate the chirality of VOC fluxes, timely and highly relevant research questions. The results show seasonal shifts in the emission and uptake of various VOCs (MTs, SQTs, isoprene, and two isoprene oxidation products) between soil and atmosphere. Dry conditions led to SQT soil emissions, but MTs and ISP uptake. Additionally, the enantiomeric compositions of VOCs emitted from soil differ from those present in ambient air, with further seasonal variation observed.

I consider this study on the rainforest ecosystem very important, given also the very limited existing knowledge. The methodologies, particularly regarding VOC flux measurements, are generally robust, and the manuscript is nicely written and well-organized. However, I have specific concerns about the number of replicates included in the study, which should be taken seriously.

Specific Comments

My primary concern is about the number of replicates used in the study. The authors utilized three chambers placed in three separate soil plots, which appears to be the minimum acceptable number of biological replicates (i.e., three). However, two of these chambers represent soil emissions with natural litter abundance, while only one chamber represents a plot without litter. Additionally, the rationale for excluding (or separating) the “spot 1” near the termite nest from the analysis and treating it separately is unclear. Are these data considered outliers? Why can’t those measurements be considered part of the biological variability that exists in the rainforest? Also, the results on soil emissions following litter removal are based on only 2 replicates, which raises questions about the reliability of the findings. The results of figure 5 should be also tested for significance.

Another point that is unclear is why the blank measurements were not subtracted from the measured data, but instead showed alongside it. The presented flux data might therefore be offset. In addition to performing a background correction, I suggest calculating and including the LOD (or LOQ) for these measurements in the graph to assess the technical limitation of the flux measurements. Was the blank measurement consistent throughout all seasons or years, or did it fluctuate?

The study design for Section 3.4, 3.4.1, and Figure 4 is unclear. The analysis aimed to investigate the effects of soil properties on soil fluxes. However, it is based on recurrent measurements from the same chamber/spot, ie, it appears based on 1 biological replicate.

The correlation between soil fluxes and environmental conditions (fig.6)  is interesting. Why was the chirality not considered here, being a central focus of the study? Also, the figure 6 shows the correlation between SWC and VOC emissions, but it is not mentioned anywhere in the text.

What is the VOC breakthrough volume in the adsorbent? Could higher ambient VOC levels cause cartridge breakthroughs and affect soil emission estimates?

The authors found that isoprene concentrations at soil level peaked at noon, while MT and SQT peaked later. Can you elaborate on this?

The SQT soil fluxes reported here are 10 to 10,000 times lower than those in other studies. Is there a technical reason in the measurements or in the data analysis? Can the authors compare and discuss their measured mean air concentrations with published data?

There is extensive evidence that soil bacteria degrade isoprene, which may be useful to refer to (e.g., El Khawand et al., 2018; McGenity et al., 2018; Murrell et al., 2020).

Technical Corrections

I recommend adhering to SI (International System of Units) when presenting emission fluxes, ie, using seconds instead of hours.

L22: "Soil" may be removed or bracketed as the functions of those VOCs extend beyond soil ecology.

L107: If measurements were taken across all seasons and the mean and SD are reported for each season, I assume the 23 refers to the total number of blank measurements rather than replicates of each season, year and field campaign conducted?

L130-135: Please clarify which standards were purchased as liquids and which were supplied by Apel-Riemer as a gas standard mixture.

L141-142: This section is unclear. Are you referring to technical variability? Please clarify and also report this variability.

Table A1: include the retention index.

Enantiomer analysis of VOC is not standard practice in the VOC research community. It would be informative to include a chromatogram in the supplement.

Ref.:

El Khawand, M., Crombie, A.T., Johnston, A., Vavlline, D.V., McAuliffe, J.C., Latone, J.A., Primak, Y.A., Lee, S.-K., Whited, G.M., McGenity, T.J. and Murrell, J.C. (2016), Isolation of isoprene degrading bacteria from soils, development of isoA gene probes and identification of the active isoprene-degrading soil community using DNA-stable isotope probing. Environ Microbiol, 18: 2743-2753. https://doi.org/10.1111/1462-2920.13345 McGenity, T.J., Crombie, A.T. & Murrell, J.C. Microbial cycling of isoprene, the most abundantly produced biological volatile organic compound on Earth. ISME J 12, 931–941 (2018). https://doi.org/10.1038/s41396-018-0072-6

Murrell JC, McGenity TJ, Crombie AT. Microbial metabolism of isoprene: a much-neglected climate-active gas. Microbiology. 2020 Jul;166(7):600-613. doi: 10.1099/mic.0.000931

General comments

The topic of this manuscript is of importance, as soil emissions have been severely neglected in the BVOC field and little is known about the processes affecting the magnitudes and types of emissions. While canopy emission especially in the tropic have been studied extensively, we still know next to nothing about how emissions and uptake from soil will change in the changing climate or due to extreme weather conditions, such as the El Niño. I also find the inclusion of stereoisomers into the larger discussion of terpenes interesting, especially if they can be used to track or estimate changes in biological processes due to environmental stressors.

While the manuscript is in general well written and has merit, I have some critical comments – especially regarding the methodology and research questions. My main concerns are:

1. As I understand, authors measured BVOC fluxes from 3 separate locations close by to each other, but all differing; 1 without litter, 1 with litter, 1 near a termite nest. This means that only one true biological replicate per location was measured, which is – in my opinion – not sufficient for an ecological study. Authors have done pseudoreplication within one chamber for seasonal changes, but as the location of two of the chambers were changed between seasons, temporal comparison even within one chamber is difficult. Same applies for blanks, where only one spot was sampled, resulting in pseudo/technical replicates, not representative of the true natural variation. While the authors express that their aim was to “screen” differing extremes by placing the chambers in distinct locations, why not have multiple chambers in those distinct locations instead of one? Why blanks were only measured in one location? As the authors themselves express, the litter density varied significantly even within a few meters – which may also be the case for soil microbiome, roots etc. – all possibly affecting the BVOC fluxes observed. It is also evident from the results (Fig.4) that spots 2 and 3 differ from spots 4 and 5. As such, the lack of replication in this study is my main concern, and authors must be careful when expressing what can be concluded based on their results.
2. Why was ambient air used as carrier gas in the BVOC sampling? As authors state in the introduction, soil fluxes are typically orders of magnitude smaller than canopy/vegetation emissions, so would it not make more sense to use zero air (VOC free air) when investigating soil emissions, rather than ambient air with potentially high levels of BVOCs? I assume authors have aimed to record the background with the other tube measured in parallel to the soil chamber, but this does not address this issue. Furthermore, in Fig.1, it looks like the ambient tube was sampled separately “near” the actual chamber. Why not have a T-piece at the inlet of the chamber with part of the ambient air going into the chamber and part into the ambient tube? This would ensure that the ambient tube captures all possible analytes and contaminants going into the chamber.
   
   Related to this, how were emission rates (Fig.3) for ambient samples calculated (what were Ct and C0)? What are the mean and standard error for (N=?); ambient air taken from the different locations, and the blank from another location? It is evident from emissions/uptakes that variation can be high between locations (e.g., monoterpenes), and for me, it’s impossible to say if they actually differ from the blank. Because no true replication for the blanks were conducted, authors cannot show what the natural variation was. Emission of monoterpenes in Jan 2023 and uptake of isoprene in Oct 2023 and 2024 are more evident, but otherwise, they may well be within blank levels.
3. Manuscript lacks hypothesis and research questions. Why measure enantiomers of terpenes or isoprene oxidation products? As the authors point out, soil BVOC fluxes are poorly understood, so the introduction would benefit from more detail for the reader’s benefit. At the moment, the introduction is vague and many important points are only mentioned but not elaborated on.
4. The section about atmospheric implications should, in my opinion, be omitted. Authors did not measure radical reactions, nor do they know what the in situ OH concentrations are. Furthermore, because of the issues with replication, the emission rates reported in this study should be considered tentative, and consequently, any estimates on atmospheric impact are rough at best and do not provide any usable information e.g., for modeling purposes.

Specific comments

Introduction:

38-41: As chirality is highlighted in this manuscript, I would like to know more about possible impacts of specific enantiomers being emitted. Authors should elaborate on what is known about (biogenic) processes and BVOC chirality and why differentiating between emissions of enantiomers is important. How can this information be used when assessing soil processes or atmospheric impacts?

45-47: Authors should elaborate how vegetation, soil properties etc. affect fluxes from soil. It would be beneficial for the reader if authors first describe some of the processes controlling BVOC fluxes from soil in general and then move on to describe what we know about tropical forests.

50-52: Authors should elaborate how these factors (water content, nutrient composition, temperature ect.) can affect soil uptake or emissions.

57-69: Again, authors should give more details about how weather conditions can affect (soil) BVOC fluxes in general and then describe what we know about their effects in rainforests. El Niño (and other extreme weather events) causes drought, which has been shown in previous studies to increase BVOC emissions, which again, can exacerbate extreme weather conditions. This cycle is worth elaborating on in the introduction, with relevant references.

61-64: What were the hypothesis and research questions? Why did you measure isoprene’s oxidation products – not otherwise mentioned in the introduction – and how do they link to the larger context or the study? 

Methods:

84: Define “close proximity”.

86: How much before sampling were the collars installed?

Fig.1. This figure would benefit from a schematic showing the different sampling spots (1-5) and which were with/without litter, effected by the El Niño etc.

103-104: Storage for up to 2 months seems excessive, especially because highly volatile compounds, like isoprene, were targeted in this study. How did the authors check that the long storage did not result in loss of analytes? Was an internal standard used?

129-130: Liquid standards were injected into the sorbent tubes under a nitrogen/helium flow I assume, not directly? Authors list the composition of the gas mixture, but what about the liquid standards?

140: Identifying enantiomers without authentic standards purely based on spectral library comparison is tentative at best. I would like to see how well compounds were separated in the chromatograms? While PARADISe is able to resolve convoluted peaks, I would be very careful with identification and quantification of compounds without an authentic standard when doing targeted analysis.

155: Define “near”.

164-168: As authors have sampled VOCs from chambers with and without litter, the litter composition should not be ignored. Did authors conduct any additional analysis of the litter or only the dry weight?

188: Why was soil moisture/temperature measured so far away from the BVOC sampling site? Was any replication conducted?

199: Statistical analysis needs to be explained more thoroughly, especially because pseudoreplication was used and the sampling sites changed in between samplings. Was the flux data normalized? Did you use repeated measurements ANOVA for dry-wet seasons and before-after litter removal? Authors need to specify which test was used for which parts of the data.

Results

Fig.2. Mean and standard deviation of what (N=?)?

264: Authors should list which signals have been summed as total MT and SQT.

280-282: How was emission highly seasonal, time-of-day dependent, and specific to soil conditions? Did you test this and their interaction with ANOVA? SQTs were significantly different in dry seasons 2023, did you test this and what was the p value?

Fig.3. See my general comment 2.

Fig.4. If these are mean fluxes, why not show standard deviation? Could you indicate in the figure which differences were statistically different.

Fig.5. Same comments as for Fig.4. Also, I would again be careful how the different spots are compared. Spot 2 and 3 are different, so authors cannot include them in their statistical analysis before and after litter removal the same way they would the same spot 5. As statistical methods were only briefly described by the authors, it’s also difficult to say what tests were used and how (e.g., repeated measures ANOVA or something else).

Discussion

I would combine the discussion about the emission rates with section 4.1 and consider very carefully what can be concluded from the results done with pseuduoreplicates. While the discussion about the different drivers behind the observed levels and blends of BVOCs is valid, the discussion about the measured emission rates (which do not reflect the natural variation in the environment) could be significantly reduced. Authors can discuss overall trends, but comparing hard numbers for emissions rates measured with pseudoreplicates is not valid and could be partly behind differences found between this study and others. As such, I would also omit the comparison with canopy emissions and the atmospheric impacts, and instead, expand on the discussion e.g., about the chirality which is a novel topic.

407-12: Could authors elaborate on how and how quickly soil microbiome can shift during extreme weather events? As no microbial analysis was done for this manuscript, it would be beneficial if authors demonstrate with relevant references if the time-scale of shifting microbiome is enough to explain the observed variations.

Appendix

Could authors provide the results from their statistical tests (p and F values, degrees of freedom) e.g., as table.

Could authors provide some example chromatograms to show the separation of enantiomers and corresponding identification for chiral compounds.

Technical corrections

In chemical formulas, numbers should be subscripts

In discussion, the verb tense should be consistent throughout, e.g. past tense.

Figure texts overall are too small.

43-45: Sentence is really long and hard to understand.

71: You already define the abbreviation (ATTO) in the introduction.

98-104: You could combine the information about the sorbent tubes and their preconditioning/storage to it’s on paragraph – separate from the description of sampling.

140: Define NIST and the version used.

199: Define ANOVA.

215: I think the sentence is missing something.

395: play only a minor role