Review of the paper: “Variability of dimethyl sulphide (DMS), methanethiol and other trace gases in context of microbial communities from the temperate Atlantic to the Arctic Ocean”

by Valérie Gros et al.

General comments

The paper by Gros et al. makes a substantial contribution to our understanding of the spatial distribution of climatically important trace gases and their potential underlying drivers. I would highlight the finding of very high DMS and significant MeSH concentrations under ice, the uncoupled DMS and MeSH distributions, and the distinct correlations between each VOC and bacterial ASVs at quite fine taxonomic resolution. Strong relationships between MeSH and bacterial ASVs in comparison to other VOCs is indicative of widespread bacterial DMSP metabolism. The paper is clearly written and well structured, and its messages are well supported by the observations. In the specific comments below I make some small criticisms that should be addressed. I suggest several additional citations, which I find important to both support the authors’ findings and give fair credit to previous studies. I was also a bit disappointed by the little use authors make of HPLC data. My feeling is that they are missing an opportunity to assess the relative importance (even if based only on correlation patterns) of phytoplankton vs. heterotrophic bacterial diversity in controlling VOCs distribution in the area north of 80N that they sampled from Niskin bottles.  My main criticism is with regards to the figures: they should be improved to make them more self-descriptive.

Specific comments

L39: Please add more up-to-date references, given that major advances in understanding of DMSP catabolism pathways have been made since 2007. 

L41: suggested citation: Kiene, 1996.  Production of methanethiol from dimethylsulfoniopropionate in marine surface waters

These articles may be of interest to provide a more complete view:

L45-48: Rodríguez-Ros et al. 2020. Distribution and drivers of marine isoprene concentration across the Southern Ocean

L49-50: Fichot and Miller, 2010. An approach to quantify depth-resolved marine photochemical fluxes using remote sensing: Application to carbon monoxide (CO) photoproduction

L56: Acetaldehyde is also photoproduced: Zhu and Kieber, 2020. Global Model for Depth-Dependent Carbonyl Photochemical Production Rates in Seawater.

L65: Lewis and Arrigo, 2020. Changes in phytoplankton concentration, not sea ice, now drive increased Arctic Ocean primary production

L65: Galindo et al. 2014. Biological and physical processes influencing sea ice, under-ice algae, and dimethylsulfoniopropionate during spring in the Canadian archipelago

L65: Wohl et al. 2022. Sea ice concentration impacts dissolved organic gases in the Canadian Arctic

L66: Galí et al. 2019. Decadal increase in Arctic dimethylsulfide emission

L68: Two good examples of changing phytoplankton species distribution

Oziel et al. 2020. Faster Atlantic currents drive poleward expansion of temperate phytoplankton in the Arctic Ocean

Orkney et al. 2020. Bio-optical evidence for increasing Phaeocystis dominance in the Barents Sea

L259-260: I don’t see how the correlation between DMS and Chl is connected to diatoms being the main photosynthetic group. Please rephrase.

L261: I cannot see the cyan squares of station 43 in the CO panel.

L305: Perhaps mention that Arctic waters feature much higher CDOM content than typical oceanic waters. The approach used by Conte et al. (2019) estimated CDOM using a bio-optical relationship between CDOM and Chl developed for typical (case I) oceanic waters (Morel et al., 2009). Arctic waters do not conform to this bio-optical type and are typically seen as optically complex waters with compound influence of oceanic, riverine and ice-melt waters with distinct signatures in terms of CDOM and particle loads. Failing to account for the high CDOM content is likely to result in underestimation of CO photoproduction.

L332: A note of caution: the values reported by Davie-Martin et al. (2020) for the NAAMES expedition are not credible in the case of DMS production rates. The highest DMS production they found in May, around 43 nM h-1 (1000 nM d-1), is about 15 times higher than any previous measurement (Galí and Simó, 2015, their figure 3a). This might have been caused by the bubbling in their experimental setup, which is known to induce DMS production in stressed cells (eg Wolfe et al., 2002. Dimethylsulfoniopropionate cleavage by marine phytoplankton in response to mechanical, chemical, or dark stress). Similar artifacts may have affected measurements of the production rate of other VOCs in that study.

L349: Suggested citation: Hayashida et al. 2020. Spatiotemporal Variability in Modeled Bottom Ice and Sea Surface Dimethylsulfide Concentrations and Fluxes in the Arctic During 1979 – 2015. Quoting from their abstract: “...model results indicate that the bottom ice DMS and its precursor dimethylsulfoniopropionate production can be the only local source of oceanic DMS emissions into the atmosphere during May prior to pelagic blooms”.

L350-351: this view can be nuanced. At high latitudes, the seasonal correlation between DMS and Chl is typically positive and quite high. See e.g. Lana et al. (2012) Re-examination of global emerging patterns of ocean DMS concentration, their fig. 4. Also:

Galí et al., 2018. Sea-surface dimethylsulfide (DMS) concentration from satellite data at global and regional scales. Table 1 and Fig. 7.

Wang et al., 2020. Global ocean dimethyl sulfide climatology estimated from observations and an artificial neural network. Table 1, section 3.

Figure 1: It would be very helpful to see the main currents and the different water masses on this map. For example, this would support the description given in L180-186, several parts of the Results and Discussion, the summary given in Table 1, etc.

Figure 2: I strongly recommend to depict somehow the water masses along the transect, for example with colored horizontal bars on top of the plot.

Figure 3: It would be useful to provide more commonly used taxonomic classifications/levels for some bacterial genera. For example, Yoonia-Loktanella and Ascidiaceihabitans tell nothing to me, but I immediately associate Rhodobacteraceae with certain types of reduced sulfur metabolism.

Figure 4: Please indicate (for example in the legend) whether stations are in pre-bloom or bloom stage, perhaps distinguishing the shelf stations as well.

Figure 5: same as Fig. 2.

Technical corrections and typos

L156: “at their” repeated

L254: “but the here found concentrations” sounds a bit awkward, please reword

L292: “as already been found”, please remove “been”

L324: nM, not nm

General Comments

This manuscript describes a comprehensive set of dissolved reactive gas and microbiological measurements taken on a research cruise from the North Atlantic to the Arctic. The authors extensively discuss the sources and relationships between the measured trace gases and microbiology, and focus their discussion on dimethyl sulfide (DMS) and methanethiol (MeSH). They show that MeSH does not correlate with DMS during the entirety of the cruise. They find that MeSH can contribute on average 20%, and up to 50%, to the total waterside sulfur budget, defined by the sum of DMS and MeSH. Overall, this manuscript is well-structured and presents new findings that are valuable to the biogeosciences and atmospheric chemistry communities and should be published after the following main comments are addressed.

My main comment for this manuscript is that a more nuanced discussion of variations in the measured MeSH/(DMS+MeSH) ratio and the dominant factors controlling it would be extremely helpful. Little information currently exists on how this ratio varies based on environmental parameters, which has impacts for how we think about SO₂ production. This dataset provides measurements of MeSH in a region for the first time with varying temperature and salinity, meaning that we now have data to form more accurate models of SO₂ production based on how this ratio scales with different waterside parameters. Suggestions for discussion on this topic include:

1. Addition of a column containing MeSH/(DMS+MeSH) to Table 1. Can any trends from the water classifications (salinity, temperature) explain the observed variations?
2. In line 371-372, it’s noted that at 78.6 ºN, MeSH contributes up to 50% of total sulfur, but only 20-40% in 70-75ºN. What is driving this difference?
3. Looking at Fig. 2, it looks like in some regions MeSH and DMS covary (>71ºN) and in some regions, there is little correlation (<68ºN). Some more statistical analysis and discussion of why there seems to be a correlation in certain water masses/time periods but not others would be useful.
4. 5 could be revised to provide more information about any environmental parameters controlling this ratio (colored points by salinity, temperature, chlorophyll?) or additional regressions against these variables instead of just latitude.

Specific Comments – Manuscript

Line 19: It would be helpful to make it clear somewhere in the abstract that all gas measurements are in the dissolved phase in the seawater and not in the air. Potentially could also add “dissolved” to title.

Line 39: Instead of “rapidly oxidized”, can you state the atmospheric lifetime of DMS?

Line 40: It would be useful to add some references to the CLAW hypothesis. Some suggestions:

• Bates, T. S., Lamb, B. K., Guenther, A., Dignon, J., and Stoiber, R. E.: Sulfur emissions to the atmosphere from natural sourees, J. Atmos. Chem., 14, 315–337, https://doi.org/10.1007/BF00115242, 1992.
• Charlson, R. J., Lovelock, J. E., Andreae, M. O., and Warren, S. G.: Oceanic phytoplankton, atmospheric sulphur, cloud albedo and climate, Nature, 326, 655–661, https://doi.org/10.1038/326655a0, 1987.

Line 41: Suggested references for DMSP demethylation producing MeSH:

• Kiene, R. P.: Production of methanethiol from dimethylsulfoniopropionate in marine surface waters, Mar. Chem., 54
• Kiene, R. P. and Linn, L. J.: The fate of dissolved dimethylsulfoniopropionate (DMSP) in seawater: tracer studies using 35S-DMSP, Geochim. Cosmochim. Ac., 64, 2797–2810, https://doi.org/10.1016/S0016-7037(00)00399-9, 2000b.

Line 42: The atmospheric impacts of MeSH are less well-characterized than DMS, but we do know some about MeSH impacts based on its oxidation and reactivity. See references below:

• Butkovskaya, N. I. and Setser, D. W.: Product Branching Fractions and Kinetic Isotope Effects for the Reactions of OH and OD Radicals with CH3SH and CH3SD, J. Phys. Chem. A, 103, 6921– 6929, https://doi.org/10.1021/jp9914828, 1999.
• Tyndall, G. S. and Ravishankara, A. R.: Atmospheric oxidation of reduced sulfur species, Int. J. Chem. Kinet., 23, 483–527, https://doi.org/10.1002/kin.550230604, 1991
• Novak, G. A.; Kilgour, D. B.; Jernigan, C. M.; Vermeuel, M. P.; Bertram, T. H. Oceanic Emissions of Dimethyl Sulfide and Methanethiol and Their Contribution to Sulfur Dioxide Production in the Marine Atmosphere. Atmospheric Chem. Phys. 2022, 22 (9), 6309–6325. https://doi.org/10.5194/acp-22-6309-2022.

Line 49: Isoprene has also been shown to have a photochemical source.

• Ciuraru, R.; Fine, L.; Pinxteren, M. van; D’Anna, B.; Herrmann, H.; George, C. Unravelling New Processes at Interfaces: Photochemical Isoprene Production at the Sea Surface. Sci. Technol. 2015, 49 (22), 13199–13205. https://doi.org/10.1021/acs.est.5b02388.

Line 52-53: Add a citation for OVOCs affecting the oxidative capacity of the remote atmosphere. Potentially this one could work:

• Singh et al. (2004). Analysis of the atmospheric distribution, sources, and sinks of oxygenated volatile organic chemicals based on measurements over the Pacific during TRACE-P. Journal of Geophysical Research Atmospheres.

Lines 55-59: Acetone, methanol, acetonitrile, and acetaldehyde can also be anthropogenic, affecting whether the net flux is positive or negative. It is worth adding this in addition to whether the flux is positive or negative depending on oligotrophic water.

Lines 58-59: Another reference for acetone and acetaldehyde flux.

• Phillips, D. P., Hopkins, F. E., Bell, T. G., Liss, P. S., Nightingale, P. D., Reeves, C. E., Wohl, C., and Yang, M.: Air–sea exchange of acetone, acetaldehyde, DMS and isoprene at a UK coastal site, Atmospheric Chem. Phys., 21, 10111–10132, https://doi.org/10.5194/acp-21-10111-2021, 2021.

Line 115: What do the different blue colors mean in Fig. 1? I suggest adding information about the water classifications to this figure as well, like Fig. S6.

Line 152: It has been shown previously that other molecules can be measured in PTR-MS at the unit mass 63 where DMS is measured, such as ethylene glycol. Has this been accounted for in background measurements? Otherwise, if these are measured along this transect, they could artificially inflate the DMS measurements. It should be explicitly stated that all VOC measurements were taken at their unit mass m/z + 1 mass.

Line 155: Cite Blake et al. (2009) again for the thermodynamics of the proton transfer reaction.

Line 155-156: While PTR-MS can be a soft ionization technique, there is still the possibility for fragmentation of larger molecules to affect your measurements, so quantifying at the m/z +1 mass may be the protonated molecule in addition to fragments of larger molecules. Have there been control experiments to support quantifying the molecules of interest only at the m/z+1 mass?

Line 157-158: What is the residence time in your tubing? Does this affect the measurements of any of your molecules, like acetonitrile?

Line 164-164: I don’t see information in the SI on how MeSH was calibrated. Was this an assumed equivalent sensitivity as DMS? This calibration should also be included in the SI.

Line 194: I don’t see any discussion of uncertainty for the trace gas measurements. Some discussion of this should be included either in the methods or in 3.1.1.

Line 195: What are your detection limits for acetone and acetaldehyde?

Line 205: I’m curious what’s causing the high MeSH between 70 and 73-75 ºN?

Line 209: I think this figure can be edited to help the story flow better. My suggestions are:

1. Since CO presumably has a different source than DMS and MeSH, having it on the same panel is distracting. I’d suggest making this a 4-panel figure with CO on its own.
2. Can some information about the water masses (info from Table 4) be included? Perhaps as a shaded background.
3. Can information about the timing of these measurements be included? By plotting against latitude, it is hard to understand how many points are represented at each latitude, especially in the horizontal transect region near 70ºN.

Line 227: I am unclear what MeSH_DMS is on Fig. 3b x-axis? Is this MeSH/(MeSH+DMS)? If so, should be updated to read more clearly.

Technical Corrections – Manuscript

Line 35: “source and sink” should be “sources and sinks”

Line 99: “some leads present ),” should be “some leads present),”

Line 123: “Chl a l concentrations” should be “Chl a concentrations”

Line 156: “at their at their” should be “at their”

Line 254: “but the here found concentrations” should be “but here the concentrations found”

Line 324: “nm” should be “nM”

Technical Corrections – Supplemental

Line 8: There is something cut off in the upper righthand corner of Fig. S1.

Line 61: Fig. S3 is blurry and hard to read.

Line 77-79: I am unclear what it means for the Henry’s law constant “whatever the solubility of the compound over 4 to 5 orders of magnitude”

Line 118: Bottom right panel y-axis in Fig. S6 is cut off. What is being plotted on the x-axis – is this both phytoplankton functional group and chlorophyll? Can the x-axis label be adjusted to be more clear?

Line 123: Should read “Supplement S7: ”

Lines 131-146: References should be alphabetized and formatting should be consistent.

Line 133: Callahan et al. reference is missing a year.

Summary

The manuscript by Gros et al. reports on a large-scale survey of DMS, methanethiol and several other gases along the gradient between the eastern subpolar North Atlantic and the Arctic. Their main goal was to assess the concentrations and spatial distribution of these compounds and relate them to the environmental (physical, microbiological) conditions via statistical correlations. Based on their results, the authors aim to improve our current understanding on trace gas cycling in the context of a changing Arctic.

General assessment

The topic of this manuscript is certainly relevant for a wide biogeochemistry community and contributes to amend the large gaps of data coverage for marine trace gases; in particular for compounds which are understudied in comparison with other climate-relevant gases such as CO2 or CH4. The paper is very well written, its structure is clear and the methodological approaches are both sound and explained with enough detail. Although the manuscript is quite descriptive, the authors state clearly that they aim to report on the results of their survey. Hence, from that perspective, they were successful in achieving that. In my opinion a significant drawback of the study is the absence of air-sea fluxes of the different compounds measured in surface waters. Based on the content of the methods presented by the authors I cannot judge whether they have information at hand to do so, but it would be worth making an effort to provide estimates of air-sea fluxes for at least some of the compounds (I do think this can be done for DMS and CO). Other than that, most of my comments to the paper are minor (see below).

Specific comments

• Title: The current title is rather long and can be misleading. The word “Variability” is should not be used generally here since the authors are addressing the spatial variability of several trace gases, not their temporal variability.
• 29-30: Revise syntax, in particular after “understanding of”. Also, I recommend stating how specifically the paper contributes to that understanding, since right now this could mean anything.
• 80: Please replace the word “levels” for “concentrations” here and it all instances where you refer to that quantity.
• 81: “numbers” is too unspecific. Please refer precisely to which measurable quantity you are referring to here.
• 82: Bacterial diversity and water masses were addressed. No sea ice data cover was presented and therefore it should not be presented as a factor in your experimental design.
• 88: The citation to Peeken (2016) is unnecessary. I know such reports have a doi number, but they do not constitute a source of peer-reviewed information and should only used when absolutely needed (see journal’s regulations)
• 92: Stating that “usually” there is not sensor drift is not enough, even if the sentence is supported by a publication. The authors need to show that this was indeed the case during their survey in order to keep the credibility of their observations.
• 106-107: Here the citation is also unnecessary and should be removed. If the authors want to refer to the data used, there are better ways such a data set doi from Pangaea.
• 117: Delete point
• 123: delete “l” after Chl a
• 123: The R2 value in S1 is different than the one shown here. Revise.
• 153-154: Revise wording. I would suggest “The measurement principle of PTRMS is (…)” or similar.
• 214-217: The details on how this statistical analysis was setup should be explained in the “Material and Methods” section (i.e. independent and dependent variables, etc.). Otherwise the statement seems arbitrary (i.e. coming from nowhere).
• 234-235: Explain the details on how the system was adjusted. This reads as if the authors used the continuous system for profiling. Was that the case? If so, a detailed description is needed.
• 259-266: I am not convinced of the approach here. Why was station 19 removed from the analysis? It appears that although stations 19 and 32 have high productivity, both isoprene and CO behave completely different. Also, if one compares CO concentration at stations 19 and 39 (having contrasting chl a concentrations), it becomes evident that CO is not affected by the same processes as other gases. The reasons for this are unfortunately not discussed at all. In order to explain the variability of some of the CO concentrations at depth (e.g. at stations 32 and 43), the authors claim that differences in the profiles are due to “decreased photochemical production following lower light penetration”. However, this is the case for all stations and therefore it is not a compelling reason to explain the decrease with depth. Perhaps the authors rather refer to the effect of different sea ice coverage percentages in light penetration (?). If so, they can easily explore this possibility by using such data which is widely available.
• 303: Is the mean value for surface waters or does it include the water column measurements? Please clarify.
• 308: Personal communications are not appropriate. Even less in this case since there are already two citations supporting the statement.
• 353: Same comment to personal communications. The authors already used Dybwad et al. (2021) as a defining criterion for the bloom stages in the study area at the time of sampling.
• 385-387: This statement is contradictory with the results presented by the authors for CO. Based on the data presented it is only clear that CO production is not necessarily tied to a biological component and that photochemistry might have had a more significant role at the time of sampling. The authors argue (L.313-315) that low CO production by diatoms might be the explanation for the low concentrations at e.g. 19. However. this is speculative and cannot be substantiated with their observations. I recommend revising this aspect of the discussion.
• Supplementary information: there are inconsistencies in the naming of Figs. S6-S8. For instance, in L.235 S7 is mentioned although it does not match what is actually shown. In Fig. S8 no CO is shown (although announced in the main text) and the caption does not match the figure.