General comments

The technical note describes outcomes from an experiment examining the suitability of three preservatives for the quantification of dissolved gas concentrations, and another experiment to determine the feasibility of HgCl2 preservation to derive CO2 fluxes from freshwater systems. Despite being toxic, HgCl2 is a commonly used chemical that prevents biological degradation of gas dissolved in water, even though alternatives exist. The study shows that these alternatives are effective and suggests substituting HgCl2 for less toxic preservatives. The results of this study are technically relevant, help reduce and avoid errors in flux estimation and support the implementation of user-friendlier substitutes for the preservation of freshwater samples. I think this study could be very valuable for many researchers in the field and I would like to thank the authors for their nice work. However, the manuscript would benefit from clarifications, and more details especially regarding the Methods are needed before publication.

Specific comments:

• Are the studied lakes representative for other lakes or waterbodies? Please clearly outline limitations of this study in terms of impact and application in a broader sense.
• Is it feasible to assume unfixed samples to represent “real” concentrations/fluxes, as control? Could you discuss this further and eventually consider renaming “control” to “unfixed” for the first experiment?
• The Methods section lacks necessary detail. I would suggest to restructure the section to make the experimental setup and respective study lakes clearer. E.g. in the study area section Lake Lundebyvannet should also be introduced, and ideally both lakes should be presented with the same level of detail relevant to the respective experiments. More importantly, I'm missing information on sampling procedures and their feasibility. Since this is a technical note, I believe the Methods should be sound. I added several comments in this regard below.
• I think the results of this study could be put into clear recommendations for future studies, and this could be part of the abstract and expressed more clearly in the discussion.
• The Introduction could benefit from adding some information about the other preservatives studied. The application of HgCl2 is broadly introduced (could be shortened), but the description of the substitutes dealt with in this study falls short. Are there any other studies where CuCl2 or AgNO3 were used to determine dissolved gas concentrations? Are there differences expected between the application of those two?

Please note, the numbers at the beginning of each comment denote the line numbers.

1. 26-27: can you be more specific about time periods (3w, 3m)?
2. 29: are low ionic strength / high DOC lakes representative?
3. 30: are these estimations valid for other lakes?
4. 31: I think explicitly adding recommendations here would be useful.
5. 59: better in regards to what?
6. 67: what is the impact of higher H+ concentrations?
7. 70: could you elaborate why this leads to an overestimation of CO2 concentration?
8. 82-92: it would help the reader if it was made clearer here that two different experiments were conducted and two different lakes were sampled for that, for example by 1)... 2)...
9. 87: This assumes that unfixed samples are the control, or "real results". Is that feasible?
10. 99: Did you collect the water from the surface? Did you use anything other than the bottles to avoid bubbling or degassing? Were samples temperature controlled (or otherwise controlled) between sampling and analysis?
11. 100: slowly poured - a bit vague? How could you guarantee no degassing?
12. 103: Are the results you got from the water samples representative for lakes in the region in terms of magnitude? Do the numbers represent means of the sub-samples or was each sub-sample used for determination of one of the parameters?
13. 105: As far as I understand, the concentration of platinum does not necessarily describe the color characteristics of water?
14. 106: how did you measure the temperature? is this an important information if the water was transported to the lab? Or did you preserve this temperature during transport?
15. 111: technically 3 treatments and one control. Which of the scenarios would presumably result in the most "real" concentration?
16. 112: why did you choose these time steps? could you elaborate if these times are representative?
17. 116: Is the preparation of the solutions part of the experiment? do the yielded concentrations have an uncertainty? or would a derivation not have an impact on the outcome?
18. 133: Did the fact that pH was not measured affect your study? Or was that one reason to use the PHREEQC model?
19. 137: Is the sampling strategy outlined different than the one for the first experiment? How did you achieve sampling water from different depths? It would also be nice if both lakes were described with the same level of detail.
20. 145: Could you clarify the purpose of DIC analysis in this study?
21. 147: Add name of TOC analyzer and/or merge with other sections below to avoid repetition. You state that samples were not fixed – why not?
22. 151: Did you compare the pH data with that measured with the pH-meter as mentioned above, or why measure twice?
23. 160: The temperature was recorded during shaking – do you mean the water temperature? What was the purpose?
24. 171: Do you mean ambient air was used for calibration? Did you know the concentrations of the ambient air?
25. 175 section: I think it would help to directly add formulas in a section in the appendix for better understanding and reproducibility.
26. 187: Could you explain the purpose of DIC analysis here or in earlier sections. Are the CO2 concentrations calculated in addition to the concentrations measured by GC for comparison? I think it may not always be clear where you used measured or calculated CO2 concentrations.
27. 244: Is it important to mention what files were input and output files? For someone who doesn't know the program, this info seems meaningless.
28. 255: Is this analysis done in retrospect to make up for not measuring sample pH directly after storage (among other things)?
29. 320: What temperature did you use to determine the Schmidt number?
30. 328 f: It is nice to have different temporal resolutions, but what is the purpose of that for this study? Would your measurements not reflect instantaneous fluxes (could maybe be considered as daily fluxes) rather than weekly?
31. 340: A rather general comment to this study: what is the assumption about the development of gas concentration in between times 0, 3w, 3m? E.g., what would the concentration after 2w or 2m supposedly look like? Did you examine that?
32. 362: Would preservation with AgNO3 then be preferable rather than with HgCl2 due to its toxicity? Can you draw conclusions regarding CH4 from your results?
33. 364, Fig. 1: Concentrations of all gases (except CO2) show largest ranges for AgNO3 addition (largest bars) after 3w. Is there an explanation for that? Add to caption: what do the boxplots show, presumably 25th and 75th percentiles and the median?
34. 375 f: Are you arguing that this process is slowed down in freshwater?
35. 380-381: Is this assumption reflected in your results by the decrease of N2 concentration? Is there also an explanation for the N2O consumption following production?
36. 385: Did you perform a statistical test here too? Is it worthwhile mentioning that the concentrations seem to have the opposite response over time than N2O?
37. 422: The opposite of what you state in the text is shown in Tab. 3. Is there a mistake in the labels?
38. 426 f: Wouldn't we expect to see a shift in pH then in Fig. 2 (top panel)? It appears as if the fixed and unfixed samples have the same pH?
39. 435, Tab. 3: What was the reason to show fluxes calculated following Cole and Caraco and not the other wind-based models here?
40. 466: Why was this cut-off of 20 µM chosen?
41. 503: Which of those shown in Tab. 3 and Fig. 4 were obtained from DIC analyses?
42. 507 f: This estimate is only valid for the tested lakes. What would be the implication for other lakes? Is this also valid for sea water samples? Do you have recommendations or a protocol that should be followed? And what about greenhouse gases other than CO2?

Technical corrections:

1. 18: what regulations are there? Or do you mean something like “complex handling” instead of regulation?
2. 49: check brackets
3. 66: use abbreviation DOC
4. 75: the paragraph could be moved to discussion
5. 87: I don't think it's necessary to mention the storage temperature here.
6. 95: determination or rather quantification?
7. 100: replace “gas loss” with “degassing” throughout.
8. 106: I think following the journal's guideline you would want to state what NIVA stands for.
9. 125: silver, not Silver
10. 132: Add name of gas chromatograph. Maybe it would make sense to merge this section with the Gas chromatography section further below, and move some information to the supplement.
11. 193: I think this description is great, but could be partially merged with sections above and equations moved to a separate section in the supplement.
12. 198: pK or K?
13. 205: for completeness state value/equation of K?
14. 206: rather than “given” use “approximated”
15. 208: add altitude “above sea level”
16. 212-214: stick to either air-water or water-atmosphere interface
17. 230: add: in percent
18. 239: without knowing (the power of) this program, I would suggest to move this section or part of it to the supplement, and maybe worth explaining briefly what the program does starting with what PHREEQC stands for. How well does the program perform in predicting variables?
19. 258: thermodynamically
20. 318: double-check equation numbering
21. 357: The concentration of CH4 [across experiments] ranged…
22. 377: Maybe reword to something like: the prevalence of N2O production in [...] was attributed to favoring more acid conditions.
23. 411: samples not sampled
24. 418: indices, not indexes (also change in table)
25. 432: maybe not necessary to mention Lake Lundebyvannet twice in the caption
26. 435, Tab. 3: Correct column labels (preservative-addition and None reversed). Remove one “Lake” in caption. Mention what Diff (%) means. No need to include % in each column. Following the journal’s guidelines, all figures and tables should be denoted with abbreviated Fig. and Tab. Double-check throughout the manuscript.
27. 445: as instead of than
28. 461, Fig. 3: What does i stand for? pH for each sample? I don't think it is needed in the x-axis label then.
29. 499: has instead of would have
30. 503: Fig. S3 shows daily fluxes, not monthly as stated in caption, right?
31. 505: "in reality" is based on samples without fixation? is that feasible? Did you mean to cite Fig. 2 instead of Fig. 3?