Please see attached pdf for paper comments.

Overall

This work presents data on greenhouse gas concentrations and estimated fluxes, along with temperature, oxygen, and chlorophyll data, from a eutrophic shallow lake over one season (April / May – October 2020). The topic is timely as the scientific community is working to reduce uncertainty around aquatic greenhouse gas emission estimates, especially from shallow systems. While I found the research interesting, I have a few overarching concerns:

1. The greenhouse gas data have been previously published (Søndergaard et al. 2023), which already describes the novel results of this paper: that temporary stratification events lead to the buildup of greenhouse gases, which are likely released upon turnover. I think the authors can more clearly describe how this work is different from the previously published paper-- I suspect the new addition is that this paper estimates the ebullitive and turnover fluxes, which I address next.
2. Ebullitive fluxes were measured using floating chambers. I have not seen chambers used to estimate ebullitive fluxes before, unless the chambers were set for a short amount of time and concentrations were measured repeatedly (e.g., using a portable gas analyzer or manual sampling)—I’ve seen this done up to 24 hours. Then over that short amount of time, the diffusive and ebullitive fluxes are teased apart. In the current study, the chambers were deployed for 2 weeks at a time, during which the chambers would have equilibrated with the water, with the exception for bubbles. I do not think it is possible to determine the total amount of ebullition with this approach. For instance, if a bubble occurred on Day 1, the CH4 could diffuse back into the water column by the time the chambers were checked two weeks later. While I am empathetic to the challenges of measuring ebullition, I do not agree with the authors that this is the “least worst method available.”
3. Turnover fluxes assume that all the CO2 and CH4 gases in the hypolimnion were released when the lake mixed. This approach assumes that there was no CH4 oxidation during turnover, which contrasts previous studies (see Kankaala et al. 2007, Thottathil et al. 2019, Zimmerman et al. 2021). If the lake mixes rapidly, oxidation may be low; however, previously published data on this lake from the same year (Søndergaard et al. 2023) shows that while thermal mixing can occur within hours, it can take 5 days for the complete mixing of oxygen. As thermoclines and oxyclines are offset in this lake (and this may be a common phenomenon, e.g. Gray et al. 2020), I don’t think it’s fair to assume oxidation is 0. Therefore, without oxidation estimates, these turnover values may be huge overestimates.

Specific Comments

Abstract

• Lines 18-19: provide details of length of the study (e.g., May to October for GHGs)

Introduction

• Broad framework of the Introduction is focused on climate change, but this is not a climate change study. While climate change will likely change the mixing regimes of lakes and ponds, it is not the major focus of this study. The novelty of this study seems to be that intermittently mixing lakes have unique biogeochemical cycles, and the oxic-anoxic cycles may explain the variability in fluxes over time. The challenge is that this story is also the framework for the Søndergaard et al. 2023 paper, which used the same dataset.
• Line 40: Provide more details on the % contribution coming from lakes and ponds instead of saying “large proportion”
• Line 53: See Deemer and Holgerson 2021 on the drivers of diffusive and ebullitive fluxes
• Lines 62-66: See Holgerson et al. 2022, which describes mixing regimes in shallow waterbodies including this category of intermittent or temporary stratification.
• Lines 71-73: The discussion of C burial is interesting but a bit of a red herring as it is not something addressed in this paper
• Lines 76-77: How is this specific to shallow lakes?

Methods

• Lines 90-98: Provide overview of the study time start and end in the first few paragraphs—the whole season study is a major strength of this study.
• Line 99: Was a solar shield used for measuring air temperature?
• Lines 113-116: Stratification should be defined by density instead of temperature because the density-temperature relationship is not linear, and water density is what determines stratification. See Gray et al. 2020 for more details on the importance of using density over temperature. Especially considering this study includes measurements from May – October, the temperature range is large and the density-temperature relationship becomes more important
• Lines 113-116: I find it confusing to determine stratification periods by both temperature and oxygen considering the thermocline and the oxycline set up at different time scales (e.g., Søndergaard et al. 2023). I recommend just using density differences.
• Lines 115-116: The statement that bottom waters remain undisturbed is an assumption—partial mixing events likely increase turbulence at the surface of the hypolimnion and gases can be exchanged.
• Lines 148-151: How far away was windspeed measured? Were the on-lake conditions compared to the institute’s measurements?
• Lines 168-170: Why is oxygen used to determine the hypolimnion here, whereas mixing was previously defined based on temperature and oxygen?
• Lines 170-171: As described above, I do not think the authors should assume oxidation is 0 when it may take 5 days for oxygen concentrations to equilibrate following isothermal conditions.
• Lines 177-178: Floating chambers need further description (surface area, volume).
• Lines 177-188: See above concerns about estimating ebullition from chambers deployed for two weeks.
• Lines 201-203: I appreciate the caveats associated with the floating chambers, but as described above, I need more convincing that these methods are appropriate. Why not measure volume displaced and collect fresh bubbles? How do the methane concentrations in the chambers compare to fresh bubbles?

Results

• Lines 218-219: Use more quantitative descriptions of time mixed vs. stratified. If you use the definition for mixed vs. stratified described in the methods (see critique on not using density), this will allow for quantifying mixed vs. stratified periods for broad summary.
• Figure 2: I recommend using the same gray backgrounds to show periods of thermal mixing—this will help highlight the offset between isothermal conditions and oxygen.
• Please use statistical tests and present the results on concentrations during mixed vs. stratified periods; e.g., lines 246-248.
• Table 1: This appears to be right from Søndergaard et al. 2023. I recommend removing it and describing in the Methods, or providing standard error or standard deviation for the 2020 season. TN:TP would be more helpful in its molar ratio to examine nutrient limitation.
• Table 2: Statistical comparisons needed to compare mixed vs. stratified periods
• Figure 3: Add where the bottom samples were taken from—Station 3?
• Figure 9: Make sure the arrows match statistical differences observed.

Discussion

• Line 294: provide number instead of saying “massive”
• Lines 295-297: provide statistical comparisons
• Lines 306-308: The partial mixing here contrasts the assumption in the introduction that bottom waters were not affected by partial mixing events
• Lines 330-335: As described above, I still need convincing that the floating chamber method is appropriate to estimate ebullition.
• Line 345: Explain why it’s an overestimate (i.e., oxidation)
• I think subheadings could help organize the discussion
• The conclusion doesn’t tie back to the introduction framework focused on climate change, which again suggests that the Introduction should instead focus on variable mixing regimes in shallow lakes and the consequences for biogeochemical cycling.

Minor Comments Not Requiring Response

• Line 27: add “for” between “need high”
• Line 48: add comma after “approach”
• Line 434: remove “crack” as in English, it often references cocaine, which I do not think is the intent here.