This manuscript describes the results of several batch experiments performed with the objective of evaluating N2O emissions response to different environmental stressors. The tested stressors were: salinity, pH, temperature, moisture and Zn. The reported results include the measurement of accumulated N2O in the head-space of the reactors for the different tested treatments and the detection of nirS and nosZ genes expression. Also, the N2O to N2 ratio is estimated because acetylene is added in some reactors to stop N2O reduction. The methodological approach is well described and the obtained results and conclusions are clear. The findings allow an improved understanding on the role of environmental perturbations on N2O emissions.

Nevertheless, I would kindly ask the authors to address the following issues:

1. In the present study, three different approaches are employed to study the role of salinity changes on N2O emissions. Instead, a single approach is employed for the other tested parameters (pH, temperature, moisture and Zn). Which is the reason of using different methodologies for different parameters? Which has been the reason of choosing these parameters instead of others (e.g. pesticides)? Also, regarding Zn, why this metal instead of others? All the responses might be related to the study site but it is not clear enough in the manuscript.
2. In the methodology section concerning the agricultural soils tests (2.1), you explain that you tested conventional and organic farming practices. However, nothing more is said about it on the results and discussion sections. Could you explain why? Also, to which type of soil correspond the results you are showing?
3. It is well stated that all experiments were performed under anoxic conditions. However, which is the gas replacing oxygen in the section 2.3 experiments? Why did you use N2 for the experiments described in 2.1 while He was used in the experiments described in 2.2? The same issue is found regarding the gas chromatography methodology employed for N2O measurements for each set of experiments. In 2.1, an Agilent GC is used, in 2.2 it is a Shimadzu GC, while no information is provided for 2.3 experiments.
4. In several sentences along the manuscript (e.g. lines 83 and 160) the authors associate the nirS gene to a nitrate reductase. This is not correct since nirS gene is associated to nitrite reductase (as well as nirK). In this context, why checking the nitrite reductase instead of the nitrate reductase or maybe both?

In the following sentences, you will find some additional specific comment where numbers correspond to the manuscript lines:

1: It should maybe be mentioned in the title that the study focuses on nitrous oxide production from denitrification because other pathways of nitrous oxide production are not considered (e.g. nitrification or chemodenitrification).

46-48: Even I understand it’s not the object of the study, it might be worth it to mention that N2O emissions can also be derived from the nitrification and chemodenitrification processes apart from denitrification. In fact, the contribution of chemodenitrification on N2O emissions is not still well known but could be a significant source. E.g. Robinson et al. (2021) https://doi.org/10.1039/D1EM00222H; Jones et al. (2015) https://doi.org/10.1021/es504862x; Cooper et al. (2003) https://doi.org/10.1128/AEM.69.6.3517-3525.2003 .

51-52: How many decades exactly?

62-63: What about the role of inorganic electron donors such as ferrous iron or sulfide?

68: Due to incomplete denitrification?

71-72: Why?

74: You have already explained the potential role of soil moisture in the previous paragraph. Try to avoid repeated information.

102-171: Even the methodology description is very clear, a table summarizing all tested conditions could be included in the revised version of the manuscript to facilitate following the results and discussion sections.

99: Gram soil/sediment per day --> dry or fresh soil/sediment?

108+125: Units for salinity are missing

119-120: Jars were incubated 12h except for the temperature treatment. Why did this treatment have a different incubation time?

177: For Zn this is not observed for the first 3 points, right?

188: Fig.3 --> Fig.2 ?

190: “above43” --> above

215: I think “moderate” is not enough specific here

231: “do not appear to” or “might”?

248: Only across ecosystem types or also across stressor types?

249-261: Could an isotope labelling approach (15N) be more appropriate than the acetylene block method?

262-277: Concerning the succession of active denitrifying microorganisms, I think it is also worth having a look at the paper published by Liu et al., in 2019 (DOI: 10.3389/fmicb.2018.03208).

284: No pattern or flat pattern?

292-229: In general, the information included in this paragraph seems somehow repetitive with respect to the statements made on the two previous paragraphs. Consider summarizing it.

319: ”nos” --> nosZ ?

Figure 1 to 4: why did you choose to fit quadratic equations? In some cases, it does not reflect the evolution of the measured parameters and induces confusion on the results interpretation.

Figure caption of Figure 1: I don’t understand to what refers the “1” for standard deviation.

Figure 4: Why did you prefer to draw joint instead of separate lines for press and control experiments? Could it be useful to include the nirS results in this Figure?   

Figure 5: I think you should state in the figure caption that this model has been performed according to your results. Maybe you could also check if it fits what has been found by other authors to perform a scheme according to all literature available up to date (including your study)? In plots b and c, following your conclusions maybe the slope at the right side should be softened to clarify adaptation after time.

The manuscript investigates the change of N2O production and denitrification (DNF) in slurries after the physicochemical perturbations artificially produced. In general, the different kind of perturbations induce an increase of the N2O/DNF ratio during about one month. Based on estuarine slurries the authors also demonstrate that communities adapted to changing environment are more adapted to the changes.

The approach seems original and provides important information about the variability of N2O emissions. An important effort is made to keep the result and discussion synthetic in order to focus the reader’s attention on the concepts presented, well summarized in the figure 5.

However, before publication, it seems important to lift three concerns I have concerning the experimental design:

1 – All the rates measured assume linearity of the N2O increase. From my experience, it is a transient species i.e. after a burst of NO3 it is going to increase then decrease. You wrote it increases linearly but did you measure it ? (it seems so for the pulse experiment but no time series is shown in the result) If not how can you justify your assumption?

For the increase to be linear, it is necessary to measure the initial rate, at least when less than a third of the NO3- has been consumed, however when I try to calculate mass budget of N in your experiment, it appears that almost the entire stock of nitrate have been consumed (for the higher rates reported). Can you give a maximum rate measurable for each of your experimental conditions? Did you measure the NO3 changes? Why do you not consider the importance of possible other intermediate species contributing to DNF such as NO2-?

Exp1

20g x 1 umol/g/d x 12h= 10umol

10mL x 1mmol/L = 10 umol

Exp 3

10mL (g?) of slurry diluted by 2 = ~5g x 2 umol/g/d x 12h = 5 umol

10mL x 0,4 mM = 4 umol

2 – The main goal is to understand the effect of an artificial perturbation on N2O production. To this aims the authors compare N2O production of perturbed slurry to a reference. However, the reference itself seems strongly perturbated by the experimental design. In particular, the anoxic conditions produced modify the community structures. Similarly, glucose addition would favor opportunistic species. Have you performed the experiment without these two modifications to obtain a more realistic reference rate measured? Why do you think it is mandatory to use these two perturbating conditions?

3 – In the first experiment, the reference incubation (for ΔT= ΔZinc = Δmoisture = ΔSalinity = ΔpH =0) for the soil treatments present in Figure 1 should correspond to the same slurry. However, very different production rates are reported (from 0 to 0.8 for N2O production and from 0.2 to 2.8 for DNF) which cast serious doubt about the reproducibility of the experiments, and the validity of the experimental design.

Additionally, the methods need much better description of the methodology used and the associated limitation (see below). It could be done in a supplementary file.

I would also appreciate the author to propose a hypothesis about the underline mechanism responsible for such common type of answer: does N2O producer adapt faster than N2 producer? Is it due to thermodynamic barrier?

Details remarks:

Sites characterization:

Why selected the 0-2 cm depth layer? did you check the absence of oxygen and the decrease of NO3  to identify the layer with the most active denitrifying community?

Do you have other ancillary parameters that could give some chemical context:

Do you know the natural NO3 concentration, does it vary between your samples, could that play a role?

Did you characterize by any way the natural heterotrophic activity? (organic matter lability? Oxygen consumption -DBO5 of your slurries? Enzymatic activities?)

Did you measure other elements that could interfere in the N cycle such as redox element (Fe, H2S, …) ? or other N species that could better describe the natural N cycle of your site such as NO2- ?

Incubation Design:

For experiment 3: “the jars were amended to a final concentration of 0.4 mM NO 3- and 0.8 mM glucose weekly” it is not clear if the chemicals were added to increase the slurries concentration by 0.4 and 0.8 mM or to 0.4 and 0.8 mM (which implicate you measured the NO3 and glucose concentration)

For experiment 1 and 2, 10mL of water is mixed with 2g of soil or sediment, where does the water come from?

Is the use of N2 purge to take off O2 could change your community behaviour ? Why do not use Ar instead ? why did you use He for the second set of experiment?

Analytical strategy :

How did you took your aliquots from your jars for N2O analyses? is there any risk of O2 contaminations during this sampling, in particular for pulse experiment where a time serie was performed with He in the headspace (a very volatile gas)? What the introduction of O2 could produce? Did you check the absence of oxygen ?

How long did you store your gas aliquots before measurement ? In which vessel ? what is the detection limit of your instrument ? are you always far above the detection limit?

How did you estimate the mass of sediment? Is it the mass of solid calculated from porosity or the mass of the bulk sediment (ie mainly water)?

Writing:

In figure 2, I recommend to put two black lines in the oligohaline and mesohaline sites to delimitated the daily changes of salinity.

For the title of sections 2.1 and 2.2, I suggest to replace “perturbation” by “Pulse disturbance” in order to keep the same expression along the entire manuscript.