Overall comments:

The manuscript compares the biological nitrogen fixation (BNF) potential at three polluted peat bogs of central Europe. The topic is of great interest taking into account the key role of BNF in the availability of N in peatlands. These ecosystems could be a source of greenhouse gases, especially disturbed polluted ones, and very few studies have been done in historically highly polluted areas. In addition, they also provide insight into biotic and abiotic controls over BNF.

The manuscript fits well with the SOIL aims and scope. In general, it is very well organised and presented. However, there are several causes of major concern that prevent accepting the paper for publication. In the first instance, from the general perspective, the authors need to build a stronger case to convince the reader of the results obtained from one single BNF measurement in time. Other parameters have been measured over years, but BNF at each site is just one single measurement, and not in-situ but in the laboratory.

In addition, some more specific questions regarding the methodology must be clearly addressed and justified:

• The authors indicate that surface bog water was collected in June 2019 at each study site (lines 175-176). And that Sphagnum mosses and peat were collected in October 2018 (line 180). How can be compared their δ15N value (e.g. lines 322-324) with sampling dates eight months apart?
• Related to the above (may answer it), in Table S3, “Data from October 2018” is for all the data provided by the table? It does not add up with the information provided in the materials and methods section as mentioned before nor in Table S2. It should be clear that the comparisons are among samples collected on the same date, or otherwise justify why they are comparable.
• The authors mention that they collected live Sphagnum and transported it to the laboratory at 6 °C (lines 182-186) and later on they talk about the incubation experiment (lines 190-214). However, several questions arise:
  • How long it took from the collection in the field to the laboratory? And to the incubation?
  • How were the live Sphagnum samples maintained in the laboratory?
  • Was there an acclimatization period before the incubation?
  • What may be the shortcomings (or reasons) of laboratory incubations instead of in-field ones? Have these been considered?
• The authors explain that their laboratory conditions during the incubation period were 16 h day at 18 °C and 8 h night at 10 ° However, they do not justify the reason. Is this setup mimicking real conditions at the time of sampling? Is it just optimal conditions for the BNF process? It needs to be justified and put in context. 
• Line 196. Here it is indicated that the incubation lasted 168 hours, which is 7 days. The authors noted (lines 207-210) that this is a longer incubation than most previous studies, but do not explain the reason. This incubation time should be justified. Here the authors should address what potential errors or shortcomings are associated with such a long incubation, aside from changing headspace concentrations of gases. This is important to explain clearly because literature suggests for this type of BNF measurements, short-term incubations, i.e., less than 4 days (Myrold et al., 1999).

Minor comments:

Line 89: delete the comma after “Zivkovic et al.”

Line 227: It is mentioned an “Appendix I”. I was not able to find it. Was it provided?

Line 310: “MMJ mean of 1.0 wt.” should say “1.1 wt.”

References:

Myrold, D.D., Ruess, W.R., Klug, M.J., 1999. Dinitrogen fixation. In: Robertson, G.P., Coleman, D.C., Bledsoe, C.S., Sollins, P. (Eds.), Standard Soil Methods for Long-Term Ecological Research. Oxford University Press, Oxford, pp. 241–257.

The biological fixation of nitrogen (or BNF) is a very important soil process yet is plagued by large spatial and temporal variability, arising from the large numbers of variables (environmental, biogeochemical and microbial) which can influence the magnitude of the process. It is particularly important in soil systems that have a limited alternative source of nitrogen or those that have been affected by pollution. One such system is peatlands, which are supplied primarily by precipitation, resulting in generally ombrotrophic conditions, and which have also been affected by atmospheric deposition of pollutants such as N and S compounds.

This manuscript is a useful addition to the literature on being able to bring together possible explanations for the variations in the rates of BNF and the manuscript contains an extensive review of the literature which addresses this topic. The primary contribution is to show that three central European peatlands at a high elevation receiving substantial atmospheric deposition of N and containing Sphagnum moss have very different rates of BNF and the study seeks to find why, using two main approaches. One is incubation of Sphagnum moss samples with labeled 15N2 and the second is to use natural abundance variations in the 15N isotope composition of the plant material, water and precipitation. The ‘usual suspects’ controlling BNF are examined with the measurements available, or deduced from alternative sources.

The main conclusion is that one site appears to be affected by a paucity of P and one by a high concentration of SO4, resulting in essentially no BNF, with the third site showing the largest rate of BNF, but without any clear indicator of why, though the weaker knowledge of hydrology at the site may be a factor. The occurrence of methanotrophic bacteria as a component of BNF requires evidence that methane is available in the location where oxidation will occur and incubation of samples with ambient methane concentration is unlikely to identify that source. The laboratory conditions for the BNF assessment were somewhat unusual and ‘one-time’, whereas there are likely substantial variations in field conditions. There is a suggestion that at the BNF-active site, microbes may have adapted to the high atmospheric N loading (from another paper), though it was the same at the other sites.

The natural abundance assessment is complicated because of all the changes in 15N that may be brought about by N transformations, and the presence of N uptake by Sphagnum from N in peat water produced by the mineralization of the peat and litter, and these uncertainties are recognized. On top of this, the 15N sampling at the site with substantial BNF showed a large spatial variability which suggested small-scale variations in BNF, or ‘hot spots’ and possibly ‘hot moments’. One question occurred to me: Sphagnum N concentration was larger at the active-BNF site than the other two (Fig. 5) but the underlying peat (0-10 cm) had a smaller N concentration (Fig. 7). Any reason for that change?

I found the manuscript to be well structured, written and illustrated with a substantial linking to previous studies. It is ‘interdisciplinary’ (as much as ‘disciplines’ still exist), drawing upon atmospheric, biogeochemical and biological controls on the BNF process in an edaphic context. On the whole, though, some of the results are inconclusive because of a lack of measurements to assess all the variables that may affect BNF, but that is the nature of the topic undertaken. I found a few typographical errors, which should be readily correctible.