Reply on RC1

wildfire regimes in forested peatlands in western Siberia: interaction between peatland moisture conditions and the composition of plant functional types" Feurdean et al. The study explores the interactions between peatland moisture, vegetation composition and fire regime from two peatland sites in Western Siberia. The study provides new and valuable insights of the relationship between the different proxies and the significance of the findings for ongoing and future climate change in boreal peatland ecosystems. Overall, I find the paper well-structured and the multiproxy data clearly presented with sound interpretations of the paleorecords. However, some of the fonts in figures larger to be print-friendly. I am an expert for amoeba and thus my review comments are focussed on the and vegetation records, which align best with my research

R: Thank you for this observation. In the revised version, we have replaced the term fire regime whenever this was incorrectly used.
Chronology: I would like to see more details on the chronology establishment in the main text, for example what material was dated and also why the authors included so many bulk dates instead of picking terrestrial macrofossils for dating purposes. The site description with local Betula growths implies that the record should contain enough datable terrestrial macrofossils for 14C purposes thus it's difficult to understand, why bulk material was dated instead.
R: Although plant macrofossils are the preferred material over bulk for radiocarbon dating, bulk peat has been shown to yield reliable age information in Europe and western Siberia (Holmquist et al., 2016). However, larger differences were observed between dates on bulk peat and wood (see details in Holmquist et al., 2016). In revising this manuscript, we will include additional information on the plant material dated. As for the question of why so many bulk dates, no identifiable plant remains were found in the bottom part of both profiles. For the remaining parts of both profiles, we have alternated radiocarbon dating on plant macrofossils with those on bulk peat as the radiocarbon measurements on plant remains yielded slightly more inconsistent age, as opposed to those on bulk peat that were in consistent chronological order. We will discuss this slightly more in the revised version of the manuscript.
Charcoal morphology: The authors state that they grouped charcoal in woody and nonwoody at both sites based on morphology criteria. However, they also state that the charcoal was classified based on length ratio for one site to achieve an additional classification of graminoids vs leaves/wood. It is unclear why they chose two methods to achieve the ratio for one site while only one method for the other site. This is especially surprising since the two classifications largely overlap for interpretation (graminae vs. woody/leafs and non-woody vs woody).
R: Our standard method to determine charcoal types was via charcoal morphological classification which can identify a series of fuel types (see Fig S3A). At a later date, we additionally tested the length to width ratio method (L:W), which is a more simple way to determine fuel type. The reason that the L:W ratio method was only tested at UC site, is because we did not have peat material left at Rybnaya. In the process of charcoal morphological identification, we often break charcoal particles, thus measuring the length and width on samples already used for charcoal morphological identification would have likely introduced errors. We, therefore, refrained from testing the L:W ratio method at Rybnaya.
Numerical analyses: I am a bit surprised to see how local indicators (watertable, macroscopic charcoal) and regional indicators (pollen) were combined for correlation calculations. Usually, microscopic charcoal >10um is used for regional reconstructions that would align better with the catchment of micron-size pollen. The authors should consider adding a justification why in this case such correlations across spatial scales are considered more useful than combining proxys with a similar catchment estimate? The difference in the geographic scale covered by the individual proxies seems also the most likely explanation why the authors couldn't find some of he statistical correlations they were expecting between watertable, fire and vegetation indicators R: We agree that the spatial scale of the water table and macrocharcoal is smaller than that of pollen. The macrocharcoal curve is strongly influenced by the smaller fraction (150-300 micron), which can have a more regional component (closer to pollen), although the comparison of sub-recent charcoal values with satellite images indicates a localize fire picture recorded by our macrocharcoal record. In revising this paper, we will include the microcharcoal fraction in statistical analysis.
Comparison between recent charcoal and satellite images: In my opinion the first paragraphs are presenting results rather than discussion of the results and should be moved to the results section. For the discussion of the relationship between charcoal based fire reconstructions and satellite-derived fire detection, I would like to see some references to other studies such as e.g. Adolf et al. (2018) or Daniau et al. (2017) R: We agree and have moved parts of this chapter at the Results under a new heading reading: ''Fire-type identification from charcoal, satellite images, and forest statistics'' and kept only a part of the text in the Discussions. References suggested are included in the revised paper. R: R: We have not attempted separation of pollen of etula tree from shrub type. However we have not noticed Betula nana locally at any of the two sites cored.

Specific comments
Line 64-65 Reference needed: When these aerosols persist in the atmosphere, it leads to a medium-term increase in albedo and ultimately to regional cooling.
R: This sentence links to the citation above: Rogers et al., 2015.
Line 65ff: The sentence needs rewording, it currently reads : fire regimes … are surface fires, see general comment above