Paine et al. presents sedimentary Hg signals and associated processes from two lakes over the past 90ka. Overall, the data quality is high, and the manuscript is well written. It is interesting to see such different Hg signals at the same periods between two lakes just ca. 10km apart. I would like to provide some comments or seek for clarifications.

Abstract:

Line 30. I would suggest adding “Hg sources” since different Hg sources also influence the local Hg cycles as discussed in the manuscript.

Introduction:

The introduction section is nice! Paine et al. showed a clear summary diagram on different processes liberating/ mobilizing/ depositing Hg (i.e. Figure 1). I would still suggest adding a few lines on the sources of Hg to the lake in lines 45-53 or in other suitable paragraphs, to help readers better understand why Hg signals could be so different in lakes Prespa and Ohrid.

Line 86. How about changes in catchment? I would suggest adding a phrase about catchment in this sentence.

Site Description:

Lines 143-146. Similar to the question above (i.e. line 86), how would changes of vegetation distribution in the catchment influence the Hg input to the lake sediment over different periods? Prespa lakes have a catchment of ca. 1300 km2. I assume there is a significant Hg contribution from the catchment (e.g., through precipitation and then runoff). If not, please clarify.

Lines 170. Does runoff from catchment belong to the category of “direct precipitation (35%)”?

Chronology:

Lines 220-239. The chronologies are very well established! But I would still suggest briefly mentioning the analytical methods of 14C, 40Ar/39Ar, and ESR in this section or in the supplementary materials, even though some relevant references are already cited. This can provide some pedagogical information to readers on age reconstruction using different techniques.

Mercury Accumulation:

Lines 316-318. Are the methods to calculate sedimentation rates and dry bulk density for Lake Ohrid the same as the ones for Lake Prespa?

Results and Discussion:

Lines 356-364. Table 1 is very well presented.

Lines 380-381 and Figure 4. HgT concentration appears consistently high during MIS2 in both Lake Prespa and Ohrid. Where does the Hg come from?

Lines 393-397. This explanation is quite superficial, even though it makes sense. I would suggest going deeper to find evidence to explain it a bit more. For example, (1) how did the catchment shift regarding vegetation? (2) what can shift the rates of Hg emissions and/or exchange between surface reservoirs? Hg loss by reduction of Hg2+ in lake ecosystems can be very important (e.g., by photoreduction, Jiskra et al., 2021. https://doi.org/10.1021/acsearthspacechem.1c00304 )

Lines 417-419. Very nice!

Lines 427-429. Makes sense! In general, Hg signals recorded in archives are net Hg input, as a result of primary deposition and post-depositional transformation processes.

Lines 535-554. It is not clear to me why Hg accumulation profile in Lake Prespa spanning 10 ka from 33 to 23 ka is much flatter than the one in Lake Ohrid. Why isn’t Hg accumulation elevated in Lake Prespa as the one in Lake Ohrid during this period? Does it link to the shallow characteristic of Lake Prespa or limited Hg input?

Key differences and implications

The whole section is overall well written, but it is lack of some interpretation on Hg loss from my perspective. Hg loss can be very different between these two lakes, therefore affecting the net Hg signals in the sediments. I would suggest adding a few lines on this information to make your interpretation more convincing.  

Supplementary information

It is well presented.

DearEditor,

 I have completed my review. I have few suggestions for the manuscript and no basic criticisms. The main conclusion is that there is no similar pattern for the two lakes and there is no an unifying hypothesis to explain the Hg variability. Different attempt to link the Hg flux to different sources fails to be conclusive for all the periods. This can be frustrating but it is a conclusion. I think, probably the statistic applied at the data is too basic to be able to give a better understanding (if different handling of the data can produce an easier intepretaation), but I’m not so expert to give some further comments on that. I can additionally note that during the Holocene TIC increase and it would be interesting to normalize the data also for TIC for this interval.

Probably, it would be useful to mention in amore direct way tectonic activity as potential source. The two lakes are placed in a very active tectonic zone, a present large gas emission from fault is present not far from lake Ohrid. Difficult to use this as argument but it needs to be considered.

An additional argument not considered (which needs at least to be mentioned) is the dust transport. Loess belt is diffuse in the Mediterranean and there are evidences of leoss deposition also in Macedonia even if not well described.

Minor comments along the text

Pag. 8 The description of the climate is not very convincing.

Line 226 be honest is correct to quote Zanchetta et al. 2018 but also Scaillet et al. 2013 QSR 78, 147-154.

Line 267 Delete zirconium

Lines 346-348 This sentence is very vague. By definition Holocene is an interglacial. You need to specify e.g. increase of forest and so on.

Lines 414-416 This sentence is obscure to me. Probably useless. What do you mean with “reminescent”. I don’t see the importance of this sentence.

Lines 475 I don’t think MIS 3 is considered anymore an interglacial (for a while). I think this sentence should be deleted.

Line 565 delete interglacial.

Lines 590-563. This seems speculative. Do you have evidence of this in the records?

Finally with minor modifications this manuscript can be accepted.