The manuscript "Mineral compounds in oak waterlogged archaeological wood and volcanic lake compartments" describes interesting research on the relationship between the elemental composition of waterlogged archaeological wood and the environment where it was buried for a long time. The study is important from the conservation perspective since the presence of various minerals can hinder proper conservation treatments or contribute to further wood degradation.

In general, the manuscript is well-written, the experiments well-planned and conducted, and the results presented clearly and understandably. However, the paper lacks conclusions and I would suggest supplementing this part carefully, considering both the conservation perspective and the environmental aspect.

General comments

The manuscript “Mineral compounds in oak waterlogged archaeological wood and volcanic lake components” provides a good starting point for understanding the mechanisms behind the degradation of waterlogged archaeological wood and its chemical interaction with its surrounding substrate. I found this an extremely interesting read with greater applications to the realm of archaeology. The study is well structured with relevant methods/analyses used.

However, the major issue is the use of EDS, which is fundamental to the study. The authors do not document the methods behind their analysis well enough for me to judge if the data they acquired is accurate. Settings, such as the acceleration voltage used, surface topography, acquisition method (maps vs spot scans) etc should be documented as standard practice, as they can severely impact the quality and error within the data sets (Newbury DE, Ritchie NWM 2013, Shirley & Jarochowska 2022). This is not to say by any means that the chemical composition recorded here is incorrect, but without providing this information, it cannot be evaluated fully or reproduced.

Without the inclusion of this information I would suggest this paper should be reconsidered after revisions and the EDS data properly evaluated. 

Specific comments

The manuscript refers to mineral content and composition throughout, including the title, but EDS does not provide mineral composition; it provides elemental composition and, under some strict assumptions, relative concentrations. Methods that provide mineral composition are e.g. XRD or EBSD. Mineral phases are also not reported in the manuscript so this needs to be corrected throughout.

Line 115 to 120: Clarification needs to be made on the parameters of the EDS analysis as the description is lacking. All user defined parameters have an impact on the data acquired and I would like to see some information. 

1) The acceleration voltage that was used should be documented. This has a direct impact on the reliability of the spectra acquired (Ie Figure 2 b and g - these are spectra, not “charts” and it should be indicated where (spot, line, area) they were taken). The general rule of thumb is that the acceleration voltage (kV) used should be twice that of the activation voltage (keV) of your element. So if we presume that the authors used 20 kV in their analysis, then only elements in the range of 0-10 keV provide reliable results. My gut feeling is the authors may have used 15 kV, which is fine, it just needs to be clarified.

2) The authors state that the samples were coated in gold (at least the thin sections). Was EDS conducted on these samples while they were coated? If so, the activation voltage of Au Kβ is 2.12 keV which, on a gold coated sample, may have some negative impacts on elements with similar activation voltages like P Lα at 2.013 keV,  S Lα at 2.37 keV. Modern systems can adjust for this in post processing and also using higher kV can be really useful, but without the information above, it's hard to say what has been done.

3) Other information, such as if the system was standardized to another material (Newbury and Ritchie 2015), dwell time, working distance are also desirable. 

Line 125: Was the Ash prepared before going into the SEM? It says there was no treatment. Are there BSE images? How were they mounted and were they coated? Also there are (generally) three different forms of EDX measurements 1) Spot scan 2) Line scan and 3) Maps (What I presume is used in Figure 2). What was used here? Each of the above methods has its benefits and pitfalls, with spotscans being generally accepted as providing the best quantitative results. However, Figure 2 leads me to believe that the spectra acquired in panel b and g were acquired from the map data? I think this is fine for this specific use case, but was the same method used for the data from ash? While okay for relative concentrations of elements, topography can have some serious impacts on the quality of data acquired (Shirley & Jarochowska 2022). I would need to see more information on how the data was acquired from the ash samples before commenting on its validity. Some SEM images with the area in which the data was acquired, is one way of achieving this.

Line 127: How were the values normalized? Most systems use the software that is provided with the EDS installation, so I’m presuming because this is Oxford Instruments either Inca or Aztec were used for this calculation. While both are “black box” for these calculations, the complexity of how their numbers are calculated is beyond the scope of the paper here, I would just like the software to be named. It should be noted that the values were acquired under variable pressure (100 kPa is very low) so there will be a lot of interference from residual air and this makes the measurements semi-quantitative at best.

Line 194: I don't think I have access to the supplementary data, but it would be necessary to follow some details, e.g. the use of “variation coefficient” of 82.5% for arsenic (l. 198), which seems very high; do the authors mean coefficient of variation or perhaps just the variance?

Lines 203-206: mean values in the text would be useful to support the text, e.g. when differences between SW and HW are mentioned, what are the actual mean values for individual elements in these two wood types?

Lines 210-215: Were P values also calculated for this regression analysis?.

• “correlations were investigated” - no correlations are reported in the manuscript; perhaps you mean “relationships”? “negative correlation was observed (R2=0.76).” - R2 is not a correlation coefficient and cannot be negative - please sort out the use of correlation/relationship/association throughout the text. Also do mention this is least squares regression, because that’s what Excel calculates.
• Line 211: What do you mean that “no linear regression was present”? I presume that the authors deemed that a R2 value below a threshold meant no relationship? But what was that threshold? A coefficient of determination can be calculated for any set of points; the text should probably say “no relationship was present”.
• Six points seems like very little, do the authors have any more they could work with? Or take multiple measurements from an individual cite on a sample?

Figure 2: I think this figure needs a bit of work before publication. While the overall layout seems fine there are some suggestions to clarify.

• The scale bars in the bottom left corner of each map need to be bigger, in the current version it is very hard to read. They appear to all be the same scale, so one would do for the whole figure. Same for the element names on the bottom right of each panel, either make them bigger or remove them.
• Should figure 2 “l” not be a “j”?
• Panels c,d,e,h,i,l all have very small and hard to read scale bars with no readable scale to represent the number of counts. 
• e has a different color scheme. Are these maps even comparable?
• Figure 2 b and g: the y axis label is a bit confusing and the overall text too small. 
• First I suggest keeping the measurements as whole numbers (20.0 counts to just 20). 
• Secondly I suggest a change is made from Counts [x1.E+3] to “Counts [x1e+3]”,”Counts (Thousands)”, or “Counts x103”.
• Similar to the y axis, the x axis should have larger text and maybe drop the decimal points because here they are unnecessary. 
• Why is there a label to Map 1 and Map 2?
• Panel g What is FeKesc?

Lines 220-224: the text does not agree with Table 1. “total organic carbon ranged from

0.17 to 21.9% while total nitrogen from 0.02 to 1.16 %” - these are different values than reported in Table 1. Is this because Table 1 reports mean or median values? Then it should be made clear and the number of measurements per each category should be shown. 

Tables 1 and 2: “P value represents the level of significance” - no, P value represents the chance of obtaining such a difference by chance, in the absence of differences between compared groups, if the assumptions of the test are met. Levels of test significance are chosen by the users.

What software were the tests carried out in? Indicate the numbers of individual samples in each group and (related) the degrees of freedom.

What does ns stand for? 

Perhaps it would be helpful to spell out the null hypotheses that are tested in both cases and discuss them as such in the text.

Technical corrections

Figure 1 - Well composed and easy to understand figure. It would be nice to be able to link panel b/c are to the panel a with a box eg. (I believe on the right of panel a correlates to b/c?).

Line 55 - The sentences starting with “Sediments composition” and “Furthermore, changes in wood” would benefit from some references.

Figure 3 needs editing: subscript should be used in oxide formulas where appropriate. What are the “in” and “out” suffixes in the X axis labels?

Figure 5: improve readability by changing MWC to Maximum Water Contents in the X axis label. R2 values are too small to read. Are the line equations used for anything in this study? If not, leave them out. 

Line 110 - Should (WAW) be used in place of waterlogged archaeological wood here?

Line 120 - What is the thickness of the sputter coating? And was EDS conducted on the sputter coated sample? The thickness of a gold coating can affect the acquisition of chemical data so it's important to just note that here.

Line 125 - was gold also excluded? I'm guessing the ash was not gold coated, but it's not really clear in the methods.

Line 180: There may be charging of the surface here, it would be good if the authors could just clarify if this is the case or not.

Line 182: Great! Double checking with XRD is a good idea.

Line 287: L-1 should be L-1

Line 334: A bit colloquial when using Anyway here

Line 361: what is pole 144S and how does it differ from pole 144?

Minor language issues, e.g. “The geology of the basin also strong influences the content of minerals and metallic ions in lake sediments.” (l. 79-80), should be e.g. “The geology of the basin also influences the content of minerals and metallic ions in lake sediments strongly.” (adverb, not adjective); similar issues elsewhere in the text