Bai et al. have analyzed a new set of sediments (2008, 2009, 2012, 2014) and filtered seawater SPM across the Pacific Arctic region using HBI biomarkers, sterols and n-alkanes. The findings of this study are consistent with other similar studies in this particular region and within similar years. These new data points and additional markers do not significantly advance our understanding of HBI synthesis or dynamics in the region, but do validate previously reported observations by Bai (2019) and Koch (2020). However, a notable contribution of this study to the existing body of knowledge is the inclusion of well-defined end members for the H-print index with respect to sea ice conditions, including ice free waters in the Bering Sea near the Aleutians up through perennial ice cover in the Chukchi borderlands and analysing alongside additional sterol biomarkers.

While the results are relatively straightforward, there are a few issues or suggestions to consider further:

• In the Introduction, perhaps it is no longer necessary to provide the extensive explanation of HBIs as it is rather well established in the growing body of literature on HBIs (lines 77-91). Perhaps this could be more consider by simply refer to the sea ice HBIs (IP25 and HBI-II) and the pelagic HBI III, and exclude the information about the history of the biomarker (e.g., initially proposed by Belt in 2007, used in paleo studies by Masse).
• It is unclear why the HBI data needs to be presented twice (once without the Bai et al. 2019 data points and this study). They don’t appear to be all that different. I suggest using the one map to show the different data sources (as done in Fig 1) and only report the whole dataset (n=88) in a figure (Fig 8). A comparison of the two could be better served in a table – and if not that different, perhaps just in the supplemental data?
• Major issue – In reviewing the supplemental data, I fear the H-print has been calculated incorrectly. The “relative abundance” of HBIs is not the value quantified against the standard – which is the absolute abundance. It is the raw data from the SIM chromatograms. It is these relative abundance values that should be used in the H-print calculation. I am happy to connect with the authors to advise on this. I believe this is why the H-print values are much lower than expected in Figs. 4d and 8d for the Bering Sea up through Bering Strait, indicating a dominant sympagic source. Figure 9 will also need to be revised with the new values.
• For Figure 9, there does not appear to be a significant difference between the means of the Chukchi Shelf and slope regions which the authors indicate are distinct.
• In this discussion section, I also recommend looking at and comparing with the Koch et al. 2020 paper from Marine Ecology Progress Series which expands the sediment dataset into 2018 and is presented similarly.
• I like how the biomarkers are presented with sea ice data in Figure 10.
• Line 461 – this was also reported by Koch et al. 2020 through the sediment trap data linking HBI III with the mentioned genera. Apologies for repeatedly referring to this paper but there is a great deal of overlap.
• Line 552 suggests weak benthic-pelagic coupling based on the sterol rations in SPM versus sediment. I’m not convinced by this and the region is known for having strong B-P coupling. Is the use of sterol ratios in this way based on previous work?
• Line 615 in the conclusion states the influence of PDO rather than global warming. This was not tested in this study.

Other editorial suggestions include:

Line 49-50: change to “at the onset of the spring bloom”; remove “still”

Line 52: change “is” to “are”

Line 62: change to “mixotrophic”

Line 78: remove “indeed”

Line 108: suggest using “community composition” rather than phytoplankton structure

Line 283: change to “higher” rather than “high”

Line 294: suggest including Koch et al 2020 a & b (PLOS One and MEPS papers)

Line 310: an issue with the symbol character displaying as a square

Line 515: should be primary production, not primarily

Line 552: change to benthic-pelagic

Bai et al. present a record of 52 surface sediment samples and 13 suspended particulate matter samples from the upper 5m of the water column from the Arctic Ocean, Chukchi Sea and Bering Strait/Sea. In all samples they have analyzed biomarkers for sea ice as well as phytoplankton productivity and terrigenous input, in order to investigate the spatial distribution of these compounds in surface sediments and waters.

The presented record confirms the distribution of the analyzed compounds as already published findings from this area and working group. Bai et al. add the H-Index as a relatively new method to calculate pelagic to sympagic productivity, which nicely underlines the distribution patterns and the applicability of this index.

I apologize, in case I have misinterpreted anything.

As mentioned before, this manuscript presents a valuable dataset, however so far, the discussion focusses mainly on the comparison with Bai et al. 2019, which raises the question of the scientific contribution of this manuscript. The presented data set however is valuable and I am sure that the scientific contribution could be higher when the discussion and data comparison is extended.

I do not see a reason why the Bai et al (2019) data is plotted for comparison but the Koch et al. (2020), Xiao et al. (2015) and Méheust et al. (2015) is not plotted if applicable. This would allow for a much better discussion of the distribution in comparison to sea ice extend and productivity. In this regard, the discussion could benefit from a discussion regarding the different seasonality’s of algae bloom/biomarker production, as the study region reached from high arctic to the middle latitudes, where light availability and seasonality differ greatly.

In this regard, it may be useful to also calculate the PIP25 indices for sea ice reconstructions. I feel the study region is not yet studies in detail since the implementation of HBI III as phytoplankton marker in the PIP25 calculation.

I feel that terrigenous sterol distribution is not given enough attention. Especially in regard to the influence of terrigenous input from sea ice and the influence on phytoplankton productivity.

Specific Comments

Abstract

The Abstract includes a lot of detail, describing the distribution of biomarkers. I feel this could be shortened and less descriptive.

L35 I apologize, if this comment is based on my lack in biological knowledge: What is the connection to pelagic sterols and Chl a with diatoms. Aren’t sterols and Chl a produced by a wide variety of phytoplankton species?

Introduction

L49 “annual bloom period” – To my understanding, there are several blooms (spring, summer, autumn), and this is, also in the Arctic Ocean, until September (sea ice minimum)

L51 increases in later summer/early autumn

L77 exchange diagnostic with geochemical

L88 This is the only time you use a hyphen for sea ice. Make sure you stay consistent.

Material and methods

L179 prior to extraction

L179 delete “with organic solvents”

L207 heating instead of hearing

L240 I strongly disagree with the arguments given by the authors using the sea ice satellite record from 1988 – 2007. Yes, Navarro-Rodriguez et al., (2014) could show that surface records are not explicitly representing a specific time frame. But I would strongly advice against using the old satellite dataset, especially as this study often discussed the most recent changes in the region and the resulting changes in sea ice, meltwater, river run off etc., and the most dramatic sea ice changes occurred after 2007. Why not choose a 20-year time frame going back from the year of sampling (1994-2014)?

L247 A appreciate that the age uncertainty of surface sediments is mentioned here. I see that it is often not applicable to confirm the age of surface sediments - However, this study would greatly benefit from it as it will strongly differ within the presented record, which makes me doubt the comparability of the individual stations.

Results

L307 common sterol found in diatoms

L307 A discussion on the uncertainties of the production of phytoplankton sterols is missing. The producers of brassicasterol and dinosterol are not as straight forwards as presented here and there are more recent publications on this than Volkman, 1986, 2003

L310 there is a problem with special characters, I only see a square throughout the manuscript

Discussion

L400 This combination of these datasets in a figure would allow a much deeper discussion of the region.

L414ff Here, for example, recent changes in the study area are discussed in detail, this is why I see the need to include a more modern sea ice satellite dataset – and also an age control...

L459 The producers of HBI III are not completely confirmed. Please be more cautious here.

L469 “recent decades” – do you have proven that your sediments are representing these? Why are you excluding the recent deca