Adding the information on PAR is a good idea, and I have done so in a revised version of the data set.  Unfortunately not all data are acdessible at this time, but will be added as soon as possible.  The data added are integrated PAR values derived from a BioSpherical sensor placed as close to the incubator as possible, and so represent the irradiance received by the samples and NOT the irradiance at the location of sampling.  All files listing these data at the BCO-DMO data site have been updated as of 2021-12-2 and will be further updated as additional archived data are accessed.

I agree that the data are valuable not only for bio-optical models but for ecological research in the Ross Sea.  I have added a statement to that effect in the manuscript (Line 248).

Line 85: I agree that to normalize the responses to photoperiods would be difficult and even misleading, but the PAR data were requested by Reviewer 1 and added to the data files.  It is of potential use in looking at effects of photoinhibition, but I did not pursue that line of investigation in this report. 

Line 156:  The anomalous values were carefully scrutinized prior to removal.  They were confined to two stations (three depths from each station), and there may have been a duplicate inoculation of radioisotope by mistake.  We don’t know that for certain, but it was clear that the ¹⁴C-uptake rates were much greater than would have been expected.  There was no relationship to bloom development or decline.

Line 199:  Corrected.

Line 211:  Corrected.

Line 211: Statistical tests and significance values now included.

Figure 4: Variations within each bin were not analyzed due to the variations in the number of data points within each bin.  Those numbers are now mentioned in the figure caption.  I did briefly look at the variations within the bin that had the most samples, and the variability in that bin was similar to the variability over a longer time span (e.g., weeks).  A complete assessment of the within-period variability is not possible.

Responses to Reviewer 4

I'd like to thank the reviewer for his/her comments. 

1. Line 50: You mentioned that “long (e.g., 24 h) measurements approximate net production, but this has not been rigorously examined in polar systems.” Have you tried to address this issue in your dataset? Does “primary productivity” in the paper mean the “net productivity” or the “gross productivity”?

I never tried to thoroughly understand and investigate the “net vs. gross” issue.  One of my colleagues in collecting some of these data, John Marra, looked at this issue more than anyone else, and concluded that 24-h incubations closely approximate net production.  Bender et al. (2000) used multiple measures of net and gross production (oxygen isotopes, nitrate inventories) and found that in 24-h incubations “high ratios of net/gross production occurred” which suggest that net production was being approached with this incubation length.   While there is undoubtedly variability in this relationship, on the whole publications on this issue strongly imply that a 24-h measurement approaches net production. 

   I have now included some of the literature that has addressed this in the revised manuscript (lines 52-53).

2.    According to table 1, your data collection sites in different periods are inconsistent. Could the year of measurement of these points be somehow reflected in Fig. 1? Will the distribution of data collection sites across different periods bring bias to the seasonal dynamic analysis (e.g., Fig. 4)? The reviewer is correct in that most of the data were collected without regard to date, and indeed, many of the studies had objectives that were linked to other oceanographic questions rather than productivity.  I tried to graphically convey the temporal (both seasonal and yearly) aspects of the station locations, but concluded that it simply became “too messy” and that the overall information on sampling locations was largely adequate.  I do note, and have added to the manuscript text (lines 81-85), that the seasonal differences are greater in magnitude than any observed interannual differences.

3:   Why not consider using “Julian Date” for the x-axis? I was a bit surprised at this suggestion, as Julian Date is to me less intuitive than the calendar date.  But it also is more quantitative, which is why I included it in the data set.  As a compromise, I have added Julian Date and the calendar date to both Figures 3 and 4. 

4.  Line 165: I think that photoinhibition is related to the photoprotection level of plankton which can be species-related. Therefore, I suggest mentioning the dominant phytoplankton groups and their photoprotection abilities here. Again, correct. While most studies on photoprotection have been done on natural assemblages, a number have investigated those in specific species and functional groups. In short, there are a variety of responses that can and do occur that are relevant to the Ross Sea. I have added some discussion and published literature reports on functional group responses to the manuscript (lines 176-184).

Responses to Heather Bouman

Thank you for your comments, especially since the paper was modeled after Bouman et al. (2018) in ESSD!  Most are easy to respond to, and below is the listing of changes I have made.

1. 18: “/” changed to “-“
2. General comment re italics: In the version I am looking at Phaeocystis and antarctica are all italicized. I can only imagine that during pdf formation those were removed, but prior to submission of the revision, I will check those carefully. Similarly, all exponents were checked and all were correct in the version I have. 
3. “Biogeochemical models” was added (now L. 77).
4. Information is now provided on the types of bottles used as well as the blue filters used (L. 90-91).
5. HPLC pigment data are not available for all cruises, but the various data sets referenced present HPLC data when available. A statement has been included (L. 152-154) that states that while phaeopigments are present, they are in very low concentrations (suggesting that microzooplankton grazing is minimal, although this is not mentioned in the manuscript), and that fluorometric and HPLC chlorophyll a values are highly correlated with a slope near 1. 
6. Heather is correct in that chlorophyll c₃ is not simply confined to diatoms. However, in the Ross Sea its presence is largely confined to diatoms that occur there (coccolithophorids do not occur in the Ross Sea), making it a reliable pigment to separate functional groups, especially when used with other HPLC pigments.  The statement has been modified (L. 149) to reflect this more accurately.
7. The statement about the comparison of assimilation numbers and has been hopefully clarified (L. 161-163). 
8. A revised statement concerning the blue filters (L. 90-91) now indicates that the same filters were used for all productivity determinations as used in controlled P-E incubations.
9. Sadly, this was not a typesetting issue, but a terrible error on my part. It has now been corrected.
10. The standard deviations were not plotted initially because they were large, and the plot became a bit unmanageable and the vertical trends obscured. Heather’s comment is correct; the standard deviations were listed in Table 3.  In this revision the standard errors in Figure 2 are removed and only the means plotted, retaining the clarity for the vertical information.  The standard deviations are retained in Table 3 but are also mentioned in the figure caption for Figure 2 as well.  An additional comment is included in the text (L. 188-189) that emphasizes the large variability in the chlorophyll-normalized rates of carbon fixation.

This review was the same as CC1; please see the review response therein.