In this study the authors investigate the impact of upwelled low O2/low pH waters on coccolithophore PIC quotas, coccolithophore PIC contribution to the total PIC pool and ultimately what impact this has on the flux of carbon to depth via the PIC:POC ratio in the oxygen minimum zone off central Chile. The general conclusion that this OMZ exhibits comparable PIC concentrations to selected open ocean areas but a greatly reduced PIC:POC ratio in the OMZ core is a nice observation that emerges from the study but the paper needs attention to clarify methodological ambiguities and the occasional lack of precision in reporting results. 

Major comments

Methods: Aspects of the methodologies used need to be clarified. In particular, sections 2.3 and 2.4 are muddled presenting mixed LM and SEM methodologies and unclear reasoning. It is particularly unclear if bulk counting or species/genus level counting was followed consistently and the low number of SEM images examined (with a low magnification) may lead to underestimation of both coccosphere and coccolith counts (see comment below regarding relationship in Figure S4b). Improved description of the methods used is required as this directly leads to ambiguity over how PIC_cocco was estimated (section 2.5) and presented (results). In particular the authors should improve description of the LM and SEM methods used, the apparent bias due to LM counting and the implications of the methodological bias on PIC quotas.

POC data: Though PIC:POC ratios are a central aspect to this study the actual POC data is not presented. The absence of the POC data is a curious omission that is not explained and weakens the study. If at all possible this data should be included and not simply alluded to.

Results: Primarily section 3.1 (but see below). There is a sense that the description of where maximum values were found emphasises Transect T1-T6 and overlooks the broader spatial distributions of parameters along Transect L1-Hyd7, particularly for PIC and coccoliths. Movement of Figure S14 into the main text may help mitigate any uncertainty caused by the unfortunate gap in data along Transect T1-T6 by providing better spatial context and allowing the authors/readers to gauge the representativeness of the in-situ data (i.e. were key spatial features, such as regions with high PIC offshore, missed by the sampling?).

Minor comments

Page 1 Line 23: This statement appears at odds with the authors conclusion that upwelling of nutrients and generation of non-coccolithophore POC may be the cause of the low PIC:POC ratios. The link specifically to low O2/low pH waters is therefore not quite correct.

P3 L79: Please specify the Copernicus product used for Kd490

P3 L83: The definition of the OMZ core was not clear to me and use of the maximum O2 concentration seems counterintuitive when defining the core of the oxygen minimum feature. Please check definition and expand how it was defined.

P3 L91: Please quantify the magnitude of the applied correction made to PIC measurements for Na residues. It is unclear if this is significant.

P3 L92: Typo, presumably this should be ‘PIC concentrations’ not ‘calculations’

P4 L94/95: The terms ‘built’ and ‘building’ seem inappropriate and the intention here is unclear. Please rephrase.

P4 L96: Direct reference to assessing the effect of this OMZ on POC concentrations suggests that the POC data are central to this study and therefore should be presented alongside the PIC data.

P4 L97-99: What criteria were used to define the depth intervals used in this study? As the subsurface interval (5-100 m) presumably crosses the mixed layer there are strong gradients to consider in the distribution of both PIC and coccolithophore diversity which may be lost by the depth bins used.

Figure 1: The figure legend needs a better description of what the black and grey lines actually represent (the odd placement of the plot legend to the right of panel b was initially overlooked). The black and grey lines need to be better distinguished either by changing the line style or line thickness.   

P5 L116/L126: Figure S4b implies a near constant bias exists between light microscopy and SEM methods when estimating coccolith abundances. As the identity of the two axes in Figure S4 are unclear I am presuming that the SEM counts are on the y-axis (and LM counts on the x-axis) in which case SEM coccolith counts are higher than LM counts, even though the general relationship between the two methods is linear. How significant is this bias and what does it mean for the results of this study? (particularly coccolith counts for stations T1-T6 given the reliance upon SEM; Line 116)

P5 L129-133: The approach used to estimate absolute abundances of species/genus coccospheres and coccoliths from SEM images is a little unclear. Based on counts from a low number of images (magnification not reported) I do not understand the rational for multiplying the counts by the total coccosphere or coccolith abundances to obtain total species counts particularly if there are biases between LM and SEM coccolith counts (mainly applicable to stations L1,L2 and L3). Should not the same approach as described in section 2.3 be used i.e. the equation in Diaz-Rosas et al 2021 thereby accounting for volumetric factors? I am concerned that there could be a scaling error here resulting from the use of mixed methodologies. Please clarify.

P5 L137: There is ambiguity in the methodology over whether species/genus level counts were obtained from SEM images and individual species/genus PIC quotas calculated or whether a size-class based approach was used with a mean size based conversion factor.

Figure 2: This is a complex figure to understand which is made harder by the splitting of this figure across 2 pages. The separation of station names (along top of panel 2e) and the lat/lon (along bottom of panel 2n) is unfortunate and this information really needs to be present on both pages to help the reader. Other issues are the difficulty in seeing the white dashed line (euphotic depth) in all panels. Given the reliance upon POC data, why is there no POC section in Figure 2? Key contours or features cited in text should be more clearly visible.

P9 L184-196: There are several unclear statements here that can be clarified. For example i) L185 near surface waters <25m vs L192 surface waters 0-30m; ii) L187 highest PICtotal in 2018 was 5.3 mmol/m3 vs L189 highest PIC in 2018 was 5.86 mmol/m3; iii) apparent bias in emphasising max PICtotal in 2018 as being along Line T1-T6 when station Hyd6 had a higher concentration. Overall, this paragraph was a little muddled and unclear and lacking detail or precision in the reporting of results.

P10 L220: It is not clear where the value of 67% was derived from. I do not see this in Figure 5c?

Figure 3: Panel b gives the impression of monospecific coccolith distributions at many sampled depths due to the approach used of grouping all liths <4um together. This is limitation that needs to be briefly addressed in the discussion.

Figure 4: Figure 4a arguably reproduces some of the data contoured in Figure 2k-n, and presented in Figure S5 so the figures could be simplified. Also, I understand PICcocco to be a bulk term representing the total contribution to PIC by both coccosphere and coccolith PIC, thus it seems wrong to present the contribution of spheres and liths to PICcocco in two subplots with axes reaching 100% in both (fig 4b & 4c). How can both axes be correct when panel 4a indicates a changing contribution by both coccospheres and liths to PICcocco? (evident for station T1,T2 but an unclear contribution by coccospheres for T3-T6).

P13 L235: Typo in legend of figure 5 (concentsdaration)

P13 L238: The phrase ‘marginally higher’ is ambiguous without a quantified value or statistical support. Is the difference significant?

P13 L245: Typo, panel 2c,g ? not 2h?

P13 L249: Typo, panel 2e-f, not 2f-g?

P15 L82: Can remove approximation by stating actual results (45-48%)

P16 L303: Would be useful to state the ratios from Balch et al 1991, Holligan et al 1993b that were used in this comparison.

P16 L303/L306/L343/L411: No need to abbreviate maximum to max.

P16 L305: It is not clear where the stated values of cell-attached coccolith contribution to PICcocco (51-72%) come from. Please clarify and highlight in the results. 

P17 L332: Missing appropriate references (for Calcite Belt, Bay of Biscay)

P19 L373: From Figure 5c I do not see how the statement that up to two-thirds of the PIC_cococ quota comes from detached coccoliths can be correct? Please clarify

P20 L388: The observation that the PIC:POC ratio (Figure 8c) is greatly reduced compared to other areas is intriguing despite the comparable PIC standing stocks (Figure 8b). Without more detail on the coincident POC dataset however it is difficult to rationalise this observation beyond the suggestion put forward by the authors that upwelling stimulates non-calcareous phytoplankton. For this reason, the authors should consider including the POC dataset in this study. It may be particularly important to ascertain the similarity or differences in POC concentrations between the various studies/sites used for comparison to validate the conclusions reached. Also, the PIC:POC results appear most comparable to results from the W. Arctic, which is not an upwelling zone. This point needs to be highlighted. What could be the cause of this similarity?

P22 L413: Conclusion 6 seems to contradict statements on P19 L379. On P19 is the statement that the PIC:POC ratio was significantly elevated in the OMZ core (due to a ballast effect), whereas on L413 the (relatively) low PIC:POC ratio of the OMZ core is highlighted for its difference to other coastal margin areas. Whilst the overall conclusion that this OMZ exhibits lower PIC:POC ratios compared to other locations is valid, it also seems that when examined in detail the core of this OMZ is associated with elevated PIC:POC ratios (Figure 8c), thus the broader significance of this could be addressed in section 4.4.

Table S1: Are the units for PIC concentration incorrect? (i.e. uM not mM?)

Figure S10: The SEM images have reproduced poorly in my copy. Maybe upload the SEM image files separately to allow greater accessibility. Missing word in Figure legend.

Figure S11: These LM images also reproduced poorly. Maybe upload the image files as well.

Figure S14: I find this figure to be a useful means of assessing the spatial variability of PIC during both cruise periods. I would encourage the authors to consider moving this figure into the main text as it provides useful context.

GENERAL COMMENTS

The article by Díaz-Rosas et al. presents an interesting examination of coccolithophore impacts on PIC standing stocks in relation to upwelling and the oxygen minimum off the coast of Chile. The authors compare coccolithophore based estimates of PIC and discrete PIC chemical measurements and show that only 50% of the PIC is sourced from coccolithophores in these waters. There is also a strong gradient in species contributions which align with the base of the euphotic zone and upper part of the oxygen minimum zone. The authors explore these observations in terms of the context of the fate of sinking organic carbon and how reduced PIC ballasting of sinking organic carbon could strengthen low oxygen conditions.

The article is well written and contains interesting observations and measurements, however several issues need to be further clarified. Firstly, the link between PIC production, ocean acidification (pH) and PIC standing stocks, sinking POC ballasting by PIC, and the oxygen minimum zone needs to be better explained in the introduction – at the moment there is no mention of PIC ballasting so that the link between pH conditions and O₂ concentrations is not clear. Secondly, a lot of material is found in the supplementary material rather than the main article and this makes following the article difficult – this is especially in the case where the authors compare their coccolithophore and PIC dynamics with environmental conditions. This is wholly in the supplementary material and if the authors consider this an important element of the paper, this statistical analysis should be in the main article. Lastly, can the authors confirm that their conversions from coccolith CaCO₃ (their Table S3) to the values reported in the article (mmol C m^-3) took into account the molecular weight of CaCO₃ – for most of the article this appears to be so, but then there is a calculation in the discussion which gives very high PIC concentrations (Ln 339) and its not clear how these values were achieved.  

SPECIFIC COMMENTS

Ln 12, Consider whether the term ‘pools’ or quotas or standing stocks (as used elsewhere in the paper) would be better.

Lns 12-13, The links between ocean acidification, PIC concentrations and production, the efficient ballasting of sinking POC, and the impact of remineralisation depth need to be further expanded on in the first couple of lines of the abstract; the links are not obvious without further explanation.

Ln 17, What do the authors mean by ‘countable coccoliths’? Do they mean detached coccoliths, or are they hinting at a portion of coccoliths that is not countable?

Ln 27, What about the important contributions from other calcifying plankton? As the paper concludes that coccolithophore PIC is only around 50% of PIC standing stocks, would it not be good to expand on other potential sources of biological PIC?

Ln 34, There is no mention of the ballast effect on POC sinking in the first paragraph of the introduction so that the link between reduced PIC production and the ‘favouring of the respiration of more organic material’ is not obvious. Suggest adding introduction of the ballast effect earlier in the introduction to make this link obvious.

Ln 43, Consider the use of the term ‘coccolithophore bloom’ in the context of the references given – Beaufort et al. (2008) is based in the subtropical S Pacific waters and did no observe what many would consider a ‘bloom’.

Ln 92 and 105, Can the authors please confirm that all the PIC concentrations presented in the paper are in mmol C m^-3 and that their conversions from the values in Table S3 have taken into account the molecular weight of calcite? Its not obvious that this is a problem or that a mistake has been made but a calculation in the discussion (Lns 337-339) gives much higher PIC concentrations that expected in units of mmol C m^-3.  

Ln 102, Fig. 1 – What are the solid lines on (b) and (c)? As these lines do not line up with the discrete samples, could the legend explain more?

Lns 112 and 118, Having given the proportion of volume counted from the filters for SEM and light microscope analysis (<20 mL or <2% of the volume filtered), it is a surprise that no comment is made on the potential impact this may have on the observations made – specifically the species diversity and the counts of large, often numerically rare species.

Ln 126, What is meant by a ‘good fit’? The statistics of the relationship should be presented in the main article rather than the supplementary material.

Ln 162, What is the oxycline? This needs to be defined somewhere as to what the authors consider either the absolute threshold, relative change or gradient of this feature.

Ln 165, Are the authors referring to fluorescence or Chl-a peaks when discussing ‘peaks of phytoplankton’? Please clarify.  

Ln 166, In Fig. 2f the surface (<25 m) nitrate concentrations look much lower (<5 µM) than the 10 µM quoted in this line. As 10 µM is not generally considered ‘low’ in terms of biologically-limiting, this line should be changed to better reflect (biologically) ‘low’ concentrations (e.g., <2 µM).

Ln 176, Fig. 2 – ODV is notorious for problems with plotting discontinuous (discrete) data in that it often ‘creates’ patterns not supported by the data. The authors should review the patterns shown in the patchier data (e.g., fluorescence, Chl-a, PIC, coccospheres, coccoliths) as to whether they are confidence in the lateral (horizontal) patterns shown where there is no data. The use of the colour scheme should also be reviewed as it is not possible to see the patterns discussed at low relative concentrations in the surface. Also, are the PIC concentrations shown in mmol C m^-3?

Ln 191-192, What does ‘over these abundance ranges, coccospheres and detached coccoliths varied in direct proportion’ mean? Please report the statistics in the main paper rather than the supplementary material.

Ln 195-196, What are figures in the supplementary material (Fig. S5-S6) referenced to support the decline below the euphotic zone in coccosphere and coccolith abundances rather than Fig. 2 in the main text?

Ln 221, Do the authors mean ‘coccoliths were estimated to account for 48% of the total PIC’ or do they mean coccolithophores? It seems from Fig. 5c that the sum of coccospheres and coccoliths is about 50% of the total PIC concentrations rather than just coccoliths.

Ln 237, Section 3.4 – All the examination of coccolithophore and PIC patterns with environmental conditions are in the supplementary material and not the main article. This should be changed as it makes it difficult to see this analysis as part of the main article. Further, Ln 249-250, what do the authors mean by ‘phosphate [being] more limiting than nitrate’? Are the relatively high concentrations of nitrate and phosphate likely to be growth limiting to the cell densities of coccolithophores observed? Also, what about light availability – this doesn’t seem to be a considered environmental condition in this section.

Ln 269-270, Could the higher Gephyrocapsa diversity observed in this study relative to the older studies referenced relate to the methods for species analysis in the different studies quoted (i.e. Hendricks et al., 2012, Venrick, 2012). Specifically, would they have been able to differentiate the weaker calcified species of the genus?

Ln 270, The phrasing of ‘standing stocks of larger taxa were only noticeable below the euphotic zone’ is rather vague and could be interpreted in two ways – either these larger species were absent in the upper ocean and only found at depth, or they made such a small contribution to the total community in the euphotic zone that they were not noticeable. Please rephrase to clarify the meaning.

Ln 275, Do the authors mean PIC contributions from other calcifying phytoplankton (such as?) or other plankton (i.e. foraminifera, pteropods). Reference to the recent study by Ziveri et al. (2023) would seem to fit here.

Ln 283, Daniels et al. (2012) was based in the Bay of Biscay, not coastal waters of Chile. This should be clarified in this use of the citation, currently it could be interpreted as supporting little resuspension of biogenic minerals or river discharge of lithogenic material in coastal waters of Chile. What evidence do the authors have of this negligible input?

Ln 299, What not put the satellite PIC images into the main article (e.g., as part of Fig. 1)?

Ln 304, What do the authors mean by ‘younger populations’ in the context used here? Do they mean in terms of stages of bloom growth and senescence when considering coccolith to coccosphere ratios?

Ln 306, The use of abbreviations (max.) here is confusing and unnecessary. Please use full word.  

Ln 310, Suggest adding ‘in culture’ to ‘inhibits the growth of many coccolithophore strains’ to emphasise that these are not field observations.

Ln 325, What about light availability – surely the 1% of surface irradiance at the base of the euphotic zone is an important limitation for growth?

Lns 337-339, The estimated PIC concentrations of 120 mmol C m^-3 from 400 x10⁶ coccoliths L^-1 sounds too high. Can the authors check their calculations – using a PIC content of one G huxleyi coccolith of 0.025 pmol C (their coccolith value; see Ln 156 and Table S3 converted from PIC to C) only leads to an estimate of 10 mmol C m^-3. To get to 120 mmol C m^-3 the authors would need a coccolith C content of 0.3 pmol C coccolith^-1 (or 3.6 pg C coccolith^-1), which is much higher than the (carbon values) given by Young and Ziveri (2000).  

Ln 369, What depth is meant by ‘subsurface’ when discussing coccolithophore PIC and detached coccoliths? Euphotic zone or deeper?

Ln 404-405, Is ‘the majority of PIC contributed by coccolithophores is dominated byG. huxleyi coccoliths shed during blooms’ a conclusion of the present study? What evidence do the authors present for ‘blooms’ preceding the observations presented?

Ln 411, Please reconsider the use of ‘max.’ in the main text.