Reviewer comment: As mentioned in the text, 49 Ma corresponds with the possible onset of Northern Component Water. The 49-47 Ma gap in BN silica deposition is intriguing. Witkowski et al. (2020, Fig 5) show a 47 Ma end of widespread chert deposition in the deep North Atlantic. Why would biosilica deposition shift from the shelf to the deep sea? This is discussed more thoroughly in Witkowski et al. (2020), but I believe a review of this discussion is warranted.

Response: The 49-47 Ma gap in BN silica deposition is actually a hiatus, so we are unable to reconstruct the events that took place through this time period. The actual timing of peak chert/porcellanite occurrences in the Atlantic presented in Witkowski et al. (2020, Palaeo3 556: 109896) is slightly different to what the Reviewer is suggesting. Peak chert/porcellanite frequency coincides with the Early Eocene Climatic Optimum, from 53 to 49 Ma rather than the interval up to 47 Ma. Also, in the Palaeo3 paper referred to by the Reviewer, we show that both the siliceous sediments that underwent diagenesis to chert/porcellanite and those sediments that retain their original biosiliceous composition occur along continental margins. Thus, there is really no distinction between “deep-sea” and “shelf” sites in the dataset compiled for the Palaeo3 paper. Finally, considering the high proportion of neritic diatoms in the Blake Nose siliceous plankton assemblages, in the Palaeo3 paper we proposed that neritic siliceous plankton production must have been higher than that in the pelagic settings, even though biosiliceous sediments are sparse on the North American margin. We assume that by “biosilica deposition shift from the shelf to the deep sea” the Reviewer is alluding to the widespread occurrence of diatom-rich sediments in pelagic settings from the Late Eocene onwards. We believe at least part of the reason for this may have been associated with diatom evolutionary events that took place through the Paleogene, including the radiation of holoplanktonic species, as discussed by Sims et al. (2006, Phycologia 45: 361-402).

Reviewer comment: Figure 4 –Biosilca and CaCO3 flux is pulsed in the equatorial Pacific, as would be expected due to changing climate driven productivity gradients. Narrow vs. broad band of equatorial flux.

Response: The lack of a productivity gradient context in the discussion is referred to below, in the response to the Reviewer’s comment regarding line 449.

Reviewer comment: Line 407 –high diatom to radiolarian ratios in BN sediments. What about diatom to sponge spicule ratios? Were these done? Could sponge spicules contributed to biosilica percent.

Response: As explained in the text (lines 423-424), no attempts were made to quantify the contribution of sponge spicules to the reconstructed biogenic opal flux. The reason is chiefly pragmatic: counting and identifying sponge spicules was beyond the scope of both the present study, and the Witkowski et al. (2020, Palaeo3 556: 109896) paper mentioned by the Reviewer, in which Blake Nose radiolarian:diatom ratios were discussed. Secondly, to our knowledge, no quantitative studies on sponge spicules from the Blake Nose cores have been performed thus far, and consequently, we were not able to use published data. Finally, several recent studies, both specifically focusing on sponge contribution to total silica flux globally, and on silica biomineralization, point to declining contribution of sponges to total silica flux that is probably synchronous with diatom expansion in the late Mesozoic (Conley et al., 2017, Frontiers in Marine Science 4: 397; Maldonado et al. 1999, Nature 401: 785-788). Sponge spicules are therefore unlikely to have made a significant contribution to opal flux at Blake Nose through the early Paleogene. A related issue is silica preservation. Silica dissolution occurs in the water column, at the sediment-water interface, and within the sediment, hence the concern as to whether biogenic opal concentrations and/or fluxes have been compromised by silica dissolution (as stated in lines 45-47, deep-time reconstructions of opal flux are founded on one central premise, that opal comprised in the studied sediments has not undergone significant dissolution). Although some attempts have been made (Warnock & Sherer, 2015, Continental Shelf Research 102, 1-8), there is currently no reliable, quantitative measure of siliceous plankton preservation in sediments, and the basic indicators of silica dissolution are chert/porcellanite and zeolite (clinoptilolite) occurrences. We have stressed in the text (lines 146-147, 297-298 in the original submission) that chert occurrences are rare in the Blake Nose cores, and that in some intervals diatom preservation can be considered pristine (lines 300-301). We therefore assume that extensive diatom silica dissolution has not occurred, which would lead to preferential preservation of the more dissolution-resistant sponge spicule silica over the more dissolution-prone diatom and/or radiolarian silica. Thus, our conclusion is that a significant contribution of sponge spicule silica to total opal flux in the Blake Nose cores is unlikely. As explained above, the best we can offer is to put still more emphasis on this issue in the revised manuscript, for instance by including the explanations given above in an expanded version of the paragraph starting in line 407.

Reviewer comment: Line 449 –diminished BN silica fluxes at 42-38 Ma –imply diminished nutrient supply? How about increased productivity gradients? Explained beginning of line 470

Response: Indeed, the discussion on the reasons for Atlantic to Pacific biogenic opal flux fractionation included in the original submission is focused predominantly on seafloor biogenic opal preservation – but, as explained in the response to the previous Reviewer comment, preservation is only one of the key factors that could potentially compromise an opal flux reconstructions. The Reviewer is right that productivity context is not presented sufficiently well, and that a productivity gradient should also be discussed. As such, discussion of possible mechanisms for altering biogenic opal flux in the Atlantic and Pacific will be better incorporated into the revised manuscript.

Reviewer comment: Line 78 –flat continental relief –Does this imply reduced rates of terrestrial runoff?

Response: Yes, Misra & Froelich (2012, Science 353: 818-823) and Froelich & Misra (2015, Oceanography 27: 36-49) invoke reduced rates of terrestrial runoff (due to inferred flat continental relief) to explain the broad low in lithium and strontium isotope records spanning the late Paleocene and early Eocene subepochs.

Reviewer comment: In general, it would be helpful for the authors to discuss in more detail the conditions in which diatom productivity is limited by silicon rather than another nutrient(s). As the authors point out, productivity can be limited by silicon supply from weathering, supply from circulation/upwelling, OR the (lack of) other nutrients. It would also be helpful to see a discussion of whether/how we know that diatom productivity is limited in the long term/large scale by silicon supply from weathering, rather than by a combination of these other factors.

Response: This issue is actually the focus of another paper in preparation by the lead author. As already remarked in response to the comments by the first reviewer, a broader discussion on factors limiting diatom productivity will be included in the revised manuscript, as requested here.

Reviewer comment: Figure 2. Please label/identify the fit lines (assuming that’s what those are) in the figure caption.

Response: The heavy red lines represent smoothed opal concentration (Figure 2A) and opal flux (Figure 2B) curves. These were derived from LOESS regression. By mistake, the span of the LOESS smoothing procedure was not included in the relevant figure caption. The figure caption will be updated accordingly in the revised manuscript.

Reviewer comment: Section 3.2 How significant are these correlative relationships given that the bioSiO2 fluxes are based on 1) smoothed sedimentation rates and 2) interpolated DBD measurements? I’m skeptical that these mean much; the authors could certainly make a case for why they do, but they haven’t done so thus far.

Response: Our presentation of the opal fluxes follows the convention adopted by Piela et al. (2012, Paleoceanography 27: PA2204). As explained in the text, the reason for smoothing sedimentation rates was to eliminate unrealistic, high-amplitude variations in sedimentation rates between consecutive age model tie-points. The focus of the study is on long-term rather than short-term trends. We argue smoothed datasets are of sufficient resolution for these purposes (i.e. in accurately reconstructing long-term trends in opal accumulation). The robustness of this approach, despite being based on smoothed and interpolated data, is best demonstrated by the multiple regression model, which reproduces the smoothed composite opal flux curve reasonably well. Unfortunately, by mistake, we have not included the multiple regression model (line 325) output in Figure 2B, despite a reference being made in the text in line 327. This omission will be rectified in the revised version of the manuscript. We will also include a comment on the suitability of smoothed datasets for the purposes of our study.

Reviewer comment: Figure 3. These axis labels won’t make sense to many readers – what do the numbers represent?

Response: We agree with the Reviewer. The axes should be labelled more clearly in the revised manuscript.

Reviewer comment: Figure 4. Some of these colors are quite hard to see, especially the light yellows and purples. There are a lot of different things to distinguish, but perhaps these can be a bit darker, or changed for different colors?

Response: We agree with the Reviewer, and intend to differentiate between the colors and line styles in a clearer manner in the revised manuscript.

Reviewer comment: Lines 460-465  The previous paragraph attributes some flux changes to changes in preservation of bioSiO2. Can that be ruled out, or at least shown to be insignificant as it relates to the changes described in this paragraph and the next? If so, the authors should make that case, and if not, should discuss the implications for their interpretations.

Response: A concern regarding preservational issues has also been expressed by the other reviewer. We agree that dissolution could potentially compromise the flux record. We repeat the relevant part of the answer to the other reviewer’s comment: silica dissolution occurs in the water column, at the sediment-water interface, and within the sediment, hence the concern whether biogenic opal concentrations and/or fluxes have been compromised by silica dissolution (as stated in lines 45-47, deep-time reconstructions of opal flux are founded on the central premise that opal preserved in the studied sediments has not undergone significant dissolution). Although some attempts have been made (Warnock & Scherer 2015, Continental Shelf Research 102: 1-8), there is currently no reliable, quantitative measure of siliceous plankton preservation in sediments, and the basic indicators of silica dissolution are chert/porcellanite and zeolite (clinoptilolite) occurrences. We have stressed in the text (lines 146-147, 297-298 in the original submission) that these are few in the Blake Nose cores, and that in some intervals diatom preservation can be considered pristine (lines 300-301): hence the assumption that no extensive diatom silica dissolution has occurred, which would lead to preferential preservation of the more dissolution-resistant sponge spicule silica over the more dissolution-prone diatom and/or radiolarian silica.

As both reviewers have voiced similar concerns, we will further expand the discussion on possible silica preservation-related issues in the revised manuscript.

Reviewer comment: Last paragraph of section 4.3  This last sentence highlights the difficulty of using bioSiO2 flux to test the effects of the weathering feedback – changes in flux are just as likely to reflect a change in ocean circulation as they are the weathering feedback. Could the authors try to disentangle these things a little more in this last section?

Response: Further insights into the relationships between opal flux, ocean circulation and weathering feedback could be gained by comparing the opal flux records included in the original submission with published silicon isotope records. A composite d³⁰Si record from the Blake Nose was published by Fontorbe et al. (2016, Earth and Planetary Science Letters 453: 67-77). A comparison between the Blake Nose opal fluxes and d³⁰Si from Fontorbe et al. (2016) will be included in the revised submission.