Dear Dr. van Leeuwen,
Thanks for your replies to both reviewers, and to JP Gattuso's comments. As noted in the reviews and as commented on in your replies, your manuscript requires some substantial modifications before it can be accepted. I would thus encourage you to submit the revised version of your manuscript, following the reviewers comments and suggestions. This revision will be subject to potential re-review.
Best regards,
Thank you for submitting your best work to Biogeosciences,
Laurent Bopp

Dear Dr. van Leeuwen,
Your revised manuscript has been seen by one of the original referees. Based on his/her report, and on my own reading of the manuscript and of your responses to the referees comments, I'm glad to accept your manuscript in Biogeosciences after some minor modifications. The referee suggests to be more cautious, thus mentionning "sensitivity of fisheries yield" rather than "potential future fisheries yields". I would advise that you keep your actual title, but mention at front that considering the numerous caveats (that you mention in the paper) these results have to be seen as a sensitivity analysis more than as an actual projection of yields. In addition, the referee has listed a number of minor modifications that I'd like you to take into consideration.
Again, I'd like to thank you for submitting your best work to Biogeosciences,
Best regards,
Laurent Bopp

Minor comments from original referee:
ew minor comments
Line 92: 1DV : what V stays for? Vertical?

Line 93: a space between water and column is missing

section 2.3: Authors point out that mesozooplankton overlap in size with the HTL model. However they do not highlight that also the benthic community represented in ERSEM-BFM partly overlap with the one described by the HTL model (e.g. deposit feeders). While I understand that given the different structure of the model, avoiding both of this overlapping it's really difficult, authors should discuss how this overlapping could impact the outcomes

line 150: what are these calibration factors?

Figure 2: is the model being validated on the annual mean, or the observed dataset is able to describe seasonal variability as well?

Line 163: it is not entirely clear to me what authors mean when they say that that high correlation coefficients reflect “the general size-based structure of the marine ecosystem and the small geographic area”.

Tables 4,5,6: I would suggest to explain in the caption what T and Act. stay for (I assume T is climate only scenario and Act. Stays for all active, but having in the caption would make more immediate to interpret the table

Tables 4, 5,6: my understanding is that the rows “fish” and “detritivore” refer to the changes in the total biomass of fish (and detritivore) simulated by the model, while “fish yield” represents the catches. If my interpretation is correct than I would suggest to change the name of the first column from “fisheries biomass” to “HTL biomass” or similar.

section 3. the title “effects on lower trophic levels” is quite misleading, because it suggests that here only impacts on LTL is described while lines 240-246, 258-265, 285-291 are all about fish and fishery. I understand that authors want to refer to their scenario definition, what about “Upscaling OA impact on nitrification”? Similarly for section 4 where I suggest something on the line of “Upscaling OA impact on detritivore growth”

Line 381: Artioli et al BGD 2013 has been published as Artioli et al., biogeosciences, 2014

line 388: As authors state later in the discussion (line 443-444), changes in circulation patterns is not the only way for global impact of climate change to affect local environment.

figures 3,4,5: I acknowledge that authors decided to keep the current format of the picture. I suggest authors to highlight which difference the bar plots (a and b) are showing. I guess is the difference between 1979-2008 and 2069-2098, however it would be clearer if they write it in the caption for a and b