Dear authors,
Thank you for posting your responses to the reviewers' comments. Both reviewers concluded that your study is of interest for BG but needs major revisions, particularly regarding justification of the experimental design and choices of data analysis. I invite you now to submit a revised manuscript that takes into account these comments. The revised ms will be sent to both reviewers for further examination.
Best regards,
Emilio Marañón

Dear authors,
Thank you for submitting a revised version of your manuscript. One of the original reviewers has examined it and considers that most comments and suggestions have been dealt with adequately. However, this reviewer lists numerous issues that still require attention before publication can be recommended. I also include below several points that need to be considered. Some of them are important, in particular those that concern 1) discussion of the implications of not having nutrient data, 2) interpretation of potential existence of feedback mechanisms, 3) comparison of properties between initial and final sampling date and 4) explanation of the antagonistic effect between high pCO2 and warm temperature.
I now invite you to address these comments in a revised version of your manuscript.
Thank you for submitting your work to Biogeosciences.

Best regards,
Emilio Marañón

Editorial comments on BG-2017-510

- As indicated by reviewer 1, the Discussion must consider the implications of the lack of information on nutrient concentration. The text should explicitly mention that differences in nutrient availability may have contributed to observed differences between control and high temperature and high CO2 treatments. The whole experiment rests on the assumption that initial nutrient concentrations were the same in every experimental bottle, but this assumption has not been verified. Assuming this was the case, and to the extent that differences in biomass were observed between treatments, differences in nutrient concentration must have arisen as well during the experiment.

- Title must be changed to make it clear that what is being studied is a bloom induced experimentally in vitro.

- Abstract line First sentence should clarify that the article deals with an experimentally induced bloom in vitro.

- In the Abstract, and throughout the manuscript, changes in properties as a result of experimental treatments must be described in terms of the difference between treatment and control, not between treatment and initial value. For instance, lines 16-17 read: ‘total phytoplankton biomass was significantly increased by elevated pCO2 (20-fold) and (…) biomass also increased under elevated temperature (15-fold)’. This passage suggests that high pCO2 and warm temperature induced very large increases in biomass, when in fact those increases (20- and 15-fold) refer to the difference between final and initial values. Given that biomass increased also in the control bottles, the relevant comparison is between treatments and control. The same problem applies to the later reference to the 30-fold increase in Prorocentrum cordatum.

Line 19 Add ‘and in the control’: ‘Throughout the experiment in all treatments and in the control…’

Lines 107-109 Note that pre-screening through 200-um mesh removes only the mesozooplankton. The microzooplankton (quantitatively more relevant, in terms of grazing pressure) are still present in the experimental bottles (see comment 2 by reviewer 2).

- The assessement of chla-specific productivity is based on a indirect method (FRRF), which, as the authors acknowledge, is based on numerous untested assumptions. In addition, FRRF measurements were conducted on a single occasion in a experiment that lasted >30 days. This represents a tenuous base to make a major conclusion of the observation that the combined high pCO2 and high temperature treatment causes no change in chla-normalised productivity.

- Related to the point above, and as mentioned also by reviewer 1, the lack of effect of the combinated high pCO2 and high temperature treatment upon productivity is by no means an instance of negative feedback (as stated at the end of the Abstract and in the Discussion). In the context of CO2-induced climate change, an example of negative feedback would be an increase in marine productivity, which would contribute to counterbalance the original disturbance (higher pCO2). If, as a result of climate change, productivity decreases, then a positive feedback would be occurring. But if the environmental changes considered (in this case, temperature and pCO2 combined) cause no effect on productivity, there is no feedback to speak of.

Lines 477-478 Many species have optimum temperatures higher than 20°C (Boyd et al. 2013 PlosOne, Chen 2015 J Plankton Res)

Lines 498-499 The logic behind this sentence is unclear: ‘This may explain the lower PBm in the combined treatment compared to elevated pCO2 and temperature individually’. If dinoflagellates have weaker CCMs compared to diatoms, how does that explain the fact that at the end of the experiment PBm is smaller in the high CO2 treatment, where DIC is more abundant? How can you connect a bulk property (PBm) with the abundance of dinoflagellates, which contribute a minor fraction (<20%) of all phytoplankton? And how can you explain the effect of temperature? The explanations of the antagonistic effect of temperature and pCO2 on phytoplankton biomass and photosynthesis are confusing. This antagonistic effect is quite puzzling and, without trying to put forward speculative mechanisms, the authors might just acknowledge that additional studies are required to figure this out.

Lines 493-494 Such high pH values denote strong consumption of DIC which is associated with nutrient depletion – hence slow growth rates likely result from nutrient limitation

Line 560 Again, the sign of the potential feedback mechanism is wrong. If high CO2 leads to higher photosynthesis and more CO2 uptake by the ocean, this represents a negative feedback, not a positive one. It is a negative feedback because the outcome contributes to counterbalance or neutralize the original disturbance (hence it is a stabilising mechanism).

Line 581 There is no reduction, if productivity in the combined high CO2 and warm temperature treatment was the same as in the control.

Figure 3, Table 1, also section 3.4. The POC:PON ratio cannot have units of mg m-3. This ratio should be computed with molar concentrations, so that its units would be molC:molN.

The text should be revised carefully for punctuation (missing spaces, missing periods, etc).

Dear authors,
Thank you for submitting a revised version of your manuscript. You have addressed adequately all the points that had been raised. However, there seems to be a problem with the reported values of the C:N ratio (Fig. 3F). Given the values of POC and PON concentration, the molar C:N ratio should range roughly between 10-16. Please correct both the figure and the text describing it (lines 326-331). Also, the units of C:N could be simplified to the more commonly used molC:molN.
Best regards,
Emilio Marañón