Thank you for the opportunity to review the manuscript titled “Role of phosphorus in the seasonal deoxygenation of the East China Sea shelf.” I think the manuscript addresses an important problem and has the potential of becoming a worthwhile contribution. The approaches are robust, and the results are elegantly presented. I have offered below a couple of comments, which hopefully can improve the manuscript.

1. Introduction: Suggest adding some more background/literature information on the relative importance of P limitation, N limitation, and light limitation in coastal environments, including literature specific to the Changjiang Estuary. This can help the readers set up expectations and better appreciate the results presented in the later sections.

2. The entire modeling analysis and results are focused on DIN and DIP. What are the roles of other subspecies of N and P? I am not suggesting the authors to re-do the work with TN and TP, but some discussion on the relative importance of DIN (DIP) vs. non-dissolved N (P) can be beneficial. This is especially true for the proposed N+P dual nutrient reduction (28% and 44%).  A 28% reduction of TN is not equivalent to a 28% reduction of DIN. The same argument is true for P. 

3. The analysis of loading reduction scenarios makes the paper stronger and more management relevant. One suggestion is to present the loading reductions in the context of major sources for this system. What are the main contributors of riverine TN and TP loads? How would the proposed reductions be achieved by management actions by targeting those major sources? Some recommendations based on literature and/or the authors’ experience with the watershed will be helpful. 

4. The majority of the analysis is based on established models. I do not have any issues with the modeling framework, but some statements on the model assumptions and uncertainties are recommended. In addition, have there been any bioassay sampling to verify the model-derived N or limitation status?

5. For the correlation result (Section 3.2), please specify what correlation method was used and justify the choice. 

6. This research is comprehensive and should have broad relevance. Some discussion that compares the results with other systems or transfer the insights to other systems will be very helpful.

7. Figure 1: How were the six zones determined. Please clarify in the caption and methods. 

8. Please be mindful on the use of significant digits. For example, 68.0 ×10³ km2 (Line 155), 75 x10⁸ mol N (line 336), and other occurrences. 

The paper investigated the impact of P limitation on hypoxic zone in the East China Sea using a coupled physical-biogeochemical model. The results shown spatio-temporal variability of P-limitation in this region in details, and provided mechanism explanation on how hypoxic region was modified by N+P, P-only, N-only scenarios from primary production to water column respiration. It also provided management suggestions that N+P reduction was the best strategy to mitigate hypoxia, and quantitative measure on how much hypoxic area would be reduced for intermediate and long-term reduction in N+P reduction.

Modelling hypoxia in China coastal ocean is a very hot topic these days. Most of them focused on mechanism explanation. The research had a very special perspective on hypoxia management strategy, which was great and important for hypoxia study in this region. The paper is very well written and organized. It should be published after minor revision.

1) Figure 1: Why zone 1 to 6 was separated like this? Any standard?

2) In results section 3.3. It mentioned the limitation factor is < 0.85. It should be mentioned in the method section, how this limitation factor is defined and why, although a citation paper has been provided. It will be wonderful to include those crucial information without go back and forth to other references

3) In the results section and Figures, except comparison for the nutrients. A comparison for surface chlorophyll concentration and salinity with satellite data, and hypoxic zone with cruise data overall should also be provided.

4) Line 270-275: A comparison with the Gulf of Mexico has been mentioned. A recent study on P-limitation on the Pearl River Estuary system. “Reversing impact of phytoplankton phosphorus limitation on coastal hypoxia due to interacting changes in surface production and shoreward bottom oxygen influx” by Yu et al. 2022 (Water research) should also be mentioned. What is the difference and similarity between the three system? Any relation with the land cover and land use change in China? (More urbanization, sewage discharge)?

5) In the code availability part. Only general ROMS code downloaded was mentioned, did the Fennel module in the most recent ROMS has incorporated all 10 state variables mentioned from Line 94-Line 100. If not, that should be uploaded to Zenodo or somewhere. Also, the parameter scheme in the biological model should be provided. I did not find that in Zhang et al. (2020)

Review of Biogeosciences manuscript “Role of phosphorus in the seasonal deoxygenation of the East China Sea shelf” by Arnaud Laurent and Coauthors.

The manuscript presents a modelling study of the effects of nutrient limitation on deoxygenation in the East China Sea shelf. The manuscript is well composed, and results are clearly presented using meaningful figures. In particular, the conceptual scheme presented in Fig. 11 well resume the main results of the manuscript. Thus, my suggestion is to publish the manuscript after minor revisions (listed hereafter).

1. 75. Consider changing “simulations with and without P” in “simulations with and without P limitation”.
2. 79. I suggest specifying that the scenario simulations are all made with reduction of nutrient loads.
3. 91-102. Even if references are provided, maybe some additional details on the simulation would be helpful for the reader. For instance, are the circulation and the biogeochemical model coupled online or offline? Since the simulation results can be in principle affected by any nutrient inputs, information about treatment of nutrient fluxes at boundaries and at surface could be of interest for the reader.
4. 106-108. My suggestion is to somehow revert how the noPlim simulation is introduced by firstly clarifying that the objective is to have a simulation without P limitation, and then by specifying that this has been implemented disabling P in the biogeochemical model.
5. 110. Maybe the final sentence of the paragraph (“Otherwise […] simulation”) can be removed if the previous suggestion 4 will be accepted by the Authors.
6. Sub-section 3.1 is inserted in the Results section. However, since riverine input is a forcing of the simulations, I am wondering if it would be more appropriate to move 3.1 in section 2 (Methods).
7. 130-135. Validation is made considering observations limited to the surface layer. I recognize that validation can be made only with available data and that the main aim of the manuscript is not the validation of the model, however the Authors could introduce in the Discussion some comments about the implications of the use of a relatively limited dataset or the expected impact of additional observations.
8. Figure S1. If I am not wrong this figure shows the same results of Fig. 2 of Zhang et al. 2020. I think that it should be somehow highlighted that this material has been already published elsewhere.
9. 191-192. Since only regions 2, 4 and 5 are shown in Fig. 7, I suggest removing the “e.g.” used into the brackets: “more concentrated in the smallest area (zone 2) and more diffuse in the larger areas (zones 4 and 5).”
10. 231. Maybe it would be clearer “The N+P and N-only reduction strategies” instead of “The 2 strategies”.
11. Figure 6. Even if one can guess that the increasing numbers on the horizontal axis are the distance along the CE-JI line, it should be better to specify this aspect in the figure or in the caption.
12. Figure 7. I think that it could be useful to specify that the change is related to P limitation.