Dear Dr. Hieronymus,
Based on the concerns raised by the reviewers, we cannot accept your manuscript for publication in Biogeosciences at this stage. However, we would be very interested in considering a revised version that addresses the reviewers concerns as specified in your answer to the reviewers comments.
Sincerely,
Christine Klaas

Dear authors,
below are a few comments that should be considered before submitting the final version for publication. While I understand that the method used requires long time series, it would be helpful to add some information (a few sentences) on how the results from the simulation and the analysis presented in the manuscript could be used to improve interpretation of field data, or alternatively be validated by field observations. Further, the comparison with field data on plankton groups (i.e. Fig. 3) could be more detailed, in particular, a discussion on whether the mismatch (for example cyanobacteria seasonal dynamics and biomass after 2005) is due to differences between field and model environmental parameters presented in the analyses (nutrients, light, mixed-layer depth, temperature and salinity) or to the representation (growth parameters) of the different plankton groups in the model? This would make a valuable addition to the manuscript.



Equation 1 (lines 78 and below): while the meaning of the subscript PHY is explained, no information is given about the parameter PHY at the end of the equation.

p. 5, lines 120-122: could be replaced with "...half-saturation constants for nitrate and ammonium uptake, respectively. The exponent in (4) accounts for inhibition of nitrate uptake in the presence of ammonium (e.g. Dortch, 1990; Parker, 1993)".

p. 5, line 122 and equation 6: Why introduce a new term "PO4LIM" when this parameter is the same as PLIMPHY already in use? The author could simply use PLIMPHY in eq. 6.

p. 5, lines 124-131: Reorganize and shorten paragraph (redundant information) as given below:
" The constant KPO4PHY is the half saturation constants for phosphate.
Nutrient limitation NUTLIM is thus described by a number between 0 and 1, where 1 corresponds to no limitation. Since NUTLIM is calculated as the minimum of NLIM and PLIM, NLIM larger than PLIM will temporally cause P limitation of phytoplankton growth rate. Hence, a different formulation e.g. of NLIM might change a models sensitivity to the limiting nutrient. Its impact on system nutrient dynamics on longer time scales is, however, difficult to judge because e.g. nitrogen fixation and denitrification potentially also may be influenced. Further experiments on this issue are out of the scope of the present paper and left for future studies."

p. 5, lines 129-130: "Its impact on system nutrient dynamics on longer time scales is, however, difficult to judge because e.g. nitrogen fixation and denitrification potentially also may be influenced". How so? Can the authors be more specific? Or remove this sentence? Since the model seems to be a standard model for the region, aren't there any sensitivity studies using SCOBI available that can be cited from the literature?

p. 6 line 154: can be replaced with " the nutrient inputs from rivers and point sources between 1970 and 2006 were compiled from..."


p. 6 line 157: can be replaced with " Atmospheric inputs were estimated in a similar manner based on data from Ruoho-Airola et al., 2012."

p. 6 lines 158-159: can be replaced with "For riverine organic phosphorus and nitrogen inputs a bioavailable fraction of 100 % and 30%, respectively, was assumed."

p. 6 lines 158-159: What about atmospheric organic N and P inputs?

p. 6 lines 171-172: can be replaced with "Organic nitrogen was implicitly added through the Redfield ratio (nitrogen to phosphorus) of detritus in the model".

p. 7, line 177: can be replaced with "The phytoplankton functional groups in the simulations and respective observations from station BY15 are shown in Figure 3 and 4."

p. 7, lines 178-181: can be replaced with "Phytoplankton biomass from field observations has been estimated through the convertion of biovolumes into carbon in accordance with Menden-Deuer and Lessard (2000). Phytoplankton biomass for the model simulation was estimated from chlorophyll (Chl) assuming a C:Chl ratio of 50. This ratio is in the middle of the salinity dependent range found by Rakko and Seppälä (2014)."

Section 2.4 (Evaluation): This section is an interesting and important addition to the manuscript but it could be transferred to results section 3.1. The comparison indicates some aspects that are not mentioned and unlikely to be simply due to the convertion factor as hypothesized in lines (188-190). These are:
- The model does not properly predict diatom blooms.
- The model significantly overestimates Cyanos from 2005 onwards while nanoflagellates are overestimated between 1999-2001.

p. 7, lines 192-194: can be replaced with "The wavelet transform method and its applications have been described in several studies (e.g. Lau and Weng, 1995; Torrence and Compo, 1998; Grinsted et al., 2004; Carey et al., 2016). Below we provide, therefore, only a brief overview of the method."

p. 8, lines 215-218: can be replaced with "The disadvantage of wavelet transform analysis is that it requires long datasets without gaps, while on the temporal scale of climate change such observations on plankton dynamics are lacking."

p. 8, lines 238-239: can be replaced with "...the spring and autumn blooms, further, the power of both periodicities increases markedly after 1950."

Further, the authors use the word "visable" in line 238 and in other parts of the manuscript. Replace visable (not an english word) with visible or synonyms (i.e. noticeable, discernible etc...).

p. 9, lines 243-245: Fig. 3 indicates that cyanobacteria biomass in the model is overestimated from 2003 onwards. Isn't that the reason the Chl maxima shifts towards August-September?

p. 9, line 248: replace "a strengthening of primary production" with "an increase in primary production"

p. 9, lines 251-252: replace with "This has led to a change in nutrient availability and dynamics as anoxia leads to a release in sedimentary phosphate (Conley et al., 2002; Savchuk, 2010)

Legend Figure 7: why is ammonium included in the term DIN? Replace with "Time-series of volume of anoxic waters (top panel), deep water concentrations of nitrate + ammonium (blue) and phosphate (red) (middle panel) as well as nitrate (blue) and ammonium (red) (bottom panel). Deep water concentrations where averaged below the mixed layer depth for the Baltic proper."

p. 9, line 271: replace with "...Fig. 8. Mixed layer values of NUTLIM increase over the 20th century indicating higher nutrient load and less nutrient limiting conditions."

p. 9-10, line 276-277: replace with "The mixed layer nutrient limitation patterns as estimated from NUTLIM and N/P ratios are shown in Figure 9."

Figure 9.: In the figure it seems as if sometimes both N and P limitation overlap. Is this simply due to the resolution of the figure? It might be helpful to plot the actual N:P ratios in the lower panel.

p. 10, line 279: replace with "during the first part of the run is consistent with the studies on pre-industrial conditions" and give the references for the studies referred to.

Figures 10, 11,12, 14, 15, 16, 17: Please explain the meaning of arrow orientations in more detail (i.e. legend in Fig. 10 does not explain how to interpret the angle, the first mention of it is on p. 11, lines 314-315).

p. 11, line 313,316: replace loads with concentrations.

p. 11, line 313 and elsewhere: What does DIN refer to? Is ammonium included like in legend of Fig. 7? Ammonium should not be included in DIN.

p. 11, lines 320-321: sentence is confusing. Figure 15 does not show large differences between phosphate and DIN.

p. 11, line 322: why would low salinity indicate stronger mixing?

p. 11, line 332: could be replaced with "The mixed layer temperature in the Baltic proper has increased..."

p. 12, lines 366-367: can be replaced with "biomass of individual phytoplankton groups increased to such an extend that inter-annual variations are small compared to the seasonal signal in primary productivity..."
Primary production is not discussed in the manuscript. What analysis (results) supports this statement?

p.13, lines 380-381: I am not sure what the message here is: does the estimate of nutrient limitation using N:P ratios better reproduce field observations as compared to the NUTLIM scheme?

p.13, lines 397-398: can be replaced with "Finally, inter-annual variations in irradiance have little influence on phytoplankton biomass accumulation".