In this manuscript the decadal years variability of the IPGP (Initiation of the Phytoplankton Growing Period) is analysed at two coastal stations located in the northern (Iroise Sea) and eastern Bay of Biscay (Bay of Vilaine). The phytoplankton biomass is related to fluorescenece measured by instrumented buoys. The sensitivity analysis for identifying the major causes of the variability is made through a 1-D biogeochemical model applied to the year 2015. The major result of this work is that, despite different environments, the IBGP days are very similar at the two stations. No significative trend in the IPGP is observed in the time series However the variability is high and seems to show an earlier IGBP in the middle of the period -around year 2010- (at about day 60 against day 90 for the beginning and the end of the period). The results are interesting but this manuscript presents some flaws. The data used are fluorescence-derived Chlorophyll-a, the identification of the factors influencing the IPGP is made through a model and the conclusions are evasive. On the first point, auxiliary analysed Chl-a concentrations collected bimonthly (which is a large interval for this purpose) corroborate nevertheless the IBGP derived from fluorescence data. The main issue comes therefore from the discussion based on model results. The model is considered as perfect and the causes of the variability are discussed from its outputs. From the introduction to the conclusion and throughout the discussion the real issues of the IGBP have not been considered with sufficient care. Blooms at local stations in river plumes may or may not occur, the true question is what happens at large scale? What is the connection with the dynamics of phytoplankton in the whole area? We know since the end of the 90's that strong early blooms within the Gironde plume may consume a large part of the winter Phosphorus stock at the beginning of March or even earlier. This has been attested also by satellite observations at broader scale. There is in the understanding of these late winter blooms (between day 50 and 90) a critical issue for identifying the “major cause” of the “major distubance” of the biological environment over the continental shelf of the Bay of Biscay. For the initiation of early offshore blooms, the light is the prevailing factor, not the SST; hence a verification of the critical depth hypothesis as formalized by Sverdrup. Although criticized with real arguments, this simple theory may be locally verified in the bay of Biscay. The blooms occur in the clear, relatively cold, and stratified waters of the outer river plumes (Loire, Vilaine, Gironde). These blooms, sometimes very strong, have a high impact in biology as they provide foods for benthos at the end of winter and they consume a large part of the phosphorus stock in the surface layer with consequences in the phytoplankton size. As the concentration of Phosphorus in the rivers has been declining at high pace for these last twenty years, these blooms could have a stronger impact in the future. My feeling is therefore that this study presents some interesting results but they have to be considered as a very local representation of much larger dynamics. Considering these time series at the stations together with satellite data of chlorophyll-a and a 2 or 3-D model appear to be the next steps to propose for future investigations. A better consideration of the atmospheric environment would also benefit the understanding of these late winter blooms as anticyclonic conditions associated to high solar irradiance and low wind (hence lower turbidity) generally prevail at the onset of the late winter blooms in the Bay of Biscay.

Specific comments:   My general comments have implications in the abstract, the introduction, the discussion and the conclusion.

Abstract: “The use of a one-dimensional vertical model coupling hydrodynamics, biogeochemistry and sediment dynamics shows that the IPGP is generally dependent on the interaction between several drivers.Interannual changes are therefore not associated with a unique driver (such as increasing sea surface temperature).” Nobody would dare say that temperature in the unique driver of the IGPB. “Extreme event also impact the IGPB”. Obvious but how is it quantified in the text? Not useful mentioning it.

Introduction: “Moreover, theories proposed for the open oceans are not relevant in coastal zones.” Really? Discussion Extreme events: “In coastal stratified regions (e.g. under the influence of river plumes), strong wind and tidal mixing can enhance the mixing and break down stratification. Such conditions can also enhance phytoplankton production (Joordens et al., 2001). During the IPGP, except during floods, both regions are weakly stratified and are then less sensitive to combined wind/tidal short events.” Not useful. In fact the stratification acts positively for initiating blooms in coastal and open waters at the end of winter.

Conclusion: You could imagine something of higher ambition than adding horizontal advection in the model …!

Figure 7: From the legend we are looking for the Chl curve. It would be better to change the legend for something similar to “Day at the IGPB and environmental drivers: . Illustrations in 2011, 2013 and 2014. ….” 

Main findings of the study

In this manuscript, the authors describe the interannual variability of the phytoplankton blooms contrasting two coastal eutrophic French bays. By using a combination of high-frequency in situ information (buoys) and simulation model (IDV) they attempted to identify main environmental drivers (climatic – hydrological) that modulate variations in observed and estimated parameters of the phytoplankton growth given some explanations of the role of water temperature and turbidity variables. Here a main time delay in triggering phytoplankton blooms was detected during 2010-2020 period. The authors also pointed out the strong influence of “extreme events” such as cold spells and floods over phytoplankton blooms during winter.

Although the authors do some interesting observational/modelling approaches with the data available, the manuscript is mostly presented as a description of the data, thus, their interpretations remain mostly speculative. I strongly suggest using additional information such as inorganic nutrients since both bays are defined as eutrophic systems. The main objective is not clearly defined, it should be written as a major one; there are statements very descriptive at the Results section with too many figures, hard to understand showing different years, etc. In addition, I strongly recommend that the authors should make a major effort to write a general hypothesis or conceptual model for a future version.

Especific comments:

1.- The first limitation that comes to mind is the low sampling effort carried out in Bay of Vilaine for the chlorophyll-a (as fluorescence) variable, with only the second period survey completed. Would be possible to fill the gap of the first period with satellite images? In Bay of Brest appears to be an increase trend during the second period and probably both bays may be affected by similar drivers.

2.- Inorganic nutrients: Although both bay are classified as eutrophic areas, the manuscript does not present data on N and Si; Si:N ands Si:P are interesting ratios to explore in the near surface layer, especially for diatoms, a groups that needs silicic acid for the frustule and for dinoflagellates which are associated to nitrogen sources. Species/functional groups could respond more to ratios than concentrations; for example, cryptophythes and dinoflagellates may respond better to N sources (i.e. nitrate, ammonia), whereas diatoms could respond better to silicic acid concentrations.

3.- Model: There are several not clear issues in the parametrization of the model (Table 1), mainly in the methodology section:  some restrictions on some parameters could be explained in more detail; for example, the initial value of O for dinoflagellates at the Bay of Vilaine; the starting values for N and Si nutrients, are they coming from a observed data base?

4.- Functional groups: Authors stated that Skeletonema spp. and Chaetoceros spp. as dominant species during both periods; would be possible that the increase in fluorescence is due to changes in taxonomical groups? Since both bays are under the influence of nutrient loading (mainly nitrogen), would be possible that the toxic dinoflagellates dominance would be part of the seasonal succession or interannual variability? Any evidence of more frequent and intense Harmful Algal Blooms (HABs) during the annual cycle or during climatological-hydrological extreme events?

5.- According to authors, Temperature and Turbidity were the main drivers (I should prefer to say factors) of the variability of phytoplankton growth, however there is poor mechanistic explanation for their effects: to the mixing/stratification process such as Sverdrup hypothesis? Turbidity could affect in both ways to phytoplankton growth: large concentrations of terrigenous particles could decrease light penetration or increase inorganic nutrients (N, P) flux adjacent coastal land.