|These authors have mostly done a good job of addressing the previous reviews. Some aspects of the statistical analysis are inadequately explained, and there are a number of ambiguities in the figure captions /legends.|
(1) I applaud the authors' efforts to add tests of statistical significance, but the description of what was done remains incomplete and unconvincing. The description on 87-94 is too terse to be useful. How exactly we get from what is described here to a significance level of 0.03 (253), 0.04 (330), or 0.001(331) is quite mysterious. Let's consider Figure 7b for example: there are 9 MHW years, and the Large Phytoplankton anomaly is negative in all of them. So it would certainly seem plausible that the probability of that occurring by chance is <5% or <1%. But one can not tell from the present text how that was demonstrated. We need more detail e.g., about the effective sample sizes. If we have 63 years of monthly data, but we are considering only a seasonal mean anomaly for the MHW years, then N2=9 and N1= at most 63. If we are considering monthly data then 63*12 is almost certainly an inflated estimate of N1 because of autocorrelation within the time series (e.g., 10.1175/1520-0442(1997)010<2147). If we are using annual or seasonal mean data then autocorrelation should be weak and N1~63 will not be too far off. In any case, whatever assumptions the authors are making, they need to spell them out in more detail.
(2) Two issues that are partially unresolved from the previous review (there are few minor ones as well, see below) are the description of iron limitation of small phytoplankton and question of cell-size-dependence of carbon-to-chlorophyll ratios. I still find the text ambiguous regarding whether the lack of iron-limitation of small phytoplankton is an emergent property or is programmed into the model a priori (e.g., 118, 283). And the repeated claims that large phytoplankton have a systematically higher Chl:C ratio (e.g., 264, 290, 323) are unconvincing. Chl:C is function of temperature, irradiance, and cell nutrient status. There may be taxonomic differences, but they are generally much smaller than variation with environmental conditions. The repeated invocations of this idea sound glib and superficial and explain very little.
(3) I think section 3.3 is too long and contains some superfluous text (e.g., 259-271). If this material is important enough to keep in the Results then the data are important enough not to be relegated to Supplementary. I would move this to Discussion or delete.
14 change "Yet" to "However" (see also 254, 350, 363)
33 delete "in the Northeast Pacific"
37 change "and/or" to "or"
51-52 "surface Chl alone provides little information on food web changes beyond primary production" Actually it doesn't necessarily tell us anything about production, just biomass.
60 Polovina not in ref list
111 "Tropics" should be "Trophics"
113 not sure what the stray "12" means
140 "Mt. Pavlof eruption" is there a literature reference for this?
142 "(2018)" something missing here
145 italicize taxonomic names
187 "west of P8 (P4–P20" P4 is east of P8
188 "climatological" maybe not best word here
190 "due to a lack of data prior to 2011" figure shows data before 2011
203 add "North" before "American"
210 "intrusion of the nitrate-depleted region from the south into the NPTZ" expansion of the nitrate-depleted region northward into the NPTZ
217-220 I would change "the temporally variable, high-nitrate near-shore region" to "the highly variable but generally nutrient-rich near-shore region" and "the depleted nitrate region" to "the seasonally nitrate-depleted region"
221 change "Observations show" to "Ship-based Line P observations show"
223 delete "lateral"
244 micromolar should be nanomolar?
246 change "primary production" to "phytoplankton production"
275 change "dropping" to "declining"
281 "starts nearly a month earlier (early April vs late April)" how do they know this if working with monthly data?
284 something missing after "-2 mmol/m^2/d"
339 "further supports this shifting of the phytoplankton assemblage" further supports the hypothesis of a shift in the phytoplankton assemblage
360 "Our results agree with this literature" Our results support these previous studies
383 change "depleted nitrate" to "nitrate-depleted"
397 "three different nitrate ..." something missing here
400 " Note that for ... the year is also used." On the one hand this seems like unnecessary belaboring of the obvious. But I understand the impulse behind it, as reviewers are often dense and demand explication of things like this. Maybe it can be worked into the figure caption.
414-415 This sentence borders on tautology. Maybe just delete everything after "both biomes".
419 change "relieved" to "reduced"
428-429 " iron returns to near-climatological levels by summer" this needs a data reference (e.g., Fig. 9d)
438 "Evidence of this shift has been observed in the AG during the “warm blob” (Peña et al. 2019) which found higher concentrations of cyanobacteria in the nitrate-depleted region of Line P and data presented in this paper, which show higher Chl a concentrations in the smaller size classes at OSP (Sect. 3.3, 3.4)." Break this into two sentences.
446 change "persist" to "recur"
Figure 1 - The meaning of the rectangular box does not appear to actually be stated in the caption (I assume it is the region averaged in panel a). The contours in panels d and e are stated to be gray but appear black, and there is an additional black line whose meaning does not appear to be stated. Anyway, why use the same color for the contours and the Argo float trajectories? and why not use the same contour levels for b/c and d/e? There are only 5 panels (a-e), but there is a reference in the text to "Fig. 1 e & f" (168).
Figure 3 - Caption is incoherent and some labels appear incorrect. (a-c) and (d-f) should be a, c, e and b, d, f? Title above panel b should be MOM6-COBALT?
Figure 4 - The map of the station positions is in Figure 1 not Figure 2. The map does not show the station numbers, so the reader is being asked to "read between the lines" here, although in this case the inference is fairly obvious. However, there are many more ticks on the plot than there are labels, or symbols in Figure 1 (i.e., all of the minor stations are shown) and the caption should explain what these are. I still think "2 μM nitrate contour" is more appropriate than "2 μM nitrate boundary" (see also e.g., 211, 215, 232).
Figure 5 - "boxes for AG and NPTZ are shown as described in Fig 1" Actually, Figure 1 has only one box and it is not explicitly defined.
Figure 6 - MLD, POC not defined. Iron should be in nM? Change "Modeled composite of the 9 marine heatwave anomalies" to "Modeled composite anomaly of the 9 marine heatwaves".
Figure 7 - The abbreviations "Lg" and "Sm" do not appear in the figure itself, and nowhere else in the text, so I would say they are expendable. Since the AG and NPTZ data are shown on separate panels, why does each warm composite need its own color?
Figure 8/9 - Again the averaging box is misspecified: the box in Figure 1 is not black and its boundaries don't fit what is stated here. Also the y axis units do not appear to be stated anywhere. And there is something wrong with the legend to panel (h): I think "Mod MHW climatology" is not correct. And the main legend has a solid line for individual years while dashed lines are actually used.
Figure 10 - Definitions of regions need to be stated (e.g., by reference to Figure 5 once its problems are fixed). Meaning of vertical bars still not explained.
Figure 11 - y axis label should be anomaly in multiples of sigma? Maybe change "winter supply" to "winter inventory".
I have doubts about the usage of "Chl" as synonymous with "chlorophyll". When this abbreviation is defined, it refers to satellite chlorophyll (31). But then it is used to refer to field measurements (e.g., 222), and there are several references to "Chl a". While oceanographers often use "chlorophyll" and "chlorophyll a" interchangeably, chlorophylls are a family of pigments of which chlorophyll a is only the most common (e.g., 101/m101p307). Satellite chlorophyll is an operationally defined quantity calculated from outgoing shortwave radiation at the top of the atmosphere, and implicitly includes absorption by other pigments. I think "Chl" should be reserved for satellite chlorophyll, and "chlorophyll" spelled out where it refers to field (bottle) data or specific pigments.
It is customary to choose a discrete significance threshold like 0.01 or 0.05. When I see a number like 0.03 (253) it makes me think the authors just chose the smallest value that all of their data exceeded. Not that this is wrong necessarily, or that it implies the test is artificial (but see my major comments above). But the convention is to e.g., use P<0.05 if 0.025<P<0.05.
"nitrate limitation" still appears in several places (e.g., 15, 375)
"sigma" is sometime spelled out, sometimes a symbol
I don't like the terms "subtropical-like" and "subpolar-like". If the outstanding characteristic of the subtropical biome is that nutrients are depleted year round (371), why does a mid-latitude seasonal convection/stratification regime become "subtropical-like" just because seasonal drawdown becomes a bit more widespread or intense? And the use of the word "mode" (e.g., 310-311) is not consistent with its usual meaning.
Multiple references within a parenthesis should be arranged either alphabetically or chronologically; pick one and apply it consistently. This paper uses both, and sometimes neither (e.g., 60).
The figures are not cited in order in the text (e.g., 90).
Always leave a space between a number and its unit (e.g., 228)