General Comments:

This study conducted a microcosm experiment using a natural Chesapeake Bay phytoplankton community to evaluate bloom dynamics over time. The authors present a thorough and well-written manuscript which captures trends in diversity, particulates, and stoichiometry during their 8-day incubation. Their data clearly shows bloom initiation and termination, as well as a negative productivity-diversity relationship, and provides well thought out explanations for each of these observations. However, there is a strong focus on grazing as the cause of bloom termination, which seems to be driven by results from the COBALT simulation. Though there is some support for this conclusion, I would urge the authors to be more cautious as I don’t believe there is sufficient data on grazing (rates, community etc.) to conclude that grazing was the main driver of bloom termination, though it may have played a role. Conversely, there seems to be a lot of data to support the impact of nutrient depletion on bloom dynamics (uptake rates, POC:Chl etc.), but this explanation was not fully explored. At elevated summer temperatures and in a closed system, nutrient depletion is likely to have contributed, and these factors should be considered in greater depth. Additionally, the authors could also discuss how a closed-system, like the one presented here, might compare to dynamics in the open ocean.  

Overall, this study was compelling and well-constructed, incorporating diverse methodology, from DNA to an NPZ model, in order to evaluate phytoplankton bloom dynamics. In particular, this study provides valuable insight into how phytoplankton diversity changes throughout a bloom. Some additional discussion could make this study even more broadly applicable.

Specific Comments:

The methods are described well and are generally easy to follow.

Line 91: Were nutrients ever replenished in these microcosms or just spiked initially? 

Line 93: What was the exact temperature during the deck-board incubations?

Line 105: This section is unclear. As I understand, DNA analyses were conducted for all carboys and technical duplicates were done for carboy C only. However, this paragraph suggests that DNA was only analyzed from carboy C. Please reword so this can be clarified.

Line 157: Is it reasonable to include a medium-sized copepod when these were filtered out from the microcosm (i.e. how big is the model grazer)? Why not a small zooplankton (e.g. small copepod or heterotrophic dinoflagellate)? How might this have impacted modeled grazing rates or grazing preferences?

Line 165: Did this assumption of phytoplankton biomass include an estimate of growth from day 1 to 2? What was the reasoning for not measuring chlorophyll on day 1?

Figure 1. This figure is nicely laid out to illustrate dynamics in the microcosm. It might be worth considering to use an average across all carboys, as in panel a, rather than just focusing on carboy C. Alternatively, figure S1 could be substituted here.

Line 240. The POC:Chl increases on day 6 are also consistent with phytoplankton becoming nutrient limited. See Jakobsen and Markager (2016), L&O or Arteaga et al. (2016), Glob. Biogeochem. Cycles. Generally, under replete conditions, phytoplankton cells tend to allocate greater resources to chlorophyll synthesis and growth, resulting in an inverse C:Chl – nutrient relationship. So, these trends could be a nutrient limitation signal as well as a grazing signal.

Line 276: How do you compare the low peridinin concentration with the high dinoflagellate relative abundance?

Figure 4g – This figure is great. It shows changes over time, biological replication, and drivers all in one figure.

Line 352-355: This is a bit confusing. There only needs to be one limiting nutrient to cause a bloom decline. Here, the authors present evidence for a limiting nutrient on days 5 and 6, which is consistent with when the bloom crashes. It seems nutrients are being prematurely dismissed, but I would argue that they should be given greater focus and discussion in this manuscript.

Line 363 – Again, POC:Chl can also be a sign of nutrient limitation. Please review the above references.

Line 365: Could these metazoan sequences result from copepod detritus?

Lines 470-477: It may also be worth noting that this incubation was a closed system design and thus likely is unable to capture all the diversity patterns that exist in an open system. A closed system prevents both immigration/emigration and nutrient replenishment which could have impacts on diversity metrics.

Line 497: While grazing may have contributed significantly to the observed trends, I still think its important not to discount the role that nutrients may have had on bloom termination. This is briefly stated on lines 509-511, but could be expanded on throughout.

Technical Corrections:

Line 36: Maybe “silica cell walls” instead of “silica shells?”

Line 336-337: This could be moved to the results.

References: Sal et al 2013 (Figure 7) is missing from the references.

Line 470: It’s difficult to distinguish the unimodal relationship in the global dataset. Could the points be made transparent (in R, use ‘alpha’), to help visualize the density of points?

This study presents the results from a microcosm experiment conducted to follow a natural phytoplankton bloom in Chesapeake Bay, using various methodological approaches (including 18S rRNA gene analysis and an NPZ model). The introduction in well-written, with a clear focus and a plausible research gap and objective. The methods are very thoroughly described and good to follow. All figures are done very nicely and informatively. I added some comments throughout the methods and introduction to consider. For the discussion, I agree with reviewer #1 that the depletion of nutrients for terminating the bloom should receive a bit more thought (as scratched upon in lines 351-355) and I added some additional comments.

METHODS

Line 87-89: How were the added concentrations of nitrogen, silica and phosphorus chosen? The reasoning for their supply ratio is clear with the follow-up sentence, but I am wondering about their concentration.

Line 100 and line 165: Why were pigment and chlorophyll measurements only started/shown on day 2? In the discussion, it says (in lines 396-398): “The variability in the timing of the bloom peak may be due to minor differences in the starting community that each carboy received, as seen in the dissimilarity present between replicate inoculum samples despite being filtered from the same stock of water.“  To get an own impression of this, it would be informative to see day 0 data for all carboys.

Lines 104-105: The DNA samples at 12:00 were collected for all carboys? Maybe to make it clearer add that it was all carboys in that case. How did you select the days on which you additionally sampled carboy C i.e., what additional information do you gain from those additional days?

Line 165: What is the assumption of a 1:100 biomass ratio of zooplankton to phytoplankton based on? Is there any literature on this that can be referenced here?

Line 192: Which alpha diversity metric was calculated and why?

RESULTS

Fig. 1: Why are nutrients and pigments only shown for carboy 3? I suggest including the other carboys as well. As written the text and as also seen in Fig S1, the carboys indeed behave quite similarly. But with only showing one carboy, it always seems a bit suspicious to me at first.

Lines 221-222: I think, it is not 100% correct to say that phosphate followed a similar pattern as the other nutrients. In the next half sentence, it is already stated that different from the other nutrients, phosphate gradually decreased while the other nutrients stayed rather constant until day 4 and then rapidly decreased. Consider rephrasing the start of the sentence.

Fig. 1: Chlorophyll c which is present in the Figure 1 is not mentioned in the text, although all other pigments are mentioned. Please clarify.

Lines 265-269: This part can be moved to the methods.

Fig. 4: Just leaving this comment here with knowing this is hard to change. While reading the paragraph from lines 274-281, I realized I am not able to tell the classes in Fig. 4 apart myself due to very similar colors. As I said, I am just leaving this here as a note.

Fig. 4: The color of the inoculum shapes is hardly visible. I know the color in this case is redundant with the shape, but maybe changing the shape to a diamond would help here.

Line 314: Please specify “dozens of diatom OTUs”.

DISCUSSION

Line 347-348; line 354-355: These two sentences read a bit like they open a line of reasoning but close it again without giving it enough credit. While the first one says that a depletion of nutrients on day 5 led to the bloom’s demise, the last sentence says that factors other than nutrient limitation need to be considered for the bloom’s termination. This leaves the question open why the nutrient limitation that is mentioned in the first part is not further discussed.

Line 433: When H was already lower than other studies, but not as low as expected during a bloom, were the other studies that are referenced here not during a bloom? Whether they measure H during a bloom or not already makes quite a difference, as also mentioned in the discussion.

Line 476-477: This statement needs some references, I think, even though some are mentioned before in the text.

SMALL CORRECTION:

Line 536: “(c) particulate organic carbon (POC) and …” instead of “particulate organic (c) carbon (POC)” in description of Fig. 3.