After reading the manuscript, I have the following ideas about the experiments carried out: Originally, the influence of upwelling water on the biogeochemical processes in surface water was apparently to be investigated. However, due to the El-Nino event, no upwelling water was available. Therefore, water from the OMZ was used to simulate the effect of nutrient-rich water upwelling into the surface layer. The OMZ water was filled into the deeper layers of the mesocosms and was kept there during the duration of the experiment. The 0 – 10 m layer above was sampled. Consequently, from my point of view, the influence of the OMZ on the overlying water layer was investigated. This is a very interesting subject. To what extent effects of upwelling water can be inferred from the effects of OMZ water could be part of the discussion. Upwelling water reaches the water surface directly and mixes with surface water during further transport and does not layer underneath. However, the biogeochemical properties could be compared using date from the literature.

The results part, especially the first chapter "Nutrients" is difficult to read. As emphasized in the title, the focus is on two scenarios. The description of the individual mesocosms does not make the two scenarios clear. You have to look constantly at the figures in order to assign the respective mesocosm to the corresponding scenario. Additionally, constantly different points in time are described in the course of the experiment (sometimes you are at the OMZ water treatment, then at the time before and then again at the end). That's even more confusing. For each scenario, 4 parallel mesocosms were examined. My suggestion would be to describe the general trend of each scenario. Boxplot charts for each sampling time and each treatment would perhaps better illustrate the trends in the two scenarios than the line charts.

In the time course of the parameters in the scenarios, a fixed structure should be maintained - the course before the addition of OMZ water, then the course in scenarios 1 and 2. In the discussion, the influence of different variables on the enzyme activity in the surface water could then be considered the addition of OMZ water and the individual OMZ treatments. In my view, such a structure would make for an interesting manuscript.

Further Comments:

Title:

The word „ production“  should be replaced by the word “activity” (in the abstract too) , because that is what was measured. From my point of view, no upwelling scenarios were examined.

Introduction:

Page 3 line 53: “The ongoing warming…”

Comment: What is meant: Warming due to climate change or warming of upwelling water during transport. Please formulate exactly.

Page 3. Line 45-46: “The fate of biomass…”

Comment: I would move this sentence to the end of the chapter (line 52) and continue with lines 63ff. The chapter describing the OMZ (line 53-62) could be placed before the objectives.  

Material and methods:

Page 5, line 93 and throughout the entire ms: Please replace deep-water by OMZ-water, because, in the ocean, water from 90 m is not deep water.

Page 5, line 99: “…. And sampled every second day…..”

Comment: Please change into “….. and sampled every second day over a period of 50days…”

Page 5 line 103: “ … on day 5…”

Comment: the words can be deleted, they are not necessary here.

Page 5, line 103: “.. from 90m depth…”

Comment: According to Bach et al. 2020 water was taken from 70 m depth.

Page 7 , line 136-142: “To measure  inorganic nutrients, total dissolved nitrogen (TDN) and phosphorus (TDP), the samples were first  filtered through pre-combusted (5 h, 450°C) Whatman GF/F filters (pore size 0.7 µm). The filtrate was collected in 50 mL acid-cleaned high-density polyethylene (HDPE) bottles and placed directly into a freezer (-20°C). Later the filtrates were thawed at room temperature over a period of 24 hours and divided in two. The first half was used to determine inorganic nutrient concentrations as described above. From the other half we determined the TDN and TDP concentrations. “

Comment: I would move this paragraph to the beginning of the Nutrient chapter

Page 9, line 192: Which extraction agent was used? Please provide a reference for the chlorophyll extraction and measurement

Page 12-13: Measurement of enzyme activity: how many parallels (pseudo-parallels) are measured for each sample?

Page 13, line 261-264: Microtiter plates were obviously used for the enzyme activity measurements, as has been described several times. I don't see how it can be done in 20ml Subsamples. Please correct the error.

Results:

According to Bach et al. In 2020 scallop larvae were added to the treatments. It should be mentioned whether the addition of these organisms affects the enzyme activities.

Discussion:

Page 23-24: line 499-502: ” The hydrolysis rates of AP were relatively low compared with most published data, probably linked to the clear surplus of PO4 3-. It is worth to note, however, that we were most likely not measuring the maximal potential hydrolysis rates as substrate addition was relatively low (100 nmol L-1) and would likely have been higher with more added substrate.

Comment: For the LAP activity, it was described in the "Methods" that the substrate saturation concentration was determined before. I assume that was done for the AP activity as well. If so, then higher substrate loading should only marginally increase the AP activity.

In summary:

The authors should consider whether upwelling water effects were actually simulated or whether the manuscript should be geared towards OMZ effects. The manuscript should then be focused accordingly. The current manuscript is not well focused and sometimes difficult to read.

Spilling et al. report a series of measurements from highly complex mesocosm experiments, in which water was initially filled into multiple mesocosms, supplemented part way through with other water, and followed for a considerable time period. Doing fieldwork – especially with mesocosms – is a difficult and often frustrating task, since there is inherent variability in natural waters, and often one (or more) mesocosms will go their own way, despite scientists’ efforts to have multiple replicates of specific treatments. From this perspective, it is understandable that Spilling and colleagues have a somewhat messy data set in which multiple parameters were measured from many mesocosms throughout a time series; figuring out what story the data are telling one is not an easy task.

However, the authors really need to spend a bit more time with their data in order to understand how the pieces do – and don’t – fit together, and above all, in order to make the readers’ journey through the manuscript as straightforward as possible. In the current version, essential information (e.g., depths from which the water for various analyses was collected) is missing or hard to find, and after reading the Methods and the Results, the reader is left confused, instead of having a general roadmap as to which parameters are being compared and how the treatments may or may not make a difference. For example, from just looking at the figures, it appears that the red and blue mesocosms (low and very low OMZ water addition) do not differ from one another in a systematic manner – sometimes the M4 mesocosm is an outlier, but not always; sometimes the M1 mesocosm might be an outlier. Does the main story lie then in the time course of evolution of these different mesocosms? Reading Bach et al leaves the impression that there might be a story in this direction; certainly the differences between the low and very low OMZ water addition does not seem to show robust differences.

Specific comments:

The abstract does not follow a clear line – it reads like a listing of observations/parameters. The authors need to portray a more coherent overview. As an example, the first sentence of the abstract discusses climate change – if this point is not carried through the manuscript, alter the introduction of the abstract so that it points the reader in the direction that does.

The explanation of the setup of the mesocosms and introduction of deep water was extremely confusing. Why was water exchanged in the mesocosms?

The authors should also discuss the effects of the introduction of brine on the microbial community. From reading Bach et al 2020, it seemed to be a very strong brine solution, so the activity/composition of organisms in this lower part of the water column was probably considerably affected. Much of the brine addition part of the manuscript (lines 117-124) was only really understandable after reading Bach et al 202; this part of the manuscript should be re-worked so that the main points are clear without reading the other manuscript for details.

Note also that any data that are re-used from Bach et al 2020 (at least some of the nutrient data?) should be noted in the methods.

It is very difficult to figure out which part of the mesocosm was measured – where did the water come from (which depths) for each analysis? This information is unclear for measurements of nutrients, FDOM, flow cytometry, chl a, and L-MCA measurements. Only the sequencing description also explicitly includes this information (line 199).

Detailed comments

Line 23: “….extracellular enzyme production of leucine aminopeptidase…” (measured activities, not production of the enzyme; reword)

Line 30: note that LAP does not degrade amino acids; it hydrolyzes terminal amino acids from larger units (the N-terminus of peptides or proteins). The amino acids themselves are degraded by other enzymes.

Line 64: organisms are productive, surface layers are not. Reword “the productive surface layer is driven by recycled production.”

Line 71: wording such as “two of the most studied ones” requires references as examples, for example ‘ (e.g. author 1, year; authors et al. , year)’

Line 73: note that the Leu-MCA substrate integrates the activities of a wide range of peptidases (Steen et al. 2015, Substrate specificity of aquatic extracellular peptidases assessed by competitive inhibition assays using synthetic substrates Aquatic Microb Ecol 75:271. In any case, there are also a wide range of leucine aminopeptidase enzymes, so it is not ‘a’ protein degrading enzyme.

Lines 108-116: explain explicitly why the deep water was put in on days 11. The entire water exchange section is very confusing; rewrite or put in a figure as a supplement to guide the reader as to which water had what characteristics

Line 204: wording: what is meant by ‘properly’ homogenizing a sample?

Line 252: How much seawater was added to each replicate, compared to the L-AMC solution? How many replicates were measured? What was the maximum time of incubation (only minimum is given)? Was fluorescence of killed controls subtracted from the live incubations?

From what depth was the water used to measure L-AMC activity? How much time elapsed between sample collection and measurements of enzyme activities?

Line 261: Where was the 20 ml subsamples obtained? What depth was this water collected at?

Line 273: what are the two sample treatments (at this point in the manuscript, the differences are not clear)

Line 296: It is not clear from which depths the measurements were made, so the effects of the addition of the deep water are a bit confusing.

Line 326: at what depth was chl a measured?

Line 343: Are any sequences available for the initial mesocosm, or for water outside the mesocosms? Why were these particular time points selected for sequencing?

Line 366: What is the difference between deep water and OMZ water? The varying terminology is confusing.

Note that a rate of 359 nmol L-1 h-1 is not low – it is far higher than most rates reported in the literature for water column measurements of LAP.

Line 369: what is the rationale for plotting ‘cumulative LAP’ activities? Presumably if rates had been measured at even more time points, then the cumulative LAP rates would have been even higher, but it is difficult to understand the biological or biochemical rationale for summing the rates in this fashion.

Line 382 – line 390: use of statistics in this manner leaves the reader with the impression that the authors have run out of ideas. For example, the LAP activities are extraordinarily high. The statistical link with the bacterial community and biogeochemical variables is ok, but what underlying biological explanation would the authors like to put forward? This point would be far more interesting than just the statistics.

Line 405: the authors state they wish to “relate the biogeochemical and microbial community to the extracellular enzyme activity and a more detailed description of the temporal development and biomass comparison of microbial groups will be presented elsewhere in this special issue (e.g. Bach et al., 2020)”. Bach et al discuss the phytoplankton, but the bacterial data could use some discussion, since presumably they are also the source of the LAP enzymes.

Line 449: first mention of integrated 0-10m sampling? Should have been easy to find in the Methods.

Line 472: what was the incubation temperature of the LAP samples? How long was the interval between water sampling and measurement?

The final part of the discussion contains considerable repetition.

Fig 1 caption: what is the difference between deep water and OMZ water? In addition, should note that the Pacific water was measured from water outside the mesocosms (as explained in Bach et al, but this term should be explained here, the reader shouldn’t have to continually refer to Bach et al.)

Fig 2 caption: see above with respect to Pacific water measurements

Fig 5 is extremely hard to read. Maybe try a bubble plot, or group the colors in a non-random manner to make it easier to determine which sequences are which colors (use perhaps patterns on the colors to distinguish them)

Figs 6 and 7: cumulative enzyme activities are meaningless (they depend on the frequency of measurement), so these panels should be deleted.