the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 13 Jan 2025
Oceanic enrichment of ammonium and its impacts on phytoplankton community composition under a high-emissions scenario
Abstract. Ammonium (NH4+) is an important component of the ocean’s dissolved inorganic nitrogen (DIN) pool, especially in stratified marine environments where intense recycling of organic matter elevates its supply over other forms. Using a global ocean biogeochemical model with good fidelity to the sparse NH4+ data that is available, we project increases in the NH4+:DIN ratio in over 98 % of the ocean by the end of the 21st century under a high-emission scenario. This relative enrichment of NH4+ is driven largely by circulation changes, and secondarily by warming-induced increases in microbial metabolism, as well as reduced nitrification rates due to pH decreases. Supplementing our model projections with geochemical measurements and phytoplankton abundance data from Tara Oceans, we demonstrate that shifts in the form of DIN to NH4+ may impact phytoplankton communities by disadvantaging nitrate-dependent taxa like diatoms while promoting taxa better adapted to NH4+. This could have cascading effects on marine food webs, carbon cycling, and fisheries productivity. Overall, the form of bioavailable nitrogen emerges as an potentially underappreciated driver of ecosystem structure and function in the changing ocean.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(5273 KB)
-
Supplement
(9518 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(5273 KB) - Metadata XML
-
Supplement
(9518 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-3639', Anonymous Referee #1, 26 Feb 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2024-3639/egusphere-2024-3639-RC1-supplement.pdfCitation: https://doi.org/
10.5194/egusphere-2024-3639-RC1 - AC1: 'Reply on RC1', Pearse Buchanan, 19 May 2025
-
RC2: 'Comment on egusphere-2024-3639', Anonymous Referee #2, 24 Mar 2025
In this study, the authors apply a global marine biogeochemical model to investigate the changes in the relative abundance of diatoms in response to shifts in NH4:DIN ratio. This topic is of interest to both the modelling community and the broader research audience. Overall, the manuscript is well-written but would benefit from some structural reorganisation for readability. Additionally, several sections require further clarification and the inclusion of more supporting information.
General comments
Typically, the presentation of model-data agreement (misfit) should precede the transient simulations, i.e., we should have “build confidence in the model” first. I suggest reorganising the discussion section so that the steady-state evaluation of the relationship between NH4:NO3 and diatom abundance appears first, followed by the transient simulations. Also, the “Model experiment” section should be moved after “Statistical analyses” in the methods.
Several statistical techniques are applied in this study. I recommend providing more background information and justification for the selected values (e.g., VIFs and spline complexity). This would help readers unfamiliar with those tools to better understand the methodological choices.
Although the biogeochemical model is based on a previously published version, this study applies a different nitrification configuration. The manuscript should provide at least a brief summary of how these changes affect key biogeochemical inventories (such as the relative abundance of the two phytoplankton types) and fluxes (including nitrogen fixation) to support the new model’s validity.
Fig. 3 shows a 70% difference in the decline of delta % diatoms between the Modelcontrol and Modelcompete. However, their delta µM C diatoms are very similar (Fig. S9). Could the changes in delta % diatoms during the transient simulation mainly result from differences in the initial conditions rather than the NH4:DIN ratio? If so, the decline in delta % diatoms might be primarily driven by a decrease in the overall nutrient pool rather than by competition with nanophytoplankton.
This brings another question to me. The manuscript does not discuss nanophytoplankton abundance during the transient simulations. Is the decline in their abundance really smaller than that of diatoms by the end of this century? From Fig. S8a, the “delta other phytoplankton” are negative in most of the low latitude regions, where the NH4:DIN increases mostly. Since the title highlights impacts on phytoplankton community composition, I believe this is an important point. More discussion is needed on how both phytoplankton groups respond to the NH4:DIN shift.
Specific comments
Title: The majority of changes in diatom abundance due to changes in the NH4:NO3 ratio occur in trophic and subtropic regions (Fig. 3a and d), where NH4 concentration actually decreases (Fig. S6). Therefore, I suggest revising the phrase “enrichment of ammonium” in the title, as it may not accurately reflect the spatial trends shown in the results.
Line 27: an -> a
Line 30: remove the extra “in”
Fig. 1: Suggest adding labels to indicate which conditions are subject to anthropogenic pressure.
Line 101: Are riverine inputs and nitrogen deposition influenced by anthropogenic forcing in the model?
Line 104-108: As mentioned in the major comments, please include some evaluations here, particularly for the nitrogen fixation since it’s also affected by the forcings. Additionally, I couldn’t find the information regarding the form of N introduced to the system through nitrogen fixation.
Line 121: Please provide a clearer description of the “changing circulation (‘Phys’)” configuration. For example, does it include stronger stratification? It is not clear which specific factors are incorporated under this forcing. Later in the text (Line 211), changes in sea-ice loss are also mentioned as part of this forcing, so clarification is needed regarding which processes are included.
Line 129: Please provide a reference or justification for this criterion (0.1 mmol C m-3).
Line 160-163: Why the rate saturates when pH > 8? Base on the equation in Fig. S5 the rate is supposed to keep increase.
Equation (1): at least one item is missing before +s1(x1).
Line 183: the intercept 𝛼 is missing in the equation.
Line 183: “thin-plate spline” is not a trivial term, it would be helpful if the author could provide a brief explanation or background information in the text.
Line 184: independent variable. -> independent variable x.
Line 193-194: Please provide a bit more information regarding the VIFs and the criterion.
Line 210-211: The full name was mentioned in the Methods and it is sufficient using only RCP8.5 here.
Line 211: “sea-ice loss” should be mentioned already in the methods instead of here.
Line 217: What does the ±6 stands for?
Line 219: Please give specific locations for examples for the “oceanographic fronts”.
Line 225-242: When comparing Fig. 3b and 3e, the major contribution to the changes in diatom abundance appears to come from the "phys", which aligns with its 55% contribution to the NH4:NO3 ratio. However, although the contribution from OA (25%) is about double that of Warming (13%), OA appears to have almost no effect on the changes in diatom between Fig.3b and 3e. This discrepancy should be addressed and further explained in the manuscript.
Fig. 3 and Fig. S9: The delta µM C diatoms of “All” are comparable between Modelcontrol and Modelcompete in Fig. S9, yet the delta % diatoms in Fig. 3 show much larger difference. Does this imply that the total diatom biomass is substantially higher in Modelcompete? Additionally, in Modelcontrol, the delta µM C diatoms decline under “Warm” is about 2.5 times greater than Phys (Fig. S9c), but their delta % diatoms declines are similar. Does this indicate that the diatom biomass is much higher in “Phys”? These discrepancies should be clarified in the text, as they are important for interpreting the results.
Line 289: thid -> the or this?
Line 296-304: move to method. Also, was the modelcompete simulation spun up before the transient simulation? Such information is missing in the text.
Line 306: Why surprisingly? Is 70% too high or too low?
Line 493: albiet -> albeit
Line 496: ellaboratiung -> elaborating
Fig. S3: If possible, please add one panel that displays model results from where observations exist.
Fig. S8: I believe the unit (or scale) for the right panels is wrong.
Citation: https://doi.org/10.5194/egusphere-2024-3639-RC2 - AC2: 'Reply on RC2', Pearse Buchanan, 19 May 2025
-
EC1: 'Comment on egusphere-2024-3639', Olivier Sulpis, 28 Apr 2025
Dear authors,
Apologies for the delay, but we have received another review for your manuscript. It was received late, therefore I am pasting it below. Please consider this new review in your reply.
Best regards,
Olivier Sulpis.
______________________________________________
Review of “Oceanic enrichment of ammonium and its impacts on phytoplankton community composition under a high-emissions scenario” by P. Buchanan et al. for BiogeosciencesThe manuscript by Buchanan and coauthors employs a physical biogeochemical model with an improved nitrogen (N) cycle representation to investigate the effect of climate change on the availability of different dissolved inorganic nitrogen (DIN) sources (mainly nitrate, NO3 and ammonium, NH4) to phytoplankton, and the consequences for phytoplankton diversity. They find that over most of the surface ocean the availability availability of NH4 increases relative to NO3, with a global mean ratio increasing from ~7% to 12% by the end of the century. The most significant changes are predicted in mid- to low-latitude regions.The model also projects a global decline in diatom biomass of about 3%. By comparing model output with geochemical rate observations and analysis of Tara Oceans’s genomic dataset, the authors suggest that this shift towards higher NH4/DIN ratio supports (1) an increase in regenerated production, and (2) a decrease in the relative abundance of diatoms, which are more dependent on NO3, in favor of smaller phytoplankton groups (pico- and nano-phytoplankton) that are more reliant on NH4.As climate change reshapes the oceanic ecosystem, it is clear that there will be ecological winners and losers, but the outcomes remain highly uncertain, both in magnitude and patterns. Buchanan and coauthors approach this question from an interesting angle, focusing on shifts in the form of dissolved inorganic nitrogen and how these changes may affect phytoplankton diversity.This is an interesting study that addresses a globally relevant topic through the use of a state-of-the-art model and a thoughtful analysis of observational data. The model projections and the analysis of the Tara dataset are stimulating and valuable, both on their own and when combined to support a mechanistic interpretation of the changes observed. In fact, the observational constraints presented here could easily become benchmark for future ocean biogeochemical models, particularly for evaluating their representation of DIN dynamics. For these reasons, I believe the study is appropriate for Biogeosciences, and I am ultimately supportive of publication.However, the manuscript is dense with information, presenting several complex analyses and interpretations that are not always clearly or convincingly explained. At times, I found myself wondering whether the results might be more effectively communicated if the study were divided into two separate papers: one focused on the present-day ocean and the role of NH4/DIN in shaping phytoplankton communities, and another dedicated to future projections and their mechanistic interpretation. The narrative structure of the paper feels somewhat meandering, moving from future model projections and sensitivities to observational analyses, then to model-observation comparisons. This steadily introduces new amounts of information, and a series of new questions are raised halfway through the result section (notably in Section 3.3), increasing the complexity of the narrative. As reflected in my detailed comments below, while I found the study rich with interesting results, I often struggled to follow the logic of the explanations and interpretations— challenges compounded by the paper’s structure.Furthermore, I am unconvinced by the more assertive interpretation that the modeled diatom decline reflects more intense competition for NH4, rather than the more straightforward effect of a general decline in NO3 supply and concentrations—on which diatoms depend directly, given the model’s DIN uptake formulation. To be fair, this interpretation is presented as a suggestion in multiple parts of the paper, and is not necessary for the paper to stand on its own, given the range of interesting results and analyses provided. For example, I find the phrasing of the abstract to be balanced, but many parts of the Results and Discussion present this idea with much less nuance. The authors themselves acknowledge at multiple points that it is difficult to disentangle the effects of an increase in NH4/DIN from those of a decrease in NO3 concentrations, and in my view there is no contradiction in proposing that diatom declines reflect both effects. I’m not sure one can simply isolate competition for NH4 as the main driver of the changes observed — especially given how important circulation driven changes are on a point by point basis. NO3 and NH4 uptake occur in parallel and can jointly affect diatom and other phytoplankton.Specific commentsSection 2.1: It is somewhat surprising that the model explicitly represents diatoms and nanoplankton, but not picoplankton, given the focus on competition between diatoms and cyanobacteria. The most abundant cyanobacteria in oligotrophic regions, Prochlorococcus and Synechococcus, fall within the picoplankton size range. But one could argue that the model’s nanoplankton functionally encompasses both pico- and nano-plankton. A brief discussion of this issue and the potential limitations it introduces could be included.The model description in the SI could be expanded for clarity. It would be helpful to include the equations for DIN uptake, as this is central to interpreting the results of decreased NO3 supply and understanding the distinction between model_control and model_compete experiments. Additionally, including the temperature dependence formulations for phytoplankton and zooplankton growth and grazing would help interpreting the warming-only experiments.The Tara analysis appears biased by the use of observations from depths shallower than 10 m only (lines 171-172). This likely skews the results towards phytoplankton communities adapted to relatively low NO3 and high NH4, in particular in oligotrophic regions, where phytoplankton are commonly found down to 100-200 m depths. This seems like a potentially important limitation, and could benefit from discussion. To my knowledge (but I may be mistaken), Tara Ocean also collected data from deep chlorophyll maxima. Why not including those data in the analysis, or at least consider them in a separate analysis?Section 2.5: the use of model-based fields in the analysis of observations makes me a bit uncomfortable, as it could introduce new, hard to control biases.Figure 2: I suggest including panel S1b, showing present-day NH4/DIN ratios with observations, in Fig. 2. This would reassure the reader that the model captures the basic N distribution patterns, and would help contextualize the changes shown in panel 2a (e.g., is a change by 10% large or small?) The text too could be more explicit , e.g., line 218, increase by 6%, could specify “from X% to Y%”.Lines 219-220: it would be interesting to report the % regenerated primary production and its change.The parameterizations for the pH dependence of ammonia oxidation are not well explained. What is the rationale behind including this pH dependence, and how were the specific functional forms shown in Fig. S5 and S7 chosen? Is it as simple as making the rate inversely proportional to the H+ concentration? (This seems the implication of the formulation invoking the pKa of NH4 dissociation.) What about the alternative formulation? The SI provides too little detail on these points, and more thorough explanation would help readers understanding the models sensitivities.Line 238-240, “in eutrophic regions, where coincidentally, shifts from low to higher NH4:DIN would have the greatest ecological impact”: This could be clarified. The phrasing is a bit confusing, as eutrophic regions are typically characterized by high NO3 concentrations. If NH4 becomes more dominant in these areas, does that imply they are no longer eutrophic? Or is the point that even in nutrient-rich waters, a shift in the form of available DIN from NO3 to NH4 could significantly affect community structure?Section 3.2, lines 258-270. The authors make the point that loss of diatoms was “driven by a combination of stimulated microbial metabolism (60%) and physical changes (40%), while ocean acidification had negligible effects”. While this may be accurate at a global mean level, it risks giving the misleading impression that increased microbial metabolism is the dominant driver of diatom loss across the ocean, as compared to decreased NO3 supply. But is is not true almost anywhere in the ocean, where on a local basis (Fig. S8) changes due to NO3 supply (or light availability at high latitudes) greatly exceed the low, but consistently negative effects of warming on metabolism. But because physical effects on nutrient supply and light availability are both positive and negative, they tend to cancel out when averaged globally. This distinction between local drivers and global mean effects is not clearly conveyed in the current discussion and should be more explicitly emphasized.Lines 272-282. I find this paragraph unconvincing in its current framing. Since diatoms preferentially take up NO3 (given the DIN uptake formulation), it seems straightforward that a decline in NO3 supply and concentrations would reduce diatom production and abundance, without the need to shift the emphasis to increased competition for NH4, which would be a consequence of the change. The reduction in NO3 and the resulting increased reliance on (and potential competition for) NH4 seem more like two sides of the same coin, rather than distinct mechanisms. I’m not sure that NH4 competition is the most correct or useful framework for interesting the model changes. The authors end the paragraph by stating: “when NO3 concentrations decline, competition for NH4 increases, and declines in diatom relative abundance follow”. But is the middle step, “competition for NH4 increases”, strictly necessary to explain the decline in diatoms? It might be more parsimonious to attribute the decline directly to reduced NO3 availability.The bathtub analogy (lines 277-278) should help understanding, here I found it confusing. If nutrients represent the volume in the bathtub, why is productivity described as the inflow? Productivity removes nutrients, it doesn’t add them. And why is recycling represented as the outflow? Conceptually, recycling returns nutrients to the system. Also, at steady state, one would expect inflow and outflow to be in balance. This needs some rethinking or clarification.Lines 291-294, “It is therefore possible that reductions in NO3 and resulting competition for NO3 was a major contributor to the losses of diatoms from the phytoplankton community in our simulations”. This seems like a straightforward explanation for diatom decline, given diatom’s functional dependence on DIN forms and the NO3 declines. But its placement here appears to undercut the argument made a few paragraphs earlier that emphasized competition fro NH4.Lines 294-304, the model_compete experiment is interesting, but I see it more as highlighting another side of the same coin, rather than challenging the idea that declining NO3 is the primary driver of diatom decline. Of course if diatoms had a stronger affinity for NH4, they would fare better under reduced NO3. But this doesn’t change the underlying cause of their decline, which still originates from the reduction in NO3.Line 311, “Physical changes no longer exerted a global negative effect on their total nor relative abundance”. The global mean masks large regional variability, where large positive and negative changes partially compensate each other. As shown in Fig. S8, physical changes are the main local driver, especially in the Southern Ocean, and in fact, across much of the ocean on a point-by-point basis. This nuance should be acknowledged more clearly.Lines 238-240, “determined that a large fraction of the projected declines in diatom relative abundance are due to their competitive exclusion by other phytoplankton in regions where NH4 becomes more important as a nitrogen source”: I am unconvinced that the authors have “determined” that the diatom decline is caused by competitive exclusion, and not by the overall decline in NO3.Section 3.3.1 and Fig. 4: this section provides a stimulating and valuable set of analyses and diagnostics that could serve as a useful benchmark for models. More ocean biogeochemical models should adopt this type of diagnostic approach as a standard practice for evaluating nutrient dynamics and phytoplankton competition.Lines 369-371: does it matter than in PISCES the half saturation constants for DIN uptake are not constant but a function of the phytoplankton biomass P? What are “typical conditions”? To help the reader, the DIN uptake functional forms used by the model should be presented in the SI.Section 3.3.2: this is a stimulating analysis—though at times it felt substantial enough to warrant its own standalone paper. Figures S10-S16 are rich with interesting information, but also dense. I wonder if “goodness of fit” metrics (some o which may be reported in Table S1?) could be included directly on the figures, so that the reader can quickly evaluate the skill of different models.Fig. 5: I got a bit lost in the interpretation of the figure. Not being familiar with GAMs, I am confused by the y-axis, “GAM residuals”, which could be clarified. Also, related to the threshold of 4%; in Fig. 5 nothing specifically makes the 4% threshold stand out. The text points to “rapid losses of diatoms as NH4:DIN became greater than 4%”, but I struggle to see anything special in this threshold, and overall it seems an arbitrary choice.Lines 449-453: points (1) and (3) are general knowledge; point (2) could be as well phrased as “how diatoms are outcompeted as less primary production is fueled by external NO3 inputs”, which would actually be closer to the mechanistic changes at play.Section 3.3.4, “The confounding effect of NO3”: framing the role of NO3 as a confounding factor underplays its central role in controlling diatom growth, and seems an example of reverse causation and post hoc reasoning. Given the well-established importance of physical changes on NO3 supply across the ocean, reductions in NO3 should be taken as the default or “null” mechanisms to explain the diatom declines, rather than a confounding effect. The fact that the whole statistical analysis could be based on NO3 instead of NH4:DIN as a covariate seem to undermine the entire argument made here.Line 473, “We therefore suggest that competition for NH4 directly controls diatom relative abundance”: this feels like a forced interpretation and a leap that is not fully supported by the evidence presented.Line 477-478, “Decreases in NO3 certainly affect diatom growth, but we propose that they mostly do so indirectly by shifting the regime towards intense competition for NH4.” Another sentence that seems to overstate the authors’ interpretation without enough supporting evidence, given the direct role of NO3 in diatom growth, which is explicitly built into the model’s functional formulation.Line 486-487, “but we recast the attribution of change in terms of competitive exclusion for NH4, rather than bulk nutrient declines”: see above.Technical comments:Line 66, “become a self-sustaining regime”: I would rephrase, partly because the message is a bit unclear, partly because it is hard to imagine a case where primary production in the euphotic zone does not involve an external supply of NO3, even in the most oligotrophic regions (were export still occurs and must be ultimately balanced by external nutrient supply).Line 45, “which would work to use up …” this sentence is a bit obscure, please clarify.Line 65, “there are numerous localized studies that showcase how phytoplankton taxa shift in response to changes in the composition of DIN”: it would be useful to add some references for these studies.Line 74: remove repeated “enrichment”.Line 129, “defined as those depths where total phytoplankton biomass was greater than 0.1 mmol C m-3.” This is a bit of an unorthodox definition of euphotic zone, perhaps clarify the rationale and reassure it is generally in line with other common light-based definitions (typically, 1% light levels)Lines 146-149: here a relevant reference could be Tang, et al., 2023, Earth System Science Data, which presents a compilation of nitrification rates. (Presumably data used there and compiled in Tang et al. overlap.)Line 151, I appreciate the authors reporting the units for primary production; for consistency, they could be added for the other major variables discussed (e.g., ammonia oxidation rates).Line 194: maybe provide some more detail on the spline approach. What is k, in k=3? What is “spline complexity”? Etc.Line 289: “thid” —> “this”.Line 307: should the reference be to Fig. 3e?Line 322, “by the in the”, remove “in the”.Line 356: since the exponent is fractional, I would call the function a “fractional-order” Monod function, not a quadratic (where the implied exponent is 2).Lines 358-359: add the units for the half saturation constants.Line 493: “albiet” —> “albeit”.- Line 496: “ellaboratiung” —> “elaborating”.Fig. S1 could also show total DIN, so that one has a complete view of the controls on the NH4/DIN ratio distribution.Fig. S8, right column: I’m confused by the units, the values go up to 0.2, but the units say %. Shouldn’t Fig. S8c be the same as S3a, were value go up to 20%? The figures also look a bit different, but perhaps it’s the contouring that make them look different.End of reviewCitation: https://doi.org/10.5194/egusphere-2024-3639-EC1 - AC3: 'Reply on EC1', Pearse Buchanan, 19 May 2025
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-3639', Anonymous Referee #1, 26 Feb 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2024-3639/egusphere-2024-3639-RC1-supplement.pdf
- AC1: 'Reply on RC1', Pearse Buchanan, 19 May 2025
-
RC2: 'Comment on egusphere-2024-3639', Anonymous Referee #2, 24 Mar 2025
In this study, the authors apply a global marine biogeochemical model to investigate the changes in the relative abundance of diatoms in response to shifts in NH4:DIN ratio. This topic is of interest to both the modelling community and the broader research audience. Overall, the manuscript is well-written but would benefit from some structural reorganisation for readability. Additionally, several sections require further clarification and the inclusion of more supporting information.
General comments
Typically, the presentation of model-data agreement (misfit) should precede the transient simulations, i.e., we should have “build confidence in the model” first. I suggest reorganising the discussion section so that the steady-state evaluation of the relationship between NH4:NO3 and diatom abundance appears first, followed by the transient simulations. Also, the “Model experiment” section should be moved after “Statistical analyses” in the methods.
Several statistical techniques are applied in this study. I recommend providing more background information and justification for the selected values (e.g., VIFs and spline complexity). This would help readers unfamiliar with those tools to better understand the methodological choices.
Although the biogeochemical model is based on a previously published version, this study applies a different nitrification configuration. The manuscript should provide at least a brief summary of how these changes affect key biogeochemical inventories (such as the relative abundance of the two phytoplankton types) and fluxes (including nitrogen fixation) to support the new model’s validity.
Fig. 3 shows a 70% difference in the decline of delta % diatoms between the Modelcontrol and Modelcompete. However, their delta µM C diatoms are very similar (Fig. S9). Could the changes in delta % diatoms during the transient simulation mainly result from differences in the initial conditions rather than the NH4:DIN ratio? If so, the decline in delta % diatoms might be primarily driven by a decrease in the overall nutrient pool rather than by competition with nanophytoplankton.
This brings another question to me. The manuscript does not discuss nanophytoplankton abundance during the transient simulations. Is the decline in their abundance really smaller than that of diatoms by the end of this century? From Fig. S8a, the “delta other phytoplankton” are negative in most of the low latitude regions, where the NH4:DIN increases mostly. Since the title highlights impacts on phytoplankton community composition, I believe this is an important point. More discussion is needed on how both phytoplankton groups respond to the NH4:DIN shift.
Specific comments
Title: The majority of changes in diatom abundance due to changes in the NH4:NO3 ratio occur in trophic and subtropic regions (Fig. 3a and d), where NH4 concentration actually decreases (Fig. S6). Therefore, I suggest revising the phrase “enrichment of ammonium” in the title, as it may not accurately reflect the spatial trends shown in the results.
Line 27: an -> a
Line 30: remove the extra “in”
Fig. 1: Suggest adding labels to indicate which conditions are subject to anthropogenic pressure.
Line 101: Are riverine inputs and nitrogen deposition influenced by anthropogenic forcing in the model?
Line 104-108: As mentioned in the major comments, please include some evaluations here, particularly for the nitrogen fixation since it’s also affected by the forcings. Additionally, I couldn’t find the information regarding the form of N introduced to the system through nitrogen fixation.
Line 121: Please provide a clearer description of the “changing circulation (‘Phys’)” configuration. For example, does it include stronger stratification? It is not clear which specific factors are incorporated under this forcing. Later in the text (Line 211), changes in sea-ice loss are also mentioned as part of this forcing, so clarification is needed regarding which processes are included.
Line 129: Please provide a reference or justification for this criterion (0.1 mmol C m-3).
Line 160-163: Why the rate saturates when pH > 8? Base on the equation in Fig. S5 the rate is supposed to keep increase.
Equation (1): at least one item is missing before +s1(x1).
Line 183: the intercept 𝛼 is missing in the equation.
Line 183: “thin-plate spline” is not a trivial term, it would be helpful if the author could provide a brief explanation or background information in the text.
Line 184: independent variable. -> independent variable x.
Line 193-194: Please provide a bit more information regarding the VIFs and the criterion.
Line 210-211: The full name was mentioned in the Methods and it is sufficient using only RCP8.5 here.
Line 211: “sea-ice loss” should be mentioned already in the methods instead of here.
Line 217: What does the ±6 stands for?
Line 219: Please give specific locations for examples for the “oceanographic fronts”.
Line 225-242: When comparing Fig. 3b and 3e, the major contribution to the changes in diatom abundance appears to come from the "phys", which aligns with its 55% contribution to the NH4:NO3 ratio. However, although the contribution from OA (25%) is about double that of Warming (13%), OA appears to have almost no effect on the changes in diatom between Fig.3b and 3e. This discrepancy should be addressed and further explained in the manuscript.
Fig. 3 and Fig. S9: The delta µM C diatoms of “All” are comparable between Modelcontrol and Modelcompete in Fig. S9, yet the delta % diatoms in Fig. 3 show much larger difference. Does this imply that the total diatom biomass is substantially higher in Modelcompete? Additionally, in Modelcontrol, the delta µM C diatoms decline under “Warm” is about 2.5 times greater than Phys (Fig. S9c), but their delta % diatoms declines are similar. Does this indicate that the diatom biomass is much higher in “Phys”? These discrepancies should be clarified in the text, as they are important for interpreting the results.
Line 289: thid -> the or this?
Line 296-304: move to method. Also, was the modelcompete simulation spun up before the transient simulation? Such information is missing in the text.
Line 306: Why surprisingly? Is 70% too high or too low?
Line 493: albiet -> albeit
Line 496: ellaboratiung -> elaborating
Fig. S3: If possible, please add one panel that displays model results from where observations exist.
Fig. S8: I believe the unit (or scale) for the right panels is wrong.
Citation: https://doi.org/10.5194/egusphere-2024-3639-RC2 - AC2: 'Reply on RC2', Pearse Buchanan, 19 May 2025
-
EC1: 'Comment on egusphere-2024-3639', Olivier Sulpis, 28 Apr 2025
Dear authors,
Apologies for the delay, but we have received another review for your manuscript. It was received late, therefore I am pasting it below. Please consider this new review in your reply.
Best regards,
Olivier Sulpis.
______________________________________________
Review of “Oceanic enrichment of ammonium and its impacts on phytoplankton community composition under a high-emissions scenario” by P. Buchanan et al. for BiogeosciencesThe manuscript by Buchanan and coauthors employs a physical biogeochemical model with an improved nitrogen (N) cycle representation to investigate the effect of climate change on the availability of different dissolved inorganic nitrogen (DIN) sources (mainly nitrate, NO3 and ammonium, NH4) to phytoplankton, and the consequences for phytoplankton diversity. They find that over most of the surface ocean the availability availability of NH4 increases relative to NO3, with a global mean ratio increasing from ~7% to 12% by the end of the century. The most significant changes are predicted in mid- to low-latitude regions.The model also projects a global decline in diatom biomass of about 3%. By comparing model output with geochemical rate observations and analysis of Tara Oceans’s genomic dataset, the authors suggest that this shift towards higher NH4/DIN ratio supports (1) an increase in regenerated production, and (2) a decrease in the relative abundance of diatoms, which are more dependent on NO3, in favor of smaller phytoplankton groups (pico- and nano-phytoplankton) that are more reliant on NH4.As climate change reshapes the oceanic ecosystem, it is clear that there will be ecological winners and losers, but the outcomes remain highly uncertain, both in magnitude and patterns. Buchanan and coauthors approach this question from an interesting angle, focusing on shifts in the form of dissolved inorganic nitrogen and how these changes may affect phytoplankton diversity.This is an interesting study that addresses a globally relevant topic through the use of a state-of-the-art model and a thoughtful analysis of observational data. The model projections and the analysis of the Tara dataset are stimulating and valuable, both on their own and when combined to support a mechanistic interpretation of the changes observed. In fact, the observational constraints presented here could easily become benchmark for future ocean biogeochemical models, particularly for evaluating their representation of DIN dynamics. For these reasons, I believe the study is appropriate for Biogeosciences, and I am ultimately supportive of publication.However, the manuscript is dense with information, presenting several complex analyses and interpretations that are not always clearly or convincingly explained. At times, I found myself wondering whether the results might be more effectively communicated if the study were divided into two separate papers: one focused on the present-day ocean and the role of NH4/DIN in shaping phytoplankton communities, and another dedicated to future projections and their mechanistic interpretation. The narrative structure of the paper feels somewhat meandering, moving from future model projections and sensitivities to observational analyses, then to model-observation comparisons. This steadily introduces new amounts of information, and a series of new questions are raised halfway through the result section (notably in Section 3.3), increasing the complexity of the narrative. As reflected in my detailed comments below, while I found the study rich with interesting results, I often struggled to follow the logic of the explanations and interpretations— challenges compounded by the paper’s structure.Furthermore, I am unconvinced by the more assertive interpretation that the modeled diatom decline reflects more intense competition for NH4, rather than the more straightforward effect of a general decline in NO3 supply and concentrations—on which diatoms depend directly, given the model’s DIN uptake formulation. To be fair, this interpretation is presented as a suggestion in multiple parts of the paper, and is not necessary for the paper to stand on its own, given the range of interesting results and analyses provided. For example, I find the phrasing of the abstract to be balanced, but many parts of the Results and Discussion present this idea with much less nuance. The authors themselves acknowledge at multiple points that it is difficult to disentangle the effects of an increase in NH4/DIN from those of a decrease in NO3 concentrations, and in my view there is no contradiction in proposing that diatom declines reflect both effects. I’m not sure one can simply isolate competition for NH4 as the main driver of the changes observed — especially given how important circulation driven changes are on a point by point basis. NO3 and NH4 uptake occur in parallel and can jointly affect diatom and other phytoplankton.Specific commentsSection 2.1: It is somewhat surprising that the model explicitly represents diatoms and nanoplankton, but not picoplankton, given the focus on competition between diatoms and cyanobacteria. The most abundant cyanobacteria in oligotrophic regions, Prochlorococcus and Synechococcus, fall within the picoplankton size range. But one could argue that the model’s nanoplankton functionally encompasses both pico- and nano-plankton. A brief discussion of this issue and the potential limitations it introduces could be included.The model description in the SI could be expanded for clarity. It would be helpful to include the equations for DIN uptake, as this is central to interpreting the results of decreased NO3 supply and understanding the distinction between model_control and model_compete experiments. Additionally, including the temperature dependence formulations for phytoplankton and zooplankton growth and grazing would help interpreting the warming-only experiments.The Tara analysis appears biased by the use of observations from depths shallower than 10 m only (lines 171-172). This likely skews the results towards phytoplankton communities adapted to relatively low NO3 and high NH4, in particular in oligotrophic regions, where phytoplankton are commonly found down to 100-200 m depths. This seems like a potentially important limitation, and could benefit from discussion. To my knowledge (but I may be mistaken), Tara Ocean also collected data from deep chlorophyll maxima. Why not including those data in the analysis, or at least consider them in a separate analysis?Section 2.5: the use of model-based fields in the analysis of observations makes me a bit uncomfortable, as it could introduce new, hard to control biases.Figure 2: I suggest including panel S1b, showing present-day NH4/DIN ratios with observations, in Fig. 2. This would reassure the reader that the model captures the basic N distribution patterns, and would help contextualize the changes shown in panel 2a (e.g., is a change by 10% large or small?) The text too could be more explicit , e.g., line 218, increase by 6%, could specify “from X% to Y%”.Lines 219-220: it would be interesting to report the % regenerated primary production and its change.The parameterizations for the pH dependence of ammonia oxidation are not well explained. What is the rationale behind including this pH dependence, and how were the specific functional forms shown in Fig. S5 and S7 chosen? Is it as simple as making the rate inversely proportional to the H+ concentration? (This seems the implication of the formulation invoking the pKa of NH4 dissociation.) What about the alternative formulation? The SI provides too little detail on these points, and more thorough explanation would help readers understanding the models sensitivities.Line 238-240, “in eutrophic regions, where coincidentally, shifts from low to higher NH4:DIN would have the greatest ecological impact”: This could be clarified. The phrasing is a bit confusing, as eutrophic regions are typically characterized by high NO3 concentrations. If NH4 becomes more dominant in these areas, does that imply they are no longer eutrophic? Or is the point that even in nutrient-rich waters, a shift in the form of available DIN from NO3 to NH4 could significantly affect community structure?Section 3.2, lines 258-270. The authors make the point that loss of diatoms was “driven by a combination of stimulated microbial metabolism (60%) and physical changes (40%), while ocean acidification had negligible effects”. While this may be accurate at a global mean level, it risks giving the misleading impression that increased microbial metabolism is the dominant driver of diatom loss across the ocean, as compared to decreased NO3 supply. But is is not true almost anywhere in the ocean, where on a local basis (Fig. S8) changes due to NO3 supply (or light availability at high latitudes) greatly exceed the low, but consistently negative effects of warming on metabolism. But because physical effects on nutrient supply and light availability are both positive and negative, they tend to cancel out when averaged globally. This distinction between local drivers and global mean effects is not clearly conveyed in the current discussion and should be more explicitly emphasized.Lines 272-282. I find this paragraph unconvincing in its current framing. Since diatoms preferentially take up NO3 (given the DIN uptake formulation), it seems straightforward that a decline in NO3 supply and concentrations would reduce diatom production and abundance, without the need to shift the emphasis to increased competition for NH4, which would be a consequence of the change. The reduction in NO3 and the resulting increased reliance on (and potential competition for) NH4 seem more like two sides of the same coin, rather than distinct mechanisms. I’m not sure that NH4 competition is the most correct or useful framework for interesting the model changes. The authors end the paragraph by stating: “when NO3 concentrations decline, competition for NH4 increases, and declines in diatom relative abundance follow”. But is the middle step, “competition for NH4 increases”, strictly necessary to explain the decline in diatoms? It might be more parsimonious to attribute the decline directly to reduced NO3 availability.The bathtub analogy (lines 277-278) should help understanding, here I found it confusing. If nutrients represent the volume in the bathtub, why is productivity described as the inflow? Productivity removes nutrients, it doesn’t add them. And why is recycling represented as the outflow? Conceptually, recycling returns nutrients to the system. Also, at steady state, one would expect inflow and outflow to be in balance. This needs some rethinking or clarification.Lines 291-294, “It is therefore possible that reductions in NO3 and resulting competition for NO3 was a major contributor to the losses of diatoms from the phytoplankton community in our simulations”. This seems like a straightforward explanation for diatom decline, given diatom’s functional dependence on DIN forms and the NO3 declines. But its placement here appears to undercut the argument made a few paragraphs earlier that emphasized competition fro NH4.Lines 294-304, the model_compete experiment is interesting, but I see it more as highlighting another side of the same coin, rather than challenging the idea that declining NO3 is the primary driver of diatom decline. Of course if diatoms had a stronger affinity for NH4, they would fare better under reduced NO3. But this doesn’t change the underlying cause of their decline, which still originates from the reduction in NO3.Line 311, “Physical changes no longer exerted a global negative effect on their total nor relative abundance”. The global mean masks large regional variability, where large positive and negative changes partially compensate each other. As shown in Fig. S8, physical changes are the main local driver, especially in the Southern Ocean, and in fact, across much of the ocean on a point-by-point basis. This nuance should be acknowledged more clearly.Lines 238-240, “determined that a large fraction of the projected declines in diatom relative abundance are due to their competitive exclusion by other phytoplankton in regions where NH4 becomes more important as a nitrogen source”: I am unconvinced that the authors have “determined” that the diatom decline is caused by competitive exclusion, and not by the overall decline in NO3.Section 3.3.1 and Fig. 4: this section provides a stimulating and valuable set of analyses and diagnostics that could serve as a useful benchmark for models. More ocean biogeochemical models should adopt this type of diagnostic approach as a standard practice for evaluating nutrient dynamics and phytoplankton competition.Lines 369-371: does it matter than in PISCES the half saturation constants for DIN uptake are not constant but a function of the phytoplankton biomass P? What are “typical conditions”? To help the reader, the DIN uptake functional forms used by the model should be presented in the SI.Section 3.3.2: this is a stimulating analysis—though at times it felt substantial enough to warrant its own standalone paper. Figures S10-S16 are rich with interesting information, but also dense. I wonder if “goodness of fit” metrics (some o which may be reported in Table S1?) could be included directly on the figures, so that the reader can quickly evaluate the skill of different models.Fig. 5: I got a bit lost in the interpretation of the figure. Not being familiar with GAMs, I am confused by the y-axis, “GAM residuals”, which could be clarified. Also, related to the threshold of 4%; in Fig. 5 nothing specifically makes the 4% threshold stand out. The text points to “rapid losses of diatoms as NH4:DIN became greater than 4%”, but I struggle to see anything special in this threshold, and overall it seems an arbitrary choice.Lines 449-453: points (1) and (3) are general knowledge; point (2) could be as well phrased as “how diatoms are outcompeted as less primary production is fueled by external NO3 inputs”, which would actually be closer to the mechanistic changes at play.Section 3.3.4, “The confounding effect of NO3”: framing the role of NO3 as a confounding factor underplays its central role in controlling diatom growth, and seems an example of reverse causation and post hoc reasoning. Given the well-established importance of physical changes on NO3 supply across the ocean, reductions in NO3 should be taken as the default or “null” mechanisms to explain the diatom declines, rather than a confounding effect. The fact that the whole statistical analysis could be based on NO3 instead of NH4:DIN as a covariate seem to undermine the entire argument made here.Line 473, “We therefore suggest that competition for NH4 directly controls diatom relative abundance”: this feels like a forced interpretation and a leap that is not fully supported by the evidence presented.Line 477-478, “Decreases in NO3 certainly affect diatom growth, but we propose that they mostly do so indirectly by shifting the regime towards intense competition for NH4.” Another sentence that seems to overstate the authors’ interpretation without enough supporting evidence, given the direct role of NO3 in diatom growth, which is explicitly built into the model’s functional formulation.Line 486-487, “but we recast the attribution of change in terms of competitive exclusion for NH4, rather than bulk nutrient declines”: see above.Technical comments:Line 66, “become a self-sustaining regime”: I would rephrase, partly because the message is a bit unclear, partly because it is hard to imagine a case where primary production in the euphotic zone does not involve an external supply of NO3, even in the most oligotrophic regions (were export still occurs and must be ultimately balanced by external nutrient supply).Line 45, “which would work to use up …” this sentence is a bit obscure, please clarify.Line 65, “there are numerous localized studies that showcase how phytoplankton taxa shift in response to changes in the composition of DIN”: it would be useful to add some references for these studies.Line 74: remove repeated “enrichment”.Line 129, “defined as those depths where total phytoplankton biomass was greater than 0.1 mmol C m-3.” This is a bit of an unorthodox definition of euphotic zone, perhaps clarify the rationale and reassure it is generally in line with other common light-based definitions (typically, 1% light levels)Lines 146-149: here a relevant reference could be Tang, et al., 2023, Earth System Science Data, which presents a compilation of nitrification rates. (Presumably data used there and compiled in Tang et al. overlap.)Line 151, I appreciate the authors reporting the units for primary production; for consistency, they could be added for the other major variables discussed (e.g., ammonia oxidation rates).Line 194: maybe provide some more detail on the spline approach. What is k, in k=3? What is “spline complexity”? Etc.Line 289: “thid” —> “this”.Line 307: should the reference be to Fig. 3e?Line 322, “by the in the”, remove “in the”.Line 356: since the exponent is fractional, I would call the function a “fractional-order” Monod function, not a quadratic (where the implied exponent is 2).Lines 358-359: add the units for the half saturation constants.Line 493: “albiet” —> “albeit”.- Line 496: “ellaboratiung” —> “elaborating”.Fig. S1 could also show total DIN, so that one has a complete view of the controls on the NH4/DIN ratio distribution.Fig. S8, right column: I’m confused by the units, the values go up to 0.2, but the units say %. Shouldn’t Fig. S8c be the same as S3a, were value go up to 20%? The figures also look a bit different, but perhaps it’s the contouring that make them look different.End of reviewCitation: https://doi.org/10.5194/egusphere-2024-3639-EC1 - AC3: 'Reply on EC1', Pearse Buchanan, 19 May 2025
Peer review completion
Journal article(s) based on this preprint
Data sets
Supplementary dataset 1: Nutrient concentration data Pearse J. Buchanan https://doi.org/10.5281/zenodo.14194938
Supplementary dataset 2: Ammonia oxidation rates Pearse J. Buchanan https://doi.org/10.5281/zenodo.14194938
Supplementary dataset 3: Coincident nutrient and regenerated to new primary production data Pearse J. Buchanan https://doi.org/10.5281/zenodo.14194938
Supplementary dataset 4: Variations in ammonia oxidation rates for pH changes. Rates normalized to a pH of 8. Pearse J. Buchanan https://doi.org/10.5281/zenodo.14194938
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 747 | 137 | 25 | 909 | 37 | 26 | 44 |
- HTML: 747
- PDF: 137
- XML: 25
- Total: 909
- Supplement: 37
- BibTeX: 26
- EndNote: 44
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Pearse J. Buchanan
Juan J. Pierella Karlusich
Robyn E. Tuerena
Roxana Shafiee
E. Malcolm S. Woodward
Chris Bowler
Alessandro Tagliabue
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(5273 KB) - Metadata XML
-
Supplement
(9518 KB) - BibTeX
- EndNote
- Final revised paper