General Comments

Reviewer’s comment:

This manuscript addresses an important question regarding diatom-driven carbon export in the western North Pacific Subtropical Gyre. However, the strength of the evidence does not match the ambition of the conclusions. The study relies on very limited spatial and temporal coverage, substantial methodological assumptions, and indirect inference to support claims of a “regional hotspot” and a trait‑based mechanistic framework. Given the extremely small effective sample size—particularly the reliance on a single station (K2b) to define a “regional hotspot”—the conclusions appear overstated and insufficiently supported. In its current form, the study reads more as a preliminary case observation than as robust evidence for a regional‑scale ecological pattern.

Response: Thanks for your honest assessment. We agree that our sample size is limited and that the term “regional hotspot” was too strong. Accordingly, we have reframed the study objective to investigate whether and how diatom‑mediated carbon export varies across spatial scales, and what factors might contribute to local heterogeneity. We have substantially revised the manuscript to present our findings as preliminary support for a dual-driver framework, in which diatom community assemblages (shaped by hydrodynamics and nutrient supply) and microbial degradation resistance (via limited FCSP-targeting enzymes) might interact to create local heterogeneity in carbon export efficiency. This is presented rather than as definitive conclusions. Specifically:

• The term “regional hotspot” has been replaced throughout with “local heterogeneity” or “K2b export maximum.”
• We have added a new “Limitations of the study” subsection in the Discussion, explicitly acknowledging the small number of stations, lack of winter flux data, and methodological biases.
• All conclusions have been toned down, and we now emphasize the need for future multi-season, multi-year, and multi-station observations.
• We have also revised the title and abstract accordingly. Detailed changes are provided below.

We hope these revisions adequately address the reviewer’s concerns

Updated title

Diatom community assemblages and microbial degradation resistance shape local heterogeneity of diatom-driven carbon export in the western North Pacific Subtropical Gyre

Updated Abstract

Oligotrophic subtropical gyres are vast but uniformly nutrient‑poor and typically support very low diatom abundances, posing challenges for understanding diatom-driven carbon sequestration. It remains unclear whether and how diatom-mediated carbon export varies across spatial scales, and what factors might contribute to local heterogeneity. Here, we presented a study integrating taxonomic community analysis, sediment trap flux measurements, and metagenomic profiling to examine diatom community assemblages, export fluxes, and potential microbial degradation resistance at five stations in the western North Pacific Subtropical Gyre (NPSG). Our observations suggested that within a vertically stratified nutrient regime, diatom communities exhibit distinct niche partitioning: Navicula and Rhizosolenia tend to be enriched in the nutrient‑depleted surface mixed layer, whereas Nitzschia, Chaetoceros, and Thalassiosira prevail in the deep chlorophyll maximum, possibly reflecting hydrographic and nutrient influences on community assembly. This trait-based (referring to functional traits such as cell size, carbon content, and fucose‑containing sulfated polysaccharides (FCSPs) production) community structuring may influence the composition and magnitude of diatom carbon export, with measured intact cell fluxes ranging from to cells and estimated carbon fluxes of 0.13–194.85 . Within our limited station coverage, total diatom carbon export and export efficiency appeared highest at the Kuroshio-influenced station K2b compared to the other two stations, where large, carbon-rich Rhizosolenia dominates the sinking flux at K2b, suggesting a localized export maximum. On the other hand, metagenomic analyses indicated a widespread but taxonomically restricted capacity for degrading common diatom polysaccharides (e.g., laminarin, mannan); however, genes encoding the essential enzyme (GH107) for cleaving FCSPs are nearly absent across all stations. This deficit in microbial degradation potential, together with the prevalence of FCSP-producing diatom taxa, raises the possibility that biochemical resistance of these polymers may locally enhance particle preservation and export. Taken together, our findings provide preliminary support for a dual-driver framework in which diatom community assemblages (shaped by hydrodynamics and nutrient supply) and microbial degradation resistance (via limited FCSP-targeting enzymes) might interact to create local heterogeneity in carbon export efficiency. We conclude that predicting the biological pump’s response to global change may require accounting not only for which diatoms are present, but also for which of their organic byproducts are protected from rapid remineralization. However, it should be emphasized that these results are based on a limited number of stations and a single season, and therefore, further testing is warranted.

Specific Comment 1

Reviewer’s comment:

Only five stations were sampled in summer, and sediment trap data were successfully obtained at just three stations (K2b, M22, K8a). Winter observations were limited to two stations and lacked sediment trap measurements entirely.

Response: Thanks for your comment. We agree that the number of stations with sediment trap data is limited. Accordingly, we have reframed the study objective to investigate whether and how diatom‑mediated carbon export varies across spatial scales, and what factors might contribute to local heterogeneity. In the revised manuscript, it is believed that the three stations with trap data (K2b, M22, K8a) are suitable for exploring local heterogeneity, given they represent distinct hydrodynamic regimes (Kuroshio, gyre interior, and North Equatorial Current, respectively). The oligotrophic North Pacific Subtropical Gyre is vast but uniformly nutrient‑poor, with very low diatom abundances. Under these constraints, focusing on representative stations allows a meaningful comparison of local export patterns. The following modifications have been made:

• In Section 2.2, we state the reasons for the unsuccessful deployment of traps in winter (due to rough sea conditions).
• Throughout the Abstract, Results and Discussion, we have replaced terms like “regional heterogeneity” with “local heterogeneity” or “observed differences among the three stations,” and added qualifiers such as “within the limited spatial scope.”
• In the Conclusions and Abstract, we now add: “Due to the limited number of stations and lack of winter flux data, our findings should be considered a preliminary study, and we emphasize the need for future multi-season, multi-year observations, and multi-station observations”, “It should be emphasized that these results are based on a limited number of stations and a single season, and therefore further testing is warranted.”

Specific Comment 2

Reviewer’s comment:

Carbon biomass was estimated using genus‑level average carbon content, largely derived from literature compilations and biovolume‑to‑carbon conversions. Oligotrophic regions are typically dominated by small and fragile diatom taxa, whose cellular carbon content may deviate substantially from literature averages derived under different conditions.

Response: Thanks for your valid concern. We acknowledge that estimating carbon biomass from genus‑level averages and biovolume conversions introduces uncertainty, especially in oligotrophic systems. Direct measurement of diatom carbon content in such low‑biomass waters is technically very challenging, often requiring impractically large sample volumes. Given this constraint, our approach follows widely used methods for oligotrophic regions. To address the uncertainty, we have:

• Performed a sensitivity analysis (±30% carbon variation) and added the resulting flux ranges in Table S8.
• Estimated 95% confidence intervals via bootstrap (Table S2).
• Explicitly discussed the limitation in Sections 2.5 and 4.6, and clarified that absolute values should be interpreted with caution while relative station comparisons remain robust.

We hope these revisions adequately address the concern.

Specific Comment 3

Reviewer’s comment:

Diatom counts were conducted after sedimentation concentration and 8 μm mesh filtration. The authors acknowledge that this procedure may underestimate large or fragile taxa such as Rhizosolenia. However, Rhizosolenia is precisely the dominant contributor to the proposed K2b “hotspot”.

Response: Thanks for your raising this critical point. We have made the following changes to address the potential bias caused by the underestimation of large/fragile taxa such as Rhizosolenia:

• In Section 2.4, we now explicitly state that the sedimentation concentration method underestimates large/fragile taxa, and that our reported water-column abundance of Rhizosolenia is likely lower than true values.
• Rhizosolenia is also dominant in sediment trap samples (which were also processed by concentration, but with fewer concentration steps than the water samples, potentially resulting in less breakage of Rhizosolenia frustules). We discuss the potential bias on export efficiency: underestimation of water‑column Rhizosolenia would lead to an overestimation of the DEP/DPP ratio at K2b. We now caution that this ratio should be considered an upper bound.
• To provide a more balanced range, we additionally counted broken Rhizosolenia frustules in the water‑column samples and included them as whole cells. This approach yields a lower bound for the DEP/DPP ratio at K2b. The revised range (lower to upper bound) is now presented in the results and discussed in Section 3.5.
• This methodological limitation is also highlighted in the new “Limitations” subsection (Section 4.6).

We believe these revisions transparently address your concern and provide a more robust interpretation of our export efficiency estimates.

Additional changes

• A new “Limitations of the study” subsection has been added to the Discussion, covering small sample size, single season, methodological biases (8‑μm filtration, carbon conversion), and the indirect nature of FCSP degradation evidence.
• The abstract, results, discussion, and conclusions have been rewritten to reflect a more cautious, hypothesis‑generating tone.

We believe these revisions substantially address your concerns and hope the manuscript is now acceptable for publication.

Response to Reviewer 2

General comment

In their paper ‘Trait-based mechanisms underpin regional hotspot of diatom-driven carbon export in an oligotrophic gyre’ the authors seek to connect the vertical structure of diatom populations in the western North Pacific Subtropical Gyre to the magnitude and composition of export flux. The topic is really interesting as the drivers of export flux in the oligotrophic ocean remain open for discussion. The main strengths of the manuscript are the dataset and the ability to interface the vertical structure of the populations with flux observations. It is difficult to generate a dataset like this, despite the small number of stations. Unfortunately, the current claims in the manuscript are extremely strong and not directly supported by the data. However, part of this could be the presentation. The figures were difficult to connect to the main text because of the many small squares and the lack of direct connection between the visuals and the processes discussed in the main text. The manuscript really needs a set of clear figures which show the proposed mechanisms and connect to the oceanographic context of the problem.

Response: Thanks for recognizing the interest and strength of our dataset, as well as the inherent difficulty of collecting such measurements in the oligotrophic ocean. We also appreciate the candid assessment that our original claims were too strong relative to the data, and that the figures were difficult to interpret. In response, we have substantially revised the manuscript to address these concerns, specifically by toning down our conclusions and claims and improving the figures to better connect with the text and the proposed mechanisms. We are grateful for the reviewer’s constructive feedback, which has significantly improved the clarity and rigor of our manuscript.

Comment 1 (Lines 26-27)

Lines 26-27: It is unclear what is meant by ‘trait-based’. Are you claiming ‘depth’ is a trait? If so, I normally think about traits in connection to nutrient uptake etc.

Response: Thanks for your comments. We have revised the abstract and main text to define “trait-based” explicitly as functional biological traits, for example, cell size, carbon content, aggregation potential, and FCSP production. Depth is not a trait. The revised abstract now reads: “This trait-based (referring to functional traits such as cell size, carbon content, and fucose‑containing sulfated polysaccharides (FCSPs) production) community structuring may influence the composition and magnitude of diatom carbon export.”

Comment 2 (Lines 32-36)

Lines 32-36: This seems out of place with the rest of the abstract.

Response: Thanks for this observation. The abstract has been reorganized to improve its logical flow. Specifically, we now define “trait‑based” early in the abstract as referring to functional traits such as cell size, carbon content, and fucose‑containing sulfated polysaccharides (FCSPs) production. In addition, we have revised the sentence in question to read: “On the other hand, metagenomic analyses indicate a widespread but taxonomically restricted capacity for degrading common diatom polysaccharides (e.g., laminarin, mannan); however, genes encoding the essential enzyme (GH107) for cleaving FCSPs are nearly absent across all stations.” These revisions integrate the sentence more naturally with the surrounding text, and we believe it is now well placed within the abstract.

Updated title

Diatom community assemblages and microbial degradation resistance shape local heterogeneity of diatom-driven carbon export in the western North Pacific Subtropical Gyre

Updated Abstract

Oligotrophic subtropical gyres are vast but uniformly nutrient‑poor and typically support very low diatom abundances, posing challenges for understanding diatom-driven carbon sequestration. It remains unclear whether and how diatom-mediated carbon export varies across spatial scales, and what factors might contribute to local heterogeneity. Here, we presented a study integrating taxonomic community analysis, sediment trap flux measurements, and metagenomic profiling to examine diatom community assemblages, export fluxes, and potential microbial degradation resistance at five stations in the western North Pacific Subtropical Gyre (NPSG). Our observations suggested that within a vertically stratified nutrient regime, diatom communities exhibit distinct niche partitioning: Navicula and Rhizosolenia tend to be enriched in the nutrient‑depleted surface mixed layer, whereas Nitzschia, Chaetoceros, and Thalassiosira prevail in the deep chlorophyll maximum, possibly reflecting hydrographic and nutrient influences on community assembly. This trait-based (referring to functional traits such as cell size, carbon content, and fucose‑containing sulfated polysaccharides (FCSPs) production) community structuring may influence the composition and magnitude of diatom carbon export, with measured intact cell fluxes ranging from to cells and estimated carbon fluxes of 0.13–194.85 . Within our limited station coverage, total diatom carbon export and export efficiency appeared highest at the Kuroshio-influenced station K2b compared to the other two stations, where large, carbon-rich Rhizosolenia dominates the sinking flux at K2b, suggesting a localized export maximum. On the other hand, metagenomic analyses indicated a widespread but taxonomically restricted capacity for degrading common diatom polysaccharides (e.g., laminarin, mannan); however, genes encoding the essential enzyme (GH107) for cleaving FCSPs are nearly absent across all stations. This deficit in microbial degradation potential, together with the prevalence of FCSP-producing diatom taxa, raises the possibility that biochemical resistance of these polymers may locally enhance particle preservation and export. Taken together, our findings provide preliminary support for a dual-driver framework in which diatom community assemblages (shaped by hydrodynamics and nutrient supply) and microbial degradation resistance (via limited FCSP-targeting enzymes) might interact to create local heterogeneity in carbon export efficiency. We conclude that predicting the biological pump’s response to global change may require accounting not only for which diatoms are present, but also for which of their organic byproducts are protected from rapid remineralization. However, it should be emphasized that these results are based on a limited number of stations and a single season, and therefore, further testing is warranted.

Comment 3 (Lines 36-39) – need a mechanism

Lines 36-39: There needs to be a mechanism to make this convincing.

Response: Thanks for this comment. We now explicitly state a mechanism in the updated Abstract and Section 4.3. The relevant sentence in the abstract has been revised to: “Taken together, our findings provide preliminary support for a dual-driver framework in which diatom community assemblages (shaped by hydrodynamics and nutrient supply) and microbial degradation resistance (via limited FCSP‑targeting enzymes) might interact to create local heterogeneity in carbon export efficiency.” The updated Section 4.3 further breaks down each step of the mechanism. We believe this addresses the concern.

Comment 4 (Lines 39-41)

Lines 39-41: It is unclear how the study does this, and if it did, what the ‘interacting controls’ are.

Response: Thanks for your pointing out the lack of clarity regarding “interacting controls.” To address this, we have removed this ambiguous term from the manuscript and toned down the conclusions accordingly. The relevant sentence in the Abstract has been revised to: “We conclude that predicting the biological pump’s response to global change may require accounting not only for which diatoms are present, but also for which of their organic byproducts are protected from rapid remineralization. However, it should be emphasized that these results are based on a limited number of stations and a single season, and therefore, further testing is warranted.” We believe this revision eliminates the confusion and presents a more cautious, evidence‑aligned conclusion.

Comment 5 (Lines 72-87)

Lines 72-87: There has been a lot of work out of the Hawaii Ocean Timeseries looking at export with a focus on Nitrogen Fixation. The authors really need to engage with these ideas I think.

Response: Thanks for your comment. We have expanded engagement with the Hawaii Ocean Time‑series (HOT). A new paragraph in the introduction explicitly cites Karl et al. (2012) and others on diatom‑diazotroph associations (DDAs). We now compare our western NPSG findings with those from Station ALOHA.

Comment 6 (Section 2.1)

Section 2.1 – Given there are only 5 stations, more care should be provided to explain why they were chosen at these sites.

Response: Thanks for your comment. A detailed justification has been added to Section 2.1, including: representation of different hydrodynamic regimes, availability of historical Chla profiles, and logistical feasibility for sediment traps. A new table (Table S1) summarizes the rationale for each station.

Comment 7 (Lines 123-126)

Lines 123-126: Given the lack of citations, more care should be given to discussion of design.

Response: Thanks for your comment. Relevant methodological citations have been added.

Comment 8 (Line 136) – nutrient flux data underutilized

Line 136: Nutrient fluxes etc. have strong importance to the plankton community. It is confusing why this manuscript was not leveraged more.

Response: Thanks for your comment. We agree that nutrient flux data are highly relevant and should be better integrated. Accordingly, we have now more thoroughly incorporated the nutrient flux data from Du et al. (2024) throughout the manuscript. Specifically:

• In Section 3.1 and Fig. 2, we explicitly link nutrient gradients to diatom community structure.
• Spearman correlations between nutrient fluxes and dominant diatom taxa are presented in Fig. 5.
• In the Discussion (Section 4.2), we have added further interpretation of how nutrient supply may shape the observed community composition and export potential at different stations.

Comment 9 (Lines 147-158)

Lines 147-158: The description was confusing to read.

Response: Thanks for your comment. The section has been completely rewritten with step-by-step numbering. A simplified flow diagram (new Fig. S1) has been added.

Comment 10 (Lines 162-164)

Lines 162-164: cite references

Response: Thanks for your comment. Standard taxonomic references have been added.

Comment 11 (Lines 182-194)

Lines 182-194: Is there any estimate as to the error bounds?

Response: Thanks for your comment. Uncertainty estimates (bootstrap 95% confidence intervals) have been added to Table S2, and the method section now describes the uncertainty propagation.

Comment 12 (Lines 203-210)

Lines 203-210: The methods, and what they conceptually mean needs to be explained clearly.

Response: Thanks for your comment. The section has been rewritten to explain the ecological meaning of richness, Simpson index, and Pielou evenness, and how they relate to export stability.

Comment 13 (Fig. 2)

Fig 2: It is really hard to decipher this plot. You need smaller numbers of panels and representative profiles to make your points clear.

Response: Thanks for your comment. Fig. 2 has been simplified to show only representative stations (K2b, M22, K8a) and key parameters (T, S, NO₃⁻, Chl a). Full profiles are moved to new Fig. S2.

Comment 14 (Fig. 3)

Fig 3: Again, this is very hard to decipher. The NMSDS plot is really busy and it is hard to see the conclusions being drawn. All plots have all stations and depths. The authors need to think carefully about presentation and the point they are trying to make. The remaining figures share these problems.

Response: Thanks for your comment. To improve clarity, Fig. 3 has been completely redesigned:

• The NMDS plot (originally Fig. 3a) has been removed, as it was overly complex and did not clearly convey the main conclusions.
• The abundance and estimated carbon biomass panels are now faceted by depth (5 m, DCM, 200 m) to allow easier comparison across stations within each depth layer.
• The heatmap has been simplified to show only the dominant diatom genera; rare taxa are aggregated into an “others” category.

Comment 15 (Section 3.3)

Section 3.3: Many things co-vary with depth. This section is confusing as it is unclear which data is used in the correlations. If all data is used unmixing the actual drivers requires careful consideration as many things vary with depth.

Response: Thanks for your comment. We now perform Spearman correlations separately for each depth layer (5 m, DCM, 200 m). This is stated in the Methods and reflected in Fig. 5.

Comment 16 (Lines 402-409)

Lines 402-409: Do flux ratios match composition ratios?

Response: Thanks for your comment. A new analysis comparing water-column and trap compositions has been added (new Fig. S6). Rhizosolenia is overrepresented in traps relative to water, consistent with its higher sinking velocity.

Comment 17 (Lines 412-417)

Lines 412-417: This all needs to be explained more.

Response: Thanks for your comment. Section 3.5 now clearly defines DEP and DPP with examples. A step-by-step calculation is provided in the text and Table 1.

Comment 18 (Lines 421-423)

Lines 421-423: I do not believe that this claim is supported. The system is heterogeneous both in space and time. This paper really needs clear biophysical reasoning to work.

Response: Thanks for your comment. We have softened the claim. The revised text reads: “In the three stations studied, diatom carbon export efficiency was not directly correlated with total POC flux, suggesting that different factors may regulate them. This observation requires further testing with larger datasets.”

Comment 19 (Lines 483-488)

Lines 483-488: Is this claim derived from the data in this paper? If so, there is not enough data to claim this. However, I believe this is well established, starting with Karl et al 2012, the references within it and the bank of work which cites it.

Response: Thanks for your comment. We agree that this claim is not directly supported by our data and is already well established in the literature (e.g., Karl et al., 2012). We revise the text to remove the claim and instead cite previous work as the basis for this understanding. Our observations are now presented as consistent with, rather than evidence for, this pattern.

Comment 20 (Lines 493-495) – vertical structuring is well established

Lines 493-495: I believe that the vertical structuring of diatom populations is well established.

Response: Thanks for your comment. Revised to: “Our results confirm previously described vertical niche differentiation (e.g., Dore et al., 2008) and, importantly, link it directly to the composition of export flux — a less documented relationship.”

Comment 21 (Section 4.3) – define “trait‑based”

Section 4.3: The authors need to define what they mean by ‘trait-based’.

Response: Thanks for your comment. A definition has been added at the beginning of Section 4.3: “By ‘trait‑based’, we refer to species‑specific functional attributes including cell size, carbon content, and biopolymer production (e.g., FCSPs).”

Comment 22 (Lines 509-511) – nitrogen fixation appears too late

Lines 509-511: It is unclear to me why nitrogen fixation makes its first appearance in the manuscript here. The standard argument for enhanced export in the NPSG comes from invoking DDAs.

Response: Thanks for your comment. The introduction now includes a sentence on DDAs and their known role in the NPSG. Rhizosolenia is linked to Trichodesmium earlier in the Results (Section 3.1).

Comment 23 (Line 547) – carbon density vs. cell size

Lines 547: Is it true that Rhizosolenia has a large carbon density or is this due to the giant cell size?

Response: Thanks for your comment. We have corrected “high carbon density” to “high cellular carbon content due to its large cell volume” throughout.

Comment 24 (Line 562)

Line 562: The authors really need to work harder to connect their data to the mechanisms invoked.

Response: Thanks for your comment. A new summary table has been added, linking each proposed mechanism to specific data sources (metagenomics, traps, taxonomy). The discussion now walks through each causal step explicitly.

Additional changes

• A new “Limitations of the study” subsection has been added to the Discussion, covering small sample size, single season, methodological biases (8‑μm filtration, carbon conversion), and the indirect nature of FCSP degradation evidence.
• All figures have been simplified or reorganized as described above; the revised figure set is provided in the supplementary materials.
• The Abstract, Introduction, Results, and Conclusions have been rewritten to reflect a more cautious, hypothesis‑generating tone.

We believe these revisions substantially address your concerns and hope the manuscript is now acceptable for publication.

Response to Reviewer 2

General comment

In their paper ‘Trait-based mechanisms underpin regional hotspot of diatom-driven carbon export in an oligotrophic gyre’ the authors seek to connect the vertical structure of diatom populations in the western North Pacific Subtropical Gyre to the magnitude and composition of export flux. The topic is really interesting as the drivers of export flux in the oligotrophic ocean remain open for discussion. The main strengths of the manuscript are the dataset and the ability to interface the vertical structure of the populations with flux observations. It is difficult to generate a dataset like this, despite the small number of stations. Unfortunately, the current claims in the manuscript are extremely strong and not directly supported by the data. However, part of this could be the presentation. The figures were difficult to connect to the main text because of the many small squares and the lack of direct connection between the visuals and the processes discussed in the main text. The manuscript really needs a set of clear figures which show the proposed mechanisms and connect to the oceanographic context of the problem.

Response: Thanks for recognizing the interest and strength of our dataset, as well as the inherent difficulty of collecting such measurements in the oligotrophic ocean. We also appreciate the candid assessment that our original claims were too strong relative to the data, and that the figures were difficult to interpret. In response, we have substantially revised the manuscript to address these concerns, specifically by toning down our conclusions and claims and improving the figures to better connect with the text and the proposed mechanisms. We are grateful for the reviewer’s constructive feedback, which has significantly improved the clarity and rigor of our manuscript.

Comment 1 (Lines 26-27)

Lines 26-27: It is unclear what is meant by ‘trait-based’. Are you claiming ‘depth’ is a trait? If so, I normally think about traits in connection to nutrient uptake etc.

Response: Thanks for your comments. We have revised the abstract and main text to define “trait-based” explicitly as functional biological traits, for example, cell size, carbon content, aggregation potential, and FCSP production. Depth is not a trait. The revised abstract now reads: “This trait-based (referring to functional traits such as cell size, carbon content, and fucose‑containing sulfated polysaccharides (FCSPs) production) community structuring may influence the composition and magnitude of diatom carbon export.”

Comment 2 (Lines 32-36)

Lines 32-36: This seems out of place with the rest of the abstract.

Response: Thanks for this observation. The abstract has been reorganized to improve its logical flow. Specifically, we now define “trait‑based” early in the abstract as referring to functional traits such as cell size, carbon content, and fucose‑containing sulfated polysaccharides (FCSPs) production. In addition, we have revised the sentence in question to read: “On the other hand, metagenomic analyses indicate a widespread but taxonomically restricted capacity for degrading common diatom polysaccharides (e.g., laminarin, mannan); however, genes encoding the essential enzyme (GH107) for cleaving FCSPs are nearly absent across all stations.” These revisions integrate the sentence more naturally with the surrounding text, and we believe it is now well placed within the abstract.

Updated title

Diatom community assemblages and microbial degradation resistance shape local heterogeneity of diatom-driven carbon export in the western North Pacific Subtropical Gyre

Updated Abstract

Oligotrophic subtropical gyres are vast but uniformly nutrient‑poor and typically support very low diatom abundances, posing challenges for understanding diatom-driven carbon sequestration. It remains unclear whether and how diatom-mediated carbon export varies across spatial scales, and what factors might contribute to local heterogeneity. Here, we presented a study integrating taxonomic community analysis, sediment trap flux measurements, and metagenomic profiling to examine diatom community assemblages, export fluxes, and potential microbial degradation resistance at five stations in the western North Pacific Subtropical Gyre (NPSG). Our observations suggested that within a vertically stratified nutrient regime, diatom communities exhibit distinct niche partitioning: Navicula and Rhizosolenia tend to be enriched in the nutrient‑depleted surface mixed layer, whereas Nitzschia, Chaetoceros, and Thalassiosira prevail in the deep chlorophyll maximum, possibly reflecting hydrographic and nutrient influences on community assembly. This trait-based (referring to functional traits such as cell size, carbon content, and fucose‑containing sulfated polysaccharides (FCSPs) production) community structuring may influence the composition and magnitude of diatom carbon export, with measured intact cell fluxes ranging from to cells and estimated carbon fluxes of 0.13–194.85 . Within our limited station coverage, total diatom carbon export and export efficiency appeared highest at the Kuroshio-influenced station K2b compared to the other two stations, where large, carbon-rich Rhizosolenia dominates the sinking flux at K2b, suggesting a localized export maximum. On the other hand, metagenomic analyses indicated a widespread but taxonomically restricted capacity for degrading common diatom polysaccharides (e.g., laminarin, mannan); however, genes encoding the essential enzyme (GH107) for cleaving FCSPs are nearly absent across all stations. This deficit in microbial degradation potential, together with the prevalence of FCSP-producing diatom taxa, raises the possibility that biochemical resistance of these polymers may locally enhance particle preservation and export. Taken together, our findings provide preliminary support for a dual-driver framework in which diatom community assemblages (shaped by hydrodynamics and nutrient supply) and microbial degradation resistance (via limited FCSP-targeting enzymes) might interact to create local heterogeneity in carbon export efficiency. We conclude that predicting the biological pump’s response to global change may require accounting not only for which diatoms are present, but also for which of their organic byproducts are protected from rapid remineralization. However, it should be emphasized that these results are based on a limited number of stations and a single season, and therefore, further testing is warranted.

Comment 3 (Lines 36-39) – need a mechanism

Lines 36-39: There needs to be a mechanism to make this convincing.

Response: Thanks for this comment. We now explicitly state a mechanism in the updated Abstract and Section 4.3. The relevant sentence in the abstract has been revised to: “Taken together, our findings provide preliminary support for a dual-driver framework in which diatom community assemblages (shaped by hydrodynamics and nutrient supply) and microbial degradation resistance (via limited FCSP‑targeting enzymes) might interact to create local heterogeneity in carbon export efficiency.” The updated Section 4.3 further breaks down each step of the mechanism. We believe this addresses the concern.

Comment 4 (Lines 39-41)

Lines 39-41: It is unclear how the study does this, and if it did, what the ‘interacting controls’ are.

Response: Thanks for your pointing out the lack of clarity regarding “interacting controls.” To address this, we have removed this ambiguous term from the manuscript and toned down the conclusions accordingly. The relevant sentence in the Abstract has been revised to: “We conclude that predicting the biological pump’s response to global change may require accounting not only for which diatoms are present, but also for which of their organic byproducts are protected from rapid remineralization. However, it should be emphasized that these results are based on a limited number of stations and a single season, and therefore, further testing is warranted.” We believe this revision eliminates the confusion and presents a more cautious, evidence‑aligned conclusion.

Comment 5 (Lines 72-87)

Lines 72-87: There has been a lot of work out of the Hawaii Ocean Timeseries looking at export with a focus on Nitrogen Fixation. The authors really need to engage with these ideas I think.

Response: Thanks for your comment. We have expanded engagement with the Hawaii Ocean Time‑series (HOT). A new paragraph in the introduction explicitly cites Karl et al. (2012) and others on diatom‑diazotroph associations (DDAs). We now compare our western NPSG findings with those from Station ALOHA.

Comment 6 (Section 2.1)

Section 2.1 – Given there are only 5 stations, more care should be provided to explain why they were chosen at these sites.

Response: Thanks for your comment. A detailed justification has been added to Section 2.1, including: representation of different hydrodynamic regimes, availability of historical Chla profiles, and logistical feasibility for sediment traps. A new table (Table S1) summarizes the rationale for each station.

Comment 7 (Lines 123-126)

Lines 123-126: Given the lack of citations, more care should be given to discussion of design.

Response: Thanks for your comment. Relevant methodological citations have been added.

Comment 8 (Line 136) – nutrient flux data underutilized

Line 136: Nutrient fluxes etc. have strong importance to the plankton community. It is confusing why this manuscript was not leveraged more.

Response: Thanks for your comment. We agree that nutrient flux data are highly relevant and should be better integrated. Accordingly, we have now more thoroughly incorporated the nutrient flux data from Du et al. (2024) throughout the manuscript. Specifically:

• In Section 3.1 and Fig. 2, we explicitly link nutrient gradients to diatom community structure.
• Spearman correlations between nutrient fluxes and dominant diatom taxa are presented in Fig. 5.
• In the Discussion (Section 4.2), we have added further interpretation of how nutrient supply may shape the observed community composition and export potential at different stations.

Comment 9 (Lines 147-158)

Lines 147-158: The description was confusing to read.

Response: Thanks for your comment. The section has been completely rewritten with step-by-step numbering. A simplified flow diagram (new Fig. S1) has been added.

Comment 10 (Lines 162-164)

Lines 162-164: cite references

Response: Thanks for your comment. Standard taxonomic references have been added.

Comment 11 (Lines 182-194)

Lines 182-194: Is there any estimate as to the error bounds?

Response: Thanks for your comment. Uncertainty estimates (bootstrap 95% confidence intervals) have been added to Table S2, and the method section now describes the uncertainty propagation.

Comment 12 (Lines 203-210)

Lines 203-210: The methods, and what they conceptually mean needs to be explained clearly.

Response: Thanks for your comment. The section has been rewritten to explain the ecological meaning of richness, Simpson index, and Pielou evenness, and how they relate to export stability.

Comment 13 (Fig. 2)

Fig 2: It is really hard to decipher this plot. You need smaller numbers of panels and representative profiles to make your points clear.

Response: Thanks for your comment. Fig. 2 has been simplified to show only representative stations (K2b, M22, K8a) and key parameters (T, S, NO₃⁻, Chl a). Full profiles are moved to new Fig. S2.

Comment 14 (Fig. 3)

Fig 3: Again, this is very hard to decipher. The NMSDS plot is really busy and it is hard to see the conclusions being drawn. All plots have all stations and depths. The authors need to think carefully about presentation and the point they are trying to make. The remaining figures share these problems.

Response: Thanks for your comment. To improve clarity, Fig. 3 has been completely redesigned:

• The NMDS plot (originally Fig. 3a) has been removed, as it was overly complex and did not clearly convey the main conclusions.
• The abundance and estimated carbon biomass panels are now faceted by depth (5 m, DCM, 200 m) to allow easier comparison across stations within each depth layer.
• The heatmap has been simplified to show only the dominant diatom genera; rare taxa are aggregated into an “others” category.

Comment 15 (Section 3.3)

Section 3.3: Many things co-vary with depth. This section is confusing as it is unclear which data is used in the correlations. If all data is used unmixing the actual drivers requires careful consideration as many things vary with depth.

Response: Thanks for your comment. We now perform Spearman correlations separately for each depth layer (5 m, DCM, 200 m). This is stated in the Methods and reflected in Fig. 5.

Comment 16 (Lines 402-409)

Lines 402-409: Do flux ratios match composition ratios?

Response: Thanks for your comment. A new analysis comparing water-column and trap compositions has been added (new Fig. S6). Rhizosolenia is overrepresented in traps relative to water, consistent with its higher sinking velocity.

Comment 17 (Lines 412-417)

Lines 412-417: This all needs to be explained more.

Response: Thanks for your comment. Section 3.5 now clearly defines DEP and DPP with examples. A step-by-step calculation is provided in the text and Table 1.

Comment 18 (Lines 421-423)

Lines 421-423: I do not believe that this claim is supported. The system is heterogeneous both in space and time. This paper really needs clear biophysical reasoning to work.

Response: Thanks for your comment. We have softened the claim. The revised text reads: “In the three stations studied, diatom carbon export efficiency was not directly correlated with total POC flux, suggesting that different factors may regulate them. This observation requires further testing with larger datasets.”

Comment 19 (Lines 483-488)

Lines 483-488: Is this claim derived from the data in this paper? If so, there is not enough data to claim this. However, I believe this is well established, starting with Karl et al 2012, the references within it and the bank of work which cites it.

Response: Thanks for your comment. We agree that this claim is not directly supported by our data and is already well established in the literature (e.g., Karl et al., 2012). We revise the text to remove the claim and instead cite previous work as the basis for this understanding. Our observations are now presented as consistent with, rather than evidence for, this pattern.

Comment 20 (Lines 493-495) – vertical structuring is well established

Lines 493-495: I believe that the vertical structuring of diatom populations is well established.

Response: Thanks for your comment. Revised to: “Our results confirm previously described vertical niche differentiation (e.g., Dore et al., 2008) and, importantly, link it directly to the composition of export flux — a less documented relationship.”

Comment 21 (Section 4.3) – define “trait‑based”

Section 4.3: The authors need to define what they mean by ‘trait-based’.

Response: Thanks for your comment. A definition has been added at the beginning of Section 4.3: “By ‘trait‑based’, we refer to species‑specific functional attributes including cell size, carbon content, and biopolymer production (e.g., FCSPs).”

Comment 22 (Lines 509-511) – nitrogen fixation appears too late

Lines 509-511: It is unclear to me why nitrogen fixation makes its first appearance in the manuscript here. The standard argument for enhanced export in the NPSG comes from invoking DDAs.

Response: Thanks for your comment. The introduction now includes a sentence on DDAs and their known role in the NPSG. Rhizosolenia is linked to Trichodesmium earlier in the Results (Section 3.1).

Comment 23 (Line 547) – carbon density vs. cell size

Lines 547: Is it true that Rhizosolenia has a large carbon density or is this due to the giant cell size?

Response: Thanks for your comment. We have corrected “high carbon density” to “high cellular carbon content due to its large cell volume” throughout.

Comment 24 (Line 562)

Line 562: The authors really need to work harder to connect their data to the mechanisms invoked.

Response: Thanks for your comment. A new summary table has been added, linking each proposed mechanism to specific data sources (metagenomics, traps, taxonomy). The discussion now walks through each causal step explicitly.

Additional changes

• A new “Limitations of the study” subsection has been added to the Discussion, covering small sample size, single season, methodological biases (8‑μm filtration, carbon conversion), and the indirect nature of FCSP degradation evidence.
• All figures have been simplified or reorganized as described above; the revised figure set is provided in the supplementary materials.
• The Abstract, Introduction, Results, and Conclusions have been rewritten to reflect a more cautious, hypothesis‑generating tone.

We believe these revisions substantially address your concerns and hope the manuscript is now acceptable for publication.