Fine scale zooplankton distribution across the North Balearic Front during late spring

Maxime Duranson et al.

General comments

This manuscript addresses a well-studied but still relevant scientific question—the effects of oceanic fronts on biological communities—and contributes valuable insights into zooplankton dynamics in the North Balearic Front (NBF). The topic fits well within the Biogeosciences scope.

This study presents new zooplankton data from net sampling at a fine spatial and temporal scale in the NBF and adjacent waters. The data were acquired with traditional sampling methods and analysed with appropriate numerical tools. The main conclusion is that the NBF in late spring acts more as a boundary between zooplankton communities than as an accumulation zone.

My main critical concerns regard taxonomic and trophic groupings. The most significant issue lies in Table 2, which oversimplifies zooplankton diversity into broad taxonomic and trophic categories. Key problems include:

-Missing taxa: Clausocalanus and Ctenocalanus, the first being a dominant genus in the open Mediterranean Sea and the latter also co-occurring commonly and with high relative abundance (e.g., Siokou-Frangou et al., 2010). I do not see these genera in any of the categories considered in Table 2. This is an important problem that must be addressed and solved.  

-Problematic trophic categories: many behavioural studies have demonstrated that copepods do not “filter” the food particles as pelagic tunicate instead do. Some copepods create feeding currents that convey water with food particles to the mouth appendages, and Temora is one example. Acartia and Centropages instead have a mixed feeding strategy, they can switch from feeding currents to ambush predation, according to the type of prey prevailing in the environment. These three genera have been pooled in the “Copepod filter-feeders” category in Table 2, where also Pleuromamma is included. But Pleuromamma swims very fast, with a motion behavior that does not allow creating feeding currents. Oncaeidae are placed in the category of “Copepods cruise-feeders”, but these cyclopoids exhibit a "jerky, hop-and-pause" motion (Hwang & Turner, 1995) as it clearly appears from observing live oncaeids.

Given these inaccuracies, trophic groups add little value and risk misinterpretation. Therefore, I recommend to 1) remove section 3.4.3 (trophic groups) and Figure 8, 2) focus the analysis on taxonomic composition, ensuring all major taxa (including Clausocalanus and Ctenocalanus) are properly represented. If trophic roles are critical to interpret the community distribution patterns, discuss them in the Discussion section, citing behavioral literature to support functional interpretations.

Another issue regards the environmental data. The manuscript mentions that physical data were recorded with CTD profiles (line 94) but does not present or discuss them. This omission should be explicitly justified. Please clarify. Are these data being analyzed in separate studies? If so, cite the relevant papers (in prep./submitted) and briefly summarize key conditions (e.g., temperature, salinity ranges) to contextualize the zooplankton findings. If not part of another study, the authors should include basic hydrographic data (e.g., depth-resolved temperature, salinity, chlorophyll) as supplementary material, with proper attribution to the data producers.

Moreover, additional needs of improvement are:

-The title reflects the content but could be more engaging by briefly highlighting the main finding (e.g., the NBF’s role as a boundary rather than an accumulation zone).

-In the abstract, "largely unknown" (line 4) should be replaced by "still insufficiently known" for greater precision.

Regarding the language, the manuscript requires thorough English editing to improve clarity and flow. The current style is heavy, with redundancies and repetitions. The Results section is overly detailed and should be streamlined for conciseness.

In synthesis, this study provides useful data on zooplankton distribution across the NBF but would be significantly improved by a substantial revision focused on 1) refining the taxonomic resolution, 2) removing the trophic classification, 3) tightening the writing.

I will be happy to provide further clarification if needed.

Specific comments

The Introduction provides a well-referenced and comprehensive background on the topic. However, I recommend the following revisions to improve organization and clarity:

The current description of the study area (lines 51-66) is too detailed for the Introduction. Please reduce this to just a few lines that introduce the study's aims. The detailed geographical and hydrological information should be moved to a dedicated "Study Area" subsection (2.1) in Materials and Methods. This subsection should be separate from the sampling strategy and should include a map of the northwestern Mediterranean showing the key hydrological structures and the BioSWOT-Med survey area (clearly framed).

At the end of Introduction, 1) the current questions about zooplankton communities should be preceded by a brief description of the cruise's general interdisciplinary scope; 2) the zooplankton study aims should be presented with clear hypotheses rather than only questions.

Additional specific suggestions:

-Improve paragraph flow by moving the first sentence to the end of the first paragraph (better transition to paragraph 2. The second sentence works well as the new opening sentence about frontal zones.

Line 35: Avoid repetition - suggest: "...concentrate high phytoplankton abundance, supporting elevated zooplankton stocks and metabolism..."

Line 41: Simplify to "...and their predators..."

Line 44: Add DVM abbreviation at first mention: “…in zooplankton Diel Vertical Migration (DVM) have also been observed…”

Line 45-46: Clarify to "...investigating zooplankton distribution at fine scales..."

Line 46: Specify what "particles" refers to (e.g., potential prey items?)

Line 48: Clarify what "varying widths" describes (height of biomass peaks? frontal features?)

Line 67: Briefly present the cruise's interdisciplinary scope before detailing the zooplankton study aims, which should be hypothesis-driven

In Materials and Methods, the section organization should be slightly changed. For better clarity and structure, subsection 2.1 should be divided into: 2.1 Study Area, and 2.2 Sampling Strategy (not "sample strategy").

The sampling approach needs a clarification: the mention of "drifting stations" (line 89) suggests a Lagrangian sampling strategy. Please clarify this point.

The bathymetric range of the sampled area should be provided.

Include details on how the filtered water volume was measured for zooplankton tows.

I do not have specific comments on numerical analysis.

Lines 76–77: This sentence appears to be a figure caption and should either be removed or rephrased.

Line 78: Explicitly define the acronym SWOT.

Line 79: Clarify what is meant by "high spatial resolution" by providing specific values.

Line 84: Revise to: "…physical, chemical, and biological properties."

Line 85: Specify the range of "fine scale" (e.g., meters, kilometers).

Line 94: Replace "sampled" with "recorded" (CTD measures properties, does not "sample" them).

Table 1: Add a column indicating sonic depths.

Line 105: State the name and location of the shore-based laboratory.

Line 106: Clarify whether "each sample" refers to: separate samples from the 200 μm and 500 μm nets, or a merged sample combining both.

Line 109: Explain how the approximate number of individuals (~1500) was determined a priori.

Table 2: See General Comments regarding content. In addition, the trophic group "All" (including "Other organisms") is misleading; I suggest renaming it "Undefined"

Line 125: Rephrase for clarity. Were the 200 μm and 500 μm samples merged before ZooScan analysis, or were they processed separately and the counts later combined (summed)?

Subsection 2.5: The method of deriving zooplankton abundance/composition for the layers 100–200 m and 200–400 m by subtracting data from 0–100 m, 0–200 m, and 0–400 m is unconventional. While inevitable due to the sampling design and gears, this approach introduces potential errors (e.g., contamination between layers, negative abundances, as observed here for Eumalacostraca and Cnidaria). Please discuss: the limitations of this method, and how potential biases were addressed (or acknowledged).

Line 155: Why were eight copepod taxa selected as the "most abundant" rather than another number (e.g., ten)? Clarify: the percentage these eight taxa represent within the total copepod assemblage, and the rationale behind choosing this specific number.

The Results section contains excessive details, making it cumbersome to read. I recommend streamlining the text to improve clarity and flow while preserving key findings (with the exception of subsection 3.4.3, which should be reconsidered -see below).

Given the methodological concerns raised in my General Comments, I suggest focusing the community composition analysis on taxonomic groups only, removing section 3.4.3 (trophic groups) and the related Figure 8. Discussion of trophic roles, particularly for key groups influencing zooplankton distribution across water masses, should instead be addressed in the Discussion section.

Additionally, the term "intermediate" (when referring to depth layers) is unnecessary and potentially misleading; it should be removed from both the text and figure captions.

Line 239: Specify that the data refer to the 0–200 m depth layer.

Line 241: From Figure 3, it appears that the 500 μm mesh net also effectively captures Appendicularia and Chaetognatha, not just the listed taxa. Include these groups.

Line 247: Remove the speculative statement “This intermediate layer likely reflects a transitional zone where DVM results in taxonomic shifts." Such interpretations should be moved to Discussion.

Section 3.3: This section is overly verbose and should be condensed.

Lines 252–256 are redundant. The term "less structured composition" is vague, define what this means. The link to diel vertical migration (DVM) is speculative without direct evidence. Move this discussion to the Discussion section. The details on layers and Hellinger distances would be better organized in a table for clarity.

Line 395: The claim that the cruise occurred during the " the post-bloom period, when phytoplankton biomass levels are already too low to sustain optimal growth of specific zooplankton groups” lacks supporting data. Either provide referenced evidence, or remove the statement unless it can be substantiated by presenting data.

Notes on the figures:

Figure 2: The current caption is confusing and needs revision.

Clarify if "Total abundance" includes all organisms across the full size spectrum?

The caption reports ESD (Equivalent Spherical Diameter), while the text refers to ECD (Equivalent Circular Diameter) (Lines 110, 128, 132, 133). Ensure consistency.

Asterisk means "the net could not be analyzed," yet data appear in the histogram. Revise or clarify this discrepancy.

The stations located at the front should be more evidently and immediately identifiable.

 Figure 3: For easier interpretation, reorganize the histograms so that Copepoda are at the base, followed by Cnidaria, Thaliacea, and other groups.

Figures 5, 9: Add letters (a, b, c) to distinguish the three panels clearly.

Figure 6: The bars are too small and hard to distinguish. Enlarge or adjust for better readability.

Figure 8: It should be removed (see my comments above on the trophic traits).

In summary, my suggestion are: Streamline the Results for better readability; remove trophic group analysis (3.4.3) and discuss functional roles in the Discussion; improve figure clarity (captions, visibility); avoid unsupported claims (e.g., "too low phytoplankton biomass"); move speculative interpretations (e.g., DVM influence) to the Discussion.

The Discussion is well-developed but could be strengthened with more detailed insights on the species-/genus-level distribution patterns of zooplankton, which would better elucidate adaptations to different water masses. Some structural improvements are necessary.

Currently, there is unnecessary mixing of result interpretation and comparisons with previous studies (e.g., second paragraph of Section 4.1). These should be separated for clarity.

A summary table comparing zooplankton abundance/biomass with prior studies in the region would be more effective than textual descriptions.

The discussion on zooplankton biomass drivers at fronts (4.2) and the front’s role as a mixing zone vs. distinct community boundary (4.3) should be merged and condensed to avoid redundancy.

Further Comments:

Line 406: What explains the contrasting responses of Cnidaria/Foraminifera (positively influenced by the front) vs. Thaliacea (negatively affected)? An attempt of  explanation is needed.

Lines 414–415: The overly generic statement "These patterns likely result from interactions between species-specific behaviors and frontal dynamics" should be rephrased with more precise reasoning (e.g., citing known behavioral or hydrographic drivers).

Lines 423–424: Is there a hypothesis for why certain taxa (Magelonidae, cyphonautes, echinoderm larvae, radiolarians, Heteronemertea) were absent at the front? If speculative, frame it as a question for future research.

Section 4.5: The title "Storm Impact?" should be assertive (e.g., "Potential Storm Effects")

Line 461: Provide a reference for the chl a-fluorescence glider data

General technical notes

 In Methods and Results, all verbs should be in the past tense, while some are now erroneously in the present tense.

In Methods and results, some taxonomic categories are given in Latin, while others are in English. Ensure uniformity throughout the manuscript (text, tables, figures).

Maintain consistency throughout the manuscript, always writing the acronyms (which should be made explicit only at the first citation).

I sincerely apologize for a comment that I have mistakenly included in my previous review, due to an oversight on the material in the Appendix.

This part about the lack of environmental data must be removed from my review:

"Another issue regards the environmental data. The manuscript mentions that physical data were recorded with CTD profiles (line 94) but does not present or discuss them. This omission should be explicitly justified. Please clarify. Are these data being analyzed in separate studies? If so, cite the relevant papers (in prep./submitted) and briefly summarize key conditions (e.g., temperature, salinity ranges) to contextualize the zooplankton findings. If not part of another study, the authors should include basic hydrographic data (e.g., depth-resolved temperature, salinity, chlorophyll) as supplementary material, with proper attribution to the data producers."

I beg the Authors to accept my sincere apologies.

I would like to begin by apologising for the delay in providing this review. The manuscript presents a study of considerable interest to me, and I read it with enthusiasm. Unfortunately, I was not able to compile my comments in a timely manner.

Main concern

The clarity and impact of the findings are hindered by a presentation that is at times diffuse and by the inclusion of numerous analyses whose relevance is not always made explicit. The manuscript would benefit from a more focused and structured approach, with a clear emphasis on its main scientific question: the evaluation of zooplankton community responses to a frontal system (and the conclusion that the community is not just a mix of those of the surrounding water masses). Its length could probably easily be reduced by ⅓, possibly ½.

While the authors acknowledge the meteorological constraints that affected the sampling strategy, the inclusion of data from the "M" stations introduces complexity that does not clearly support the core narrative. These stations appear to add variability that confounds rather than clarifies the analysis of the frontal signal. In particular, the inclusion of M stations significantly alters multivariate patterns such as those visible in the PCAs in Figure 5 and following, making it more difficult to discern the contrasts between the A, B, and F stations, which are those associated with the frontal gradient. I would strongly recommend excluding the M stations entirely from the analyses and focusing the manuscript on the transects most relevant to the front.

Moreover, while the breadth of analyses presented is commendable, their number may overwhelm the reader and dilute the central message. A clearer selection of the most pertinent analyses to highlight your main message (e.g., PCA on concentration or biovolume, barplots, NBSS) would help maintain the reader's focus. Other analyses—such as those related to diel vertical migration or the storm—could be treated as potential confounding factors. These could then be addressed more briefly in the discussion with some plots presented in supplementary materials if necessary.

In addition, the manuscript attributes several observed differences between cruise legs to the passage of a storm. However, the current dataset does not make it possible to unambiguously separate storm effects from general temporal or spatial variability. This uncertainty should be acknowledged more explicitly in the discussion, and interpretations emphasising the storm as a dominant factor should be tempered accordingly.

Major comments

Line 113: The authors limit the analysis to eight categories derived from the ZooScan. It is unclear why only these eight were used, given that the system allows for much finer taxonomic resolution. These categories were likely aggregates of finer taxonomic groups. Why didn't you use the finer level data? Indeed, as acknowledged in lines 423–425, the inclusion of rarer taxa enhanced the ability to distinguish between water masses, indicating that finer-scale groupings may be more informative for detecting ecological responses to fronts. Coarse groupings such as "copepods" may be too ubiquitous to reveal significant patterns.

Lines ~145: The vertical distribution estimation based on two net tows at different depths, with subtraction, is methodologically weak compared to dedicated stratified sampling using e.g. a multinet. Variability in the upper layer between tows could significantly affect results. Ideally, replicate shallow tows should be presented to estimate intra-station variability and compare it with the inter-station variance. If this is not possible, a discussion of this methodological limitation should be included, supported where possible by literature.

Line 173: The use of the Hellinger transformation shifts the analytical focus to relative composition rather than absolute concentrations. This is not inherently "preferable" to the use of Euclidean distance on raw data, but represents a different analytical approach. This choice should be justified explicitly, as it influences the interpretation of the results.

Lines ~215: Since the fT stations are constructed as linear combinations of stations A and B, their positions in PCA space should likewise fall between the positions of those stations. Due to the Hellinger transformation, this may not be a straight line but a curved path. Still, the optimal mixing between a and b should be computable exactly (i.e. without iteration) from the PCA space. In any case, recalculating the PCA with each added station alters the structure of the PCA space (even if this is likely small here) and this hinders comparisons. It would be more appropriate to construct a PCA using A, B, and F stations, and then project the fT stations as supplementary points.

Section 3.3: While this section demonstrates the presence of diel vertical migration, it does not quantify its importance relative to other sources of variability. A multivariate analysis such as PCA including all stations could help illustrate whether day and night samples from the same station are more similar to one another than to samples from other stations (and you have it… so cite it here).

Section 3.4.3: This section appears to replicate earlier PCAs using trophic groups rather than taxonomic ones. It is not clear what new insight this re-analysis is intended to provide. If there is a hypothesis suggesting that trophic groups respond differently to frontal structures, it should be clearly stated. Otherwise, the patterns described may simply reflect underlying taxonomic distributions.

Lines 479 and 484: The discussion mentions the value of high-resolution observations and autonomous platforms. If I am not mistaken, a Zooglider was deployed during the BioSWOT campaign. It would be extremely valuable to discuss these data alongside the net samples. The integration of these two datasets could significantly enhance the interpretation of the observed patterns compared to studying them in isolation. I was actually expecting to read about this when I accepted the review and was disappointed to see only the net data.

Detailed comments

• The manuscript contains inconsistencies in citation formatting. References should follow the format (Author Year) rather than (Author (Year)).
• Line 36: The assertion that fronts concentrate plankton should be moderated, as this is not always the case. Indeed, you already provide a more nuanced statement just a few lines above.
• Figure 1: It would be helpful to include a contextual map showing general ocean circulation as introduced in the background. The ship’s trajectory should also be overlaid. Each subpanel should show only the relevant stations for its respective transect.
• Line ~80: A clearer description of the software tools used would be beneficial. Their specific functions and utility during the cruise should be outlined.
• Lines 86–88: This material reiterates content already provided in the introduction. It would be advisable to consolidate this information in one section and refer to Figure 1 directly from the introduction.
• Line 89: If the station drift over 24 hours is not negligible relative to the map scale, this should be depicted as a trajectory rather than as discrete points.
• Line 90: Two "f2" stations are presented: f2_D and f2_N… but the day/night distinction has not been made explicit yet.
• Lines 95–96: The notation for water masses and stations should be presented earlier to guide the reader from the outset.
• Table 1: The column listing station names does not provide essential information and could be removed.
• Line 115: The term "abundance" is used, but the actual metric is concentration (ind/m³). This should be corrected throughout the manuscript.
• Line 152: It is not clear what the "groups" are at this point. Overall, in the methods section, I would advise to always start by explaining the ecological purpose of the analyses and only then, describe (which tests, which hypotheses, etc.) you will carry them out. Currently, the rationale of the analyses is often not clear.
• Line 152: Normality should be assessed on residuals rather than raw data. If normality assumptions were not met, alternative methods should be justified. Indicate whether data transformations were applied to reach normality (it's likely that a transformation was used).
• Line 155: The aim of the copepod subgroup test should be more explicitly explained.
• Line 163: The x-axis of the NBSS should be in units of mm³.
• Line 165: The y-axis should be expressed in mm³/m³/mm³; the denominator represents Δvolume in mm³.
• Line 167: If the ellipsoid volume approximation is deemed superior, as you state at lines 135-136, explain why the spherical approximation is still used in this section.
• Line ~170: Again, the manuscript should provide the rationale ("why") for each analysis before presenting the methodology ("how").
• Line 172: Clarify that observed noise in the NBSS at large sizes is due to the rarity of large individuals rather than size per se.
• Line 178: I would suggest replacing the shorthand notation (y1+) with an explicit sum sign. Also, indicate that concentrations—not frequencies—are being summed, and define all variables (y_ij are not explicitly defined).
• Line 181: Please explain what is meant by "asymmetric" in this context.
• Line 185: Specify whether normality tests were performed before or after the Hellinger transformation. Note that PCA does not absolutely require normal data but is appropriate only with approximately normal input, so an actual normality test may be excessive. Also, please clarify the purpose of testing correlations between variables (since this seems to me that assessing correlations is what the PCA does already).
• Line 196: Provide details on how "dispersion" was calculated.
• Line 199: Was the significance of the pseudo-F statistic tested? If so, specify the method.
• Line 202: The notation "fT" is potentially ambiguous. A clearer notation such as f{t}_D, where {t} is a subscript and indicates theoretical interpolation, would help avoid confusion.
• Figure 2: Indicate in the axis title that (b) refers to the biovolume of small organisms.
• Figures 3 and 4: Consider using a more refined and perceptually balanced colour palette, such as those offered by Tableau or ColorBrewer.
• Line 252: Why is biovolume analysed here but not in the previous section? Indeed, biovolume provides a valid view of the taxonomic composition. I am not asking for an additional analysis (there are many already); rather I would recommend choosing an angle of analysis, justifying it and sticking by it. Also, the claimed similarity between different depth layers should be demonstrated using multivariate analyses (e.g., PCA based on Hellinger distances), which would better capture the structure of the data.
• Line 283: The PERMANOVA test should be described in the methods section. Clarify which factors were tested.
• Lines 293–294: "This indicates...dynamics of the water masses": I am not sure I understand what you mean. It is unclear why variation in the proportion of group A is interpreted as evidence for vertical migration. Could this not be attributed to bathymetric differences between the regions of A and B water masses for example?
• Line 299: The reconstructions of fT station values are assessed on relative concentrations only, within the PCA framework (since the Hellinger transformation was performed). You should not state that you reconstruct "absolute" concentrations.
• Lines 300 and 303: Specify what the relative increases or decreases are in reference to.
• Line 312: Avoid abbreviations such as Cop_CCF or Cni unless defined. Using full names would not be much longer but would be clearer.
• Line 314: Specify what is meant by "non-carnivorous" (e.g., non-carnivorous copepods?).
• Figure 6: If there are only three samples per station (corresponding to depth layers), boxplots may not be appropriate (a boxplot summarises a data distribution with 5 values; if you have only 3, this does not make sense). Consider an alternative method of data representation if that is indeed the case.
• Figure 9: Much of the description of these results are in terms of shifts between depth layers or between regions, but these are difficult to see since the depth layers are in different subplots. A single PCA plot with region and depth encoded by colour or symbol would facilitate interpretation.
• Line 337: Rather than referencing previous literature, the nutrient-rich nature of water mass A should be demonstrated using nutrients data collected during the cruise, if possible (I image that basic oceanography variables were collected).
• Line 369: If prior studies in the NWMS do not address the NBF specifically, this literature review may be condensed.
• Lines 370–379: These results from the literature are not clearly linked to your findings. Consider moving this paragraph later, where the discussion is more integrative.
• Line 395: Fronts may actually have their strongest effect when nutrients are limiting, such as during the normally post-bloom period of the year, when the cruise occurred. Indeed, they can then enhance nutrient availability and prolong productivity later in the season.
• Line 408: Clarify what is meant by "higher taxa."
• Line 411: "highlighting the importance of considering individual taxonomy groups rather than just overall abundance patterns when analysing community dynamics": this claim that taxonomy matters for community analysis is self-evident: community dynamics is the dynamics of various species, so, of course, it cannot be assessed with only the overall concentration. Consider removing or rephrasing.
• Line 429: The observation that zooplankton differences are stronger at 100–200 m despite the fact that the front is stronger in 0-100m is intriguing and warrants further discussion.
• Lines 431–432: "emphasises the stronger influence of hydrology and biological productivity at the surface": and of the front! The fact that the two water masses that meet at the front have a different history is also a good explanation for this observation.
• Line ~440: Diel vertical migration should be introduced early in the results as a potential confounding factor. Explain how this was controlled for/avoided (e.g., comparing only daytime samples) so that you can safely go on with the analyses despite this confounding factor.
• Line 461: Provide specific details regarding what you observed on the chlorophyll a fluorescence profiles.
• Table 3: Ensure that this table is referenced appropriately in the text. Note that Centropages typicus should be italicised and use lowercase for the species epithet.
• Line 484: Satellite data are mentioned but not utilised. If available, these should be incorporated into the analysis or explicitly discussed.