Macrofaunal Communities Associated with Ecklonia radiata Holdfast Along the Southeast Coast of South Africa

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This study represents a significant contribution as it highlights the importance of kelp forests as refuges for fauna, especially in a less-studied region like South Africa. The topic is well introduced, the references are relevant, and the objectives are clearly defined. However, there are major concerns, mainly the incomplete or missing statistical analyses, an excessive number of figures, and a disorganized Discussion section that includes unnecessary data and comments unrelated to the results.

Below are some observations and suggestions, which I hope will help improve the study:

Materials and Methods (M&M)

-. This section appears disorganized, includes irrelevant information, and describes analyses that are not presented in the results. It seems to have been adapted from a similar study. Specific comments follow:

-. Figure 2.1: The key should arrange sites geographically (e.g., east to west) rather than alphabetically, as the current order might confuse readers unfamiliar with the area. Additionally, the Xhorha site is not shown on the map. Why are coastal hotels included? This information does not seem relevant to the study.

-. Line 135: pH, nitrate, phosphate, and sodium were measured, but their use and purpose in the study are unclear. Only variables relevant to the study should be mentioned, even if recorded with a multiparameter probe.

-. Lines 159-168: Laboratory processing details are overly detailed and should be summarized. For example: “All organisms retained on a 0.5 mm sieve screen were sorted and preserved in 70% ethanol. Most animals were identified to the highest possible taxonomic resolution.”

-. Lines 175-177: This sentence describes sampling design, not laboratory processing.

-. Lines 187-193: PCA and PERMANOVA based on Euclidean distances are described here but are missing in the results section!!.

-. Lines 200-202: This information is unnecessary for understanding the study.

-. Sections 2.2.3 to 2.2.6: These subsections should be rewritten to remove unnecessary information. The methodology should flow logically, from simpler analyses (univariate) to more complex ones (multivariate). Currently, dbRDA is mentioned in both 2.2.3 and 2.2.6. Consider merging all subsections into one for clarity. Writing errors such as “holdfast across site PRIMER (Clarke and Gorley, 2015)” and “Environmental and Ecological ASSOCIATIONS” should also be corrected.

Results

-. “Spatiotemporal Variation in Environmental Factors Across Sites”: Statistical tests are needed to verify potential differences in all variables. For instance, two-way ANOVA (or Kruskal-Wallis) could identify significant differences, aligning with the study's objectives. Alternatively, a univariate PERMANOVA based on Euclidean distance could be used to avoid ANOVA assumptions.

-. Lines 258-267: Figure 2.2 is not cited in the text.

-. Lines 270-280 (Salinity): Potential salinity differences between sites or times seem exaggerated in Figure 2.3 due to the chosen scale. A table presenting means and standard deviations may be more suitable, which applies to other environmental variables as well. This would also reduce redundant figures.

-. Figure 2.3: Are data available only for 2023 at four of the sites?

-. Lines 298-312: Figure 2.5 is not properly cited; only placeholders like “Fig. A” and “Fig. B” are mentioned. Statistical tests are also necessary to verify potential differences.

-. Sections 3.1.1 to 3.1.5: These should be combined into one section titled “Univariate Community Variables,” which could be numbered 3.2.

-. Lines 320-323: This sentence should be moved to the Materials and Methods section.

-. Richness, H, Abundance, and J: The use of the epsilon statistic when Kruskal-Wallis tests are non-significant seems unnecessary. For abundance, which appears significant (p = 0.05), this is not acknowledged in the text. In this case, pairwise post-hoc tests, such as Dunn's test, are required. A univariate PERMANOVA based on euclidean distance with pairwise tests is also recommended if possible.

-. Figure 2.6: The figure caption should provide more information, such as specifying that it shows box plots and describing the symbols (e.g., the thick horizontal line represents the median or mean?). Sites should also be ordered geographically.

-. 3.2. Holdfast Features vs. Ecological Indices: This section should be renumbered as 3.3. There are too many figures with limited value. These could be moved to supplementary materials or replaced by a table summarizing Pearson or Spearman correlations. Additionally, the figure captions for Figures 2.7 to 2.12 should include more details, such as clarifying that these are linear regressions (which were not mentioned in the M&M section).

-. 3.3. Multivariate Patterns: This section should be renumbered as 3.4.

-. Line 411: Add parentheses after "LCBD."

-. Line 416: Cite Figure 2.13. However, this figure seems unnecessary, especially since the results are non-significant, and it should be removed.

-. Dendrogram and PERMANOVA: The statement "The dendrogram unveiled a clear geographic pattern..." is overstated. An MDS (Multidimensional Scaling) plot would be more informative than a dendrogram.

-. Figure 2.14: delete "Dendrogram showing…"

-. PERMANOVA: Table 2.1 should include the number of unique permutations. The table is poorly formatted and difficult to read. The F-statistic should be referred to as pseudo-F.

-. Pairwise Test: The table columns do not align with typical results for pairwise analyses. R² is not commonly reported (at least when using PRIMER). DF can be omitted in pairwise comparisons, as it is always 1. Include only the t-statistic (not F-model), p-value, and the number of permutations.

-. Line 449: "Families" should be replaced with "species."

-. Table 2.2: This table is also poorly formatted. In the caption, specify "contribution to dissimilarity."

-. SIMPER Analysis: More detailed information is required. Currently, only the percentage contribution of each species is presented, which might not provide useful insights. For example, instead of assuming all sites are different, consider grouping sites based on dbRDA or a MDS patterns (e.g., Kob-Inn and Mazeppa; Nqabara, Xhorha, and Dwesa) and use these groups in the SIMPER analysis. Additionally, the ratio value has been omitted. The ratio value is an even more interesting statistic than the percentage of contribution to dissimilarity, as it informs us about the consistency of the results. For example, a species may have a high percentage but only appear at one site, while a sparsely abundant species may better discriminate between groups if it consistently appears within one group. Large tables resulting from this analysis should be moved to supplementary materials.

-. 3.4. Environmental Drivers of the Species Composition Across Sites: This section should be renumbered as 3.5.

-. Lines 462-463: Avoid underlining text.

-. dbRDA Analysis: Important details are missing to validate the significance of the model, such as Monte Carlo permutation tests, variance inflation factors (VIF) for variables, or an ANOVA by terms. Were the environmental variables previously standardized? In the plot, some vectors are very short, suggesting low influence. Consider repeating the analysis with only significant variables (determined via ANOVA by terms). It is recommended to perform this analysis in R.

-. Figure 2.15: Specify the type of analysis in the caption, e.g., "Plot of the dbRDA analysis based on macroinvertebrates associated with Ecklonia radiata holdfast."

Discussion

-. Overall Structure: The Discussion contains inconsistencies, is somewhat disorganized, and does not align precisely with the Results sections. It should begin with a paragraph summarizing the main findings and focus on results relevant to the revised analyses.

-. First Paragraph: This paragraph is confusing, as it references other geographic regions as if they were part of this study. The variable “wave fetch” is highlighted as crucial for species variation, but it was not measured in this study. Incorporating and analyzing this variable (e.g., in dbRDA) would strengthen the study.

-. Section 4.1.3 on Ecological Notes: This section should be rewritten to provide a more integrated discussion. Many statements include information irrelevant to the study’s objectives.

-. General Recommendations: Each subheading in the Discussion should align with the Results sections to improve clarity. Avoid including irrelevant data or unsupported commentary. Ensure all points are tied directly to the study's results and objectives.

 

Author Response

Dear Reviewer 1,

 

I really appreciate the comments, and I believe they will certainly improve my work.

Kindly find my response attached below.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

I liked this MS and think aspects of it have some real scientific value and novelty. Ditch the enviornmental data & analysis and make it a straightfoward faunal ecology of holdfasts/holdfast ecosystem engineers MS. 

 Major comments

I am concerned the output from the dbRDA and think a mistake may have been made somewhere. If I look at the raw salinity for the sites, then there is no difference between sites. However, the dbRBA tells me salinity is a major factor driving differences between two sites to the remaining four. How can this be correct? You sort of acknowledge this with the relate  function. Apart from the impact of longitude (which is sort of obvious) nothing else has an impact.

 

To be brutally honest, I can’t see how any of the environmental data and analysis adds anything to the MS and/or tells us anything about what is driving differences between sites in species composition. I would re-write the MS omitting this and frame it as a study of the fauna of holdfasts between sites for the variables you measured (e.g. depth, sediment wet weight) and as related to the dominant fauns found which mat structure that community. In my opinion, this is the novel and interesting aspect of your work.

I would also remove all references to where you think propagules are coming from and remove the biology/ecology descriptions for the important fauna in the discussion – weave it into the text where it needs to be.

The introduction is well written, you will need to modify the materials & methods and results section in accordance with my comments and the discussion will need a substantial re-write (sorry).

Minor comments

Results section

Values presented, means and standard errors?

Figure 2.2 Y axis should be SST.

Salinity data why is only 2023 data presented (Figure 2.3) for 4 sites? Parts per thousand shouldn’t be used should be PSU (i.e. ppt of what?).

Make sure all if your numbers in a section follow the same numerical format e.g. all to 3 or 4 decimal places.

Lines 415 to 417 – if the result is not significant then you can’t say anything about, i.e. there is no association.

Line 416. Figure ?

Author Response

Dear Reviewer 2,

I really appreciate your comments, and thank you for the positive feedback.

 

Please find the response to your comments attached below.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

After a thorough evaluation, the significance of your study in understanding macroinvertebrate community assemblages and environmental drivers in South African kelp forests is acknowledged. However, I recommend major revisions before considering it for publication. Some parts need clarification or a better explanation. Below are the key areas that require improvement:

1. Methodology and Statistical Interpretation

• It is unclear whether the 2019–2023 environmental data were averaged or treated as a time series. Please clarify.
• The section labeled Multivariate Analysis should include all the multivariate analyses i.e. PCA, PERMANOVA, and dbRDA are also multivariate techniques.
• Definitions and ecological relevance of beta diversity components (e.g., turnover, nestedness) should be more clearly explained for readers unfamiliar with these indices.

2. Data Presentation and Results

• Please indicate which statistical comparisons are significant (e.g., by using asterisks in tables and figures where applicable).
• The table presenting species contributions to community dissimilarity (SIMPER analysis) is difficult to follow and needs restructuring.
• The relationship between depth and site characteristics (e.g., why offshore sites like Kob-Inn and Mazeppa appear very shallow) should be clarified.
• Some figures are partially obscured, making interpretation difficult. The scale bar is missing from the study site map, and locations like Xhorha need better visibility.
• The figures presenting environmental data (temperature, salinity, and primary productivity) lack clarity and should include data sources (e.g., Copernicus).
• I recommend overlaying in-situ measurements with satellite-derived environmental data to allow for a more comprehensive comparison.
• Figures such as the dendrogram should be of higher resolution for clarity.
• A brief presentation of the most abundant species found during the study would add value to the manuscript.

3. Discussion

• The manuscript should use present tense when discussing general scientific knowledge to improve clarity and avoid confusion.
• The Discussion should be re-written, please refer to specific comments in the attached file

I believe these revisions will significantly strengthen your manuscript and improve its clarity, methodological rigor, and scientific impact. 

Comments for author File: Comments.pdf

Comments on the Quality of English Language

The English could be improved.

Author Response

Dear Reviewer 3,

Thank you so much for your comments and positive feedback. Your insights in this study will certainly improve its scientific rigor. I have attempted to fix the issues you pointed out. Kindly find my response, attached, below.

Thank you

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I think that authors have improved the manuscript according to the reviewers' comments, although some inconsistencies remain and should be addressed before possible publication. Most of these are formatting issues rather than content-related concerns. Below, I provide specific comments:

-. Line 204: Delete the information on the dendrogram, as it has been removed from the study. Replace it with a statement about MDS.

-. In the main text, the correct notation for degrees of freedom is df, not DF. However, I believe it is unnecessary to include this information in the text since it is already presented in the tables.

-. Table 3.1: Adjust the data columns to align with their headers. Remove the "number of permutations" column, as only “unique permutations” are relevant. Additionally, the number of unique permutations for species richness is too low and does not match the values reported for abundance or H'. These results should be reviewed. If the total number of possible permutations to obtain p-values is low (<100), Monte Carlo sampling estimates should be used (see Anderson & Robinson, 2003). Furthermore, if H' was not significant, a pairwise test should not have been conducted for this parameter.

-. Line 268: Delete “Table…”.

-. Figures 3.3 to 3.8: These should be numbered as supplementary material, not as part of the main text. They could be labelled as Figures S1 to S6.

-. Table 3.2: Again, adjust the data columns to align with their headers. In the PERMDISP section, there is an extra column that does not indicate what data it corresponds to. Additionally, the table caption should be similar to that of Table 3.1, explicitly mentioning that these are PERMANOVA results.

-. Figure 3.9 should be renumbered as Figure 3.3. The figure caption should state that the site names in the plot correspond to centroids. The stress value should also be included.

-. Line 384: The sentence “The dendrogram… clusters of organisms” should be deleted, as this analysis has been removed. The statement should be revised to reflect the MDS results.

-. Lines 388 and 393: Table 2.2 should be referred to as Table 3.2.

-. Table 3.3 (SIMPER analysis): The results are not correctly presented. The average abundance for both groups should be provided, and the cumulative contribution to dissimilarity should be expressed as a percentage.

-. Line 487: Indicate that Plate 1 belongs to the supplementary material and renumber it according to the journal's formatting guidelines.

-. Lines 491-494: Be cautious with this statement, as non-significant differences were found for H'.

-. Line 576: The purpose of Figure 3.10 is unclear. If retained, it should be introduced earlier in the Results section and designated as supplementary material.

-. Lines 612-613: Salinity was removed from the study and should not be mentioned here.

-. Why is there a second reference list at the end of the document?

Author Response

Dear Reviewer 1,

 

Thanks once again for your insights into our work. Your comments are welcomed, and I believ they have added a significant value to the paper. The response to your comments is attached below. 

 

Thank you

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

The MS is much better, however, you have still left material in the MS referring to the environmental variables and items you said you have corrected. Please work carefully through the MS and correct it.

1. i) lines 24 to 29 – delete: “Longitude and salinity accounted for approximately 22% of the ma- 24 croinvertebrate assemblage patterns. Our results indicates that westernmost sites may 25 serve as possible sources of propagules for eastern sites, hence we advocate for their in- 26 clusion in the next phase of the marine protected area (MPA) network expansion. Addi- 27 tional investigation into genetic connectivity among sites and ecological interactions 28 within holdfasts is advised to support these findings.”;
2. ii) in the Introduction – condense the SST paragraphs – this isn’t measured/used anywhere else;

iii) line 268 – Table?

Author Response

Dear Reviewer 2,

 

I really appreciate the valuable input you have made in this ms. It keeps improving with every comment you make.

The response to your comments is found below.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript has been improved after the last revision, but some important issues still need attention. The statistical analysis of the diversity variables is inappropriate and needs to be changed from PERMANOVA to ANOVA. A species list is also required since it will give additional value to the study as a baseline for future projects. Specific comments can be found in the attached document.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Some corrections are still needed.

Author Response

Dear Reviewer 3,

Thank you so much for the valuable inputs in the ms. They have really made an impact in this paper.

I really appreciate it.

Kindly find the response below.

 

 

Author Response File: Author Response.docx

Round 3

Reviewer 1 Report

Comments and Suggestions for Authors

I think that authors have improved the manuscript according to the comments, and I can recommend its publication. However, some errors remain which can be corrected in the printing proofs:

-. Figure 3.9 should be renamed as Figure 3.3 (both in the text and caption)

-. All supplementary material must be cited in the main text. For example, Tables S1 and S2 are missing in the main text. Additionally, Figures S1 to S7 still appear in the main text as Figures 3.3 to 3.8 and 3.10.

Author Response

Dear Reviewer 1

 

Thank you so much for your insights. Your inputs have significantly improved our work, and I appreciate that you have recommended our ms for publication.

Below, I have made the adjustments you have suggested.

Comment 1: Figure 3.9 should be renamed as Figure 3.3 (both in the text and caption)

Response 1: Thank you for picking that up. I have revised the numbering as suggested.

Comment 2: All supplementary material must be cited in the main text. For example, Tables S1 and S2 are missing in the main text. Additionally, Figures S1 to S7 still appear in the main text as Figures 3.3 to 3.8 and 3.10.

Response 2: Agreed. The numbering has been revised and changed according to your instruction.

 

Thank you

Reviewer 2 Report

Comments and Suggestions for Authors

I believe the MS is much improved and will recomend it for publication.

Author Response

Dear Reviewer 2,

 

Thank you so much for recommending our work for publication. None of it would have been possible without your valuable imputs.

 

Thank you so much

Reviewer 3 Report

Comments and Suggestions for Authors

I appreciate your willingness to address all my comments; however, I insist that PERMANOVA is not the appropriate analysis for testing differences between sites for a single variable. PERMANOVA is designed for multivariate data, and I am unsure how it was applied in your study. Given this concern, I am unable to recommend your paper for publication.

Author Response

Dear Reviewer 3,

 

Comment 1: I appreciate your willingness to address all my comments; however, I insist that PERMANOVA is not the appropriate analysis for testing differences between sites for a single variable. PERMANOVA is designed for multivariate data, and I am unsure how it was applied in your study. Given this concern, I am unable to recommend your paper for publication.

 

Response 1: Thank you so much for your inputs and your eagerness to improve our work by ensuring that we do not compromise the quality along the way. Although, univariate PERMANOVA was suggested by reviewer 1, I decided to take up your comments and rerun the analysis using the generalized linear mixed model since I could not use ANOVA since our data violated the assumptions of the test. 

 

I hope that GLMMs are as acceptable as an ANOVA that you suggested. I have noticed that the patterns we got from the GLMM output were comparable to the ones generated from univariate PERMANOVA, therefore, the writing and the conclusions remain the same. I appreciate your firm grip on the usage of an alternative test. 

 

I have uploaded the ms with new analysis and have removed univariate analysis on ecological indices. Please kindly reconsider your decision based on the improved version.

Thank you