All reviews of published articles are made public. This includes manuscript files, peer review comments, author rebuttals and revised materials. Note: This was optional for articles submitted before 13 February 2023.
Peer reviewers are encouraged (but not required) to provide their names to the authors when submitting their peer review. If they agree to provide their name, then their personal profile page will reflect a public acknowledgment that they performed a review (even if the article is rejected). If the article is accepted, then reviewers who provided their name will be associated with the article itself.
Thank you for the final revision. Your manuscript is accepted for publication in PeerJ.
# PeerJ Staff Note - this decision was reviewed and approved by Rob Toonen, a PeerJ Section Editor covering this Section #
Thank you for the revised manuscript and attention to details provided by reviewers. One minor revision has been identified by reviewer #1:
> There is only one issue I have left; the fact only one clutch was used. While a lack of genetic diversity may be great for running statistics care should be taken when drawing conclusions from one genotype as a number of papers have demonstrated family dependent effects of ocean acidification. Seeing as your results are largely in agreement with previous studies that used multiple clutches it doesn't seem to be too big of an issue. However, I strongly feel that this should be mentioned as a caveat in the discussion, just so the reader is aware of this point (as is may get missed in the methods) This is a very well written paper, great work.
Kindly address the suggestion with revision to the tracked change manuscript. We look forward to accepting your manuscript as soon as you can respond to this communication.
Firstly, I would like to apologise for misinterpreting that data table you provided. You have done an excellent job in addressing the concerns that I raised. There is only one issue I have left; the fact only one clutch was used. While a lack of genetic diversity may be great for running statistics care should be taken when drawing conclusions from one genotype as a number of papers have demonstrated family dependent effects of ocean acidification. Seeing as your results are largely in agreement with previous studies that used multiple clutches it doesn't seem to be too big of an issue. However, I strongly feel that this should be mentioned as a caveat in the discussion, just so the reader is aware of this point (as is may get missed in the methods) This is a very well written paper, great work.
All comments and request of clarifications that I have provided on the first version of the manuscript were substantially fulfilled. I do not have any further comments on the second version of the manuscript.
No comment on the experimental design as in the first version.
All the request were fulfilled.
Authors have modified their manuscript taking into account all the comments I made during my first review. Their manuscript is still well written, the research question is well defined and their methods and investigation are rigorous, well described and relevant. Data is robust, statistically sound and controlled.
Improvements made to the manuscript are significant. In particular, authors reduced the length of their manuscript, re-worded properly some ambiguous or erroneous sentences or terms, without reducing the clarity and meaning of their manuscript.
I recommend the article to be accepted as it is.
The reviewers remarked on the well written and crafted manuscript, however, they provided a number of points to consider in your revised manuscript. Concerns were expressed by all three reviewers regarding experimental design involving issues of replication, control conditions and related details. For example, how many times was the experiment repeated and is a regression with n = 3 sufficient (Figure 1 A-B) to support your conclusions?
Your careful assessment of comments regarding the statistical analyses and their significance will be important in your revisions and next round of reviews.
The use of the term “settlement” in the context of the stage of development of the larval clownfish was also of concern as was the history and experimental details of the brood-stock pairs.
Reviewers suggested re-organizing and re-wording the results and discussion sections as well as decreasing the length of the manuscript. A numbered list or chart of the diverse metrics used was suggested to assist the reader in understanding the methods and analyses completed.
The reviewers offer detailed suggestions in each section that will improve the overall manuscript. We look forward to receiving your revised manuscript for further review.
This paper reports the effect of ocean acidification on otolith development in larval clownfish. This study includes SEM analysis in addition to otolith parameters usual measured which makes it one on the most in depth studies looking at ocean acidification effects on otolith development.
The MS is very well written and crafted. Its contains all the appropriate references. The raw data has been shared but in the form of mean tank values. Raw data should be provided for each individual fish. Figures and table are clear.
I don't understand why you present the survival, settlement and growth data first in the results and discussion. The focus of this paper is otoliths so lead with this.
The experimental design and methods appear good, however, there is some vital information missing with regards to the brood-stock which make it difficult to fully assess the strength of the experimental design.
1. You need to describe the history of the brood-stock pairs, were they wild caught, or lab bred (if lab bred what generation)?
2. Need information on how many pairs and clutches were used in the experiment and how the clutches were split between tanks, so the reader has an idea of how much genetic diversity was in the experiment. Its also difficult to assess the strength of the experimental design without this info.
3. When reading the experimental design section it sounds like you had 12 tanks and ran the experiment only once but looking at your supplementary data file I can see you ran it 7 times? This needs to be made clear. Where the same pairs used each time? Addressing point 2 should clear this up.
I have an issue with the way the data has been analysed. I can understand taking the mean of tank as the individual fish aren’t independent replicates. However, to take the mean of a tank across the 7 times (taking the mean of a mean) you ran the experiment is inappropriate as you are masking a lot of natural variation and thus forcing an R2 that is unlikely to be representative of what is actually going on. The mean for each tank each time the experiment was performed should be included in the analysis, so you should have 21 points per treatment not 3. Then tank should be included in the model as a random factor. Parental pair or clutch (these should be uniquely labelled) should also really be included as random factors in the model nested within tank. It would also be interesting to see how the relationship looks if you plot the data for each individual fish, this could go into supplementary materials. I would also be interested to see how the relationships looked if the 3000uatm treatment was excluded as it appears some of the significant trends are being driven by this treatment which is unrealistic for coral reef fish. Just out of curiosity did you try analyzing the data looking at comparisons between treatments and were they any significant differences, especially between the 390-1600 treatments which are more relevant levels for coral reef fish in the future.
Additionally, you don’t specify/it isn’t clear why you choose to do a regression analysis over a standard treatment comparison approach. This relates to the question you are asking. Perhaps re-defining/refining your aims could clear this up.
The discussion in well written. However, care needs to be taken when comparing your results to some past studies as the data was analysed differently and it does seem that some of your significant trends are being driven by the non-ecologically relevant 3000uatm treatment.
There is an extensive use of tables while it would be better to use charts as they are easier to be visually understood. There is a specific comment in my revision.
I found too many speculations in the discussion part, that is unnecessarily long. I think that authors should keep speculations at minium and always keeping in mind the central topic of the work (pretty much what is stated in the title).
In their paper, Holmberg and colleagues investigate the potential effects of future sea acidification on the morphology of otoliths and other biological characters in the Clark’s anemonefish. They used a lab experiment testing the impacts of different water pH, eventually similar to those foreseen for the future. Most of their results are in line with the ones from other studies on the same topic, even though in their experiment they considered a set of non-ordinary morphometric characters. Overall, I found the experiment designed and conducted in a solid way, even though, considering the rationale and the final objective of the study, it would have been better to increase the number of aquaria for each level of treatment. The micro-structure analysis of otoliths would have also helped to shed some light on different issues on which authors are not particularly sure (e.g. the assessment of settlement). I do not have any major concerns about otolith analyses while I found some problems concerning the estimation of mortality, settlement timing and somatic growth that I explain below.
As pointed out by authors, it is not surprising that the extreme value of pH investigated did not result in a higher mortality rate in larvae. The high natural mortality that occurs during settlement has probably blurred the effect of acidification on this variable. As in many others studies, in my opinion assessing and comparing mortality in fish larvae is a not easy task and the outcomes should be discussed with caution, also considering that it is an experiment conducted in controlled conditions.
One concern regards the assessment of acidification effects on the number of settled individuals. Authors found that SL was reduced by increased pH. This could be an indication that settlement was going to be delayed by the experimental conditions. But in order to claim that there was a delay you have to know which is the average age at settlement in control conditions, that authors do not have as they stop the experiment even before the end of the larval phase in controlled conditions. From this perspective the outcome and the relative conclusions made by the authors are very doubtful in my opinion. In this sense, which was the rationale for choosing just 10 days for rearing larvae and not waiting, at least, till the end of the potential larval phase (that in the species should be around 15 days)? I also wonder if eggs hatched at the same time. If not, this would add further uncertainty to the assessment of settlement delay. I found the conclusions and the hypothesis about this part to be improper and over-speculative. In my opinion this part of the study should be removed.
It is interesting that SL decreases with increased acidification, something on which there are controversial results coming from the few studies carried out till now. Concerning this result I have a concern about the potential influence of size at hatching on the estimation of the final standard length. Size range at hatch for the species considered can be 0.3 mm, so it is not that different from the range found by authors on the final standard length between aquaria averages (that seems to be <0.5 mm from the regression). Considering these comparable ranges, how can authors be sure that the final difference in terms of standard length was not due to an initial difference of size in recently hatched larvae (even considering a random distribution of larvae in the tanks)?
Finally, even though I found the discussion interesting and well written, in my opinion it is too long and not very focused on the results of the study. Even though the journal accepts certain speculations I think that they should be kept at minimum and avoiding to go out of context. In this sense, in my opinion it is not proper to speculate about potential impacts of ocean acidification on behavior or sensory reception as authors did not have any data on these topics at all. For this reason I suggest to shorten the discussion.
line 27: no need to specify what ocean acidification is in the abstract, in my opinion. The concept is self-explanatory.
Table 1. I would directly specify in the table the effect like ‘increased S area’ or ‘decreased S roundness’. It would be clearer. Also because for some metrics like ‘S shape’ an increase is difficult to understand.
Line 134: it is not very clear what ‘lateral face’ is, even looking at the supplementary material. Is it the dorsal part of the otolith?
Material and Methods
line 233-236: why are these variables visually scored by operators? Especially for lateral development that is a measure of convexity, was there no way to ‘measure’ it and then assign it to a score?
line 243: how many otoliths (and % on the total) were discarded? Also considering the unsettled individuals.
There is an extensive use of tables. In many cases I would use graphical representations like bar-charts to show results. Tables, especially when they are large, are very difficult to read.
Results are very wordy and probably too long. I would reduce them a bit keeping the main output of the analyses. I do not find useful to report all the detail for all the otoliths analyzed.
You should explain what RC1 and RC2 are in the text, for a non expert reader.
Why did you show a PCA in a table?
Figure 1. Please, do not extrapolate regression lines and confidence intervals.
line 402-403: how is this possible?
The manuscript is clear and unambiguous, with professional English used throughout
Literature references are "over-sufficient", the authors cannot be blame for that and I guess the peerJ format allows that, but it is true that the long lists of references (3-4 lines of only references sometimes) is sometimes disturbing. But this shows the good knowledge of the authors and this provide a powerful contextual frame.
Article structure is correct, but authors should consider to reduce its length significantly.
Figures are of good quality, while being not very "graphically" attractive. This is purely esthetism, but I cannot say that more attractive figures would be a great value here.
Tables are fine, so are the raw data shared.
This is an interesting, substantial, and coherent body of work, but as written in my detailed review, the authors should consider re-organizing or re-wording some of their paragraphs and ideas before this manuscript gets published.
This is original primary research within aims and scope of the journal
The research question is well defined, relevant and meaningful
The investigation is super rigourous, well backgrounded, and performed to high technical & ethical standard
Some methods may be detailed more (despite my general comment to, most of the time, there are too many details and the overall manuscript is too long and has to be shortened), see my list of comments for that.
Rational & benefif to literature is clearly stated, even if this could be improved in the discussion as detailed in my list of comments.
Data is robyus, statistically sound & controlled
Conclusion are well stated but can still be improved, especially after considering the real meaning of the "settlement" term used by the authors. I recommend the authors to change this wording.
The authors should consider re-writting part of their Results and Discussion to meet with what is expected from such sections. Results section should presents... results only, and Discussion section is not a review section. Overall, the authors have all the sufficient data and knowledge to perform these improvements, so there should be no problem with that. I highly recommend the authors to read my details comments in regard to that.
Speculation is appropriate.
This is a great manuscript relating a substantial amount of work on a clownfish otolith metrics in the context of acidification pressure. This manuscript is very professional and globally well written. The changes that have to be performed are not related to the quality of the writing, but to the manuscript organization between Introduction, Mat&Met, results and discussion. Some wordings are inappropriate and require modifications. I highly encourage the authors and read carefully all the detailed comments provided below.
I believe that this study can be published if these comments are taken into account.
I’d rather summarize here the main idea and hypothesis of your discussion on your results, ocean acidification and otolith’s functions, instead of saying that “you reviewed the literature”, as this does not provide any important information. Be more specific perhaps.
missing a word I guess “these impacts may have consequences on teleost fitness in the near-future ocean”?
“serving to abate”: awkward wording, please change to something that does not imply that the absorption of CO2 by the ocean intentionally “serves” to decrease the CO2 greenhouse effect. This is an inappropriate “personification”.
You should simplify this sentence “impacts seawater chemistry by decreasing pH and carbonate (CO32-) concentration (Doney et al., 2009).
I would be more direct: “both can impact” instead of “both are expected to impact”. Please also refer to more, including more recent, literature to justify this point.
I would delete the first sentence, as it introduces well you research interests, but does not serve this introduction. Then I would organize this paragraph a bit differently, by introducing that Teleostii dominate most aquatic habitats and are internal calcifiers (which would make the connection with your previous paragraph). And then I would write that in addition to the already known impacts of acidification on other of these fishes’ biological functions, their calcified structures might also be vulnerable in acidified contexts.
This a very well written paragraph, but as the entire manuscript actually, this is too long. You could summarize that very briefly by saying that otoliths are calcified structures of the inner ear that function as sensory organs. The pH gradient in the endolymph is also a good point in the context of your study, but apart from that I advise the authors to be more concise in their writing.
I would add “therefore” or a synonym in “… elevated seawater pCO2 therefore stimulates…” to precise that you refer here to the demonstration you made with Table 1.
I don’t understand what the authors actually mean here.
I suggest the authors to use (i), (ii), … etc. to list the different metrics that they investigated (as this is again very long).
While references are over-abundant in the previous parts of the introduction, I would like the authors to here justify their arguments with more literature.
Authors should precise here that the otolith morphology they will look at corresponds to the metrics they described in the previous paragraph. As previously, the authors should consider to make the transition clearer between their paragraphs, and this could be done most of the time by a simple word (e.g. here: “we investigate ocean acidification impacts on these otolith metrics in A. clarkii” or something similar).
Authors should consider to simplify their point here and move the most part of these information in the Mat&Met section. I advise the authors to be more direct, and so in the whole manuscript, to shorter it significantly.
Authors should check if the volume metric required by peerJ guidelines is liter of if gal is also accepted.
If possible, I recommend the authors to use (i) (ii), etc. instead of “1.”, “2.” etc… as this will be clearer given that lists start with numbers.
As authors use liter here for volume metric, they should consider homogenizing their manuscript by using always the same metric type all along the manuscript (in reference to my previous comment on gal).
I do not agree with the terms “settlement” and “settle” used by the authors. I recommend the authors to use “settlement-stage” instead, as ‘settlement’ refers to an ecological transition that does not occur here in aquaria, especially as you euthanized individuals at that time. For instance: “Due to the natural variability in developmental time required for individuals to reach their settlement-stage” would be more appropriate wording than what the authors have used. I recommend the authors to consider re-wording all the sentences where they clumsily refer to “settlement” and “settle” instead of “settlement-stage”.
Unnecessary sentence as it does not provide any clear information.
Mat&Met – general comment
The use of “we” at the beginning of every sentence is probably one of the reason why this section is, similarly to the introduction, too long. I recommend the authors to be more synthetic, and to not hesitate to use the passive form instead of “We + verb”.
I would add this paragraph in Mat&Met to confirm the proper functioning of the pH treatments and conditions, but not here in the results section. Or, in the case the authors comment these results, then OK for the results section, but here the authors only cite the Table 2, without any particular interest of that being in the results section.
lines 288-289, 305/308, 322-324
I recommend the authors to switch directly to the results. There is no point of presenting what they will then present. Authors should just present directly their results. This sentence is typical of a figure or table caption, but should not stand in a results section. The manuscript is well written, but really not synthetic, and these particular cases are symptomatic of what should be improved by the authors.
I do not feel convinced by these regressions with n=3. In addition the variability is important, and the differences are also not convincing, as evidenced in Table 3 with the means and standard deviations. I would therefore rather limit these results to comparisons between the pH conditions, not taking risks with the estimations made by the authors with the “each 100 µatm increase in pCO2”.
Again, I do not think that authors measured settlement by putting fish in a habitat context and observing the number of fish settling in an anemone and the number of fish staying in the water column. What their measure is the acquisition of certain developmental features, related to the settlement stage in situ. I recommend the authors again to change their wording appropriately, as this result actually represents the proportion of fish reaching certain anatomical and developmental criteria, but not an ecological step.
I do not understand the point of this text. Perhaps in the Mat&Met, but still, this does not provide any result here.
Authors should cite their Table 4 here.
This is the interesting part. The previous sentences are unreadable or at least too long. In this result section, the authors should consider justifying their “biological” results with statistics. If they want to detail more the statistical significance of all that, authors should consider to write extra files or extra Mat&Met for the reader to have access to such details, but here, in the result section, the messages have to remain clear.
lines 334-340, 347-359, 367-382
Same comment as above
lines 341-346, 360-366
Very interesting! But maybe more a discussion than a result? Anyway, this is great and way more interesting than the long list of statistical details… The authors should consider rewriting their results section taking this comment into account.
Again, the authors should avoid to use the term “settlement” in this context, and talk more about “settlement-stage” pigmentation or morphology.
Also, the author should consider to add some literature references here, the work by Victor 1986 is great, but there are several other publications relating that: authors should have a look in particular at McCormick 1999 (delayed metamorphosis in A. triostegus) and Holzer et al., 2017 (A triostegus metamorphosis).
I would be more careful with the conclusions given here (but maybe this is because of the inappropriate “settlement” wording): mortality and somatic growth are not impacted, but competency (developmental ability to settle) is impacted. This is very interesting and the authors should give more emphasis on that in my opinion.
What do the authors conclude from these 24 studies reviewed? It would be a great value to put in parallel the great point made by the authors (importance of investigating each otolith independently) with their review (as done in this article), but also to dress some sort of a conclusion or perspectives taking into account both these things.
This is great and interesting, however I wonder if it is essential to mention the “number” of paper consistent with the authors’ findings, or in the case it would also be appropriate to mention the number of papers inconsistent with these findings? I would remain simple here and simply mention that this is consistent, with some specificity that the authors mention very well, with most studies, and I would also provide small insights on findings that are new and/or inconsistent.
This is too long and too much a bibliographical work rather than a discussion of the results.
I cannot agree more. This is all very interesting and warrants review, but this is not the topic of this manuscript, as I previously evoked many times.
This is a great paragraph, but this has, in my opinion, nothing to do in this particular research article. Authors should consider writing a separate review, as they barely refer to the findings of their study here.
This is finally some real discussion challenging the authors’ findings with existing literature and hypotheses. I recommend the authors to organize their entire discussion like this paragraph if they review their manuscript.
Could the author discuss more their results on asymmetrical responses in the context of lateralized processes in coral-reef fishes? I recommend the authors to have a look at studies relating lateralized behavioral, anatomical and developmental processes in coral reef fishes at the settlement stage, in particular when facing external perturbations such as acidification (see Ferrari and Besson papers). Putting the results of this study into such a context would be a great value for this manuscript.
I would like the authors to justify or discuss more the reliability of their curves given the small number of replicates used for these curves (n=3).
All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.