Molecular Phylogeny of Holarctic Aeshnidae with a Focus on the West Palaearctic and Some Remarks on Its Genera Worldwide (Aeshnidae, Odonata)^†

Round 1

Reviewer 1 Report

The manuscript by Schneider et al. focuses on a phylogenetic and taxonomic revision of the Aeshnidae, an important dragonfly family, with a focus on the Holarctic fauna. According to the authors, this is the largest dataset of DNA data on this family to date, which I accept, but the manuscript has many weaknesses in my opinion. I list several comments in the attached PDF file, and summarize the most significant ones here.

1. DNA data analysis and tree construction using various techniques is used to arrive at some consensus result that supports/refutes some claim or hypothesis of ours. Describing (at great length) partial results that may contradict each other or have no substantial support and thus no predictive value is therefore completely unnecessary and makes it seem more like the authors have nothing to fill the manuscript with.

2. I don't find it useful to use haplotype networks for multiple species, because this information is provided by the trees that have already been constructed. I could understand it if HT networks dealt with e.g. relationships of samples of one species from different geographic areas, or if the authors wanted to show that multiple species are in fact one species. However, this is not the case in this MS.

3. The authors suggest that COI is more variable than ITS, referring to the work of Hasegawa,Kasuya 2006. However, this work analyzes 16S and 28S rDNA, and most other works suggest rather that ITS is a highly variable marker. I may be wrong, but to my mind the authors have not presented any relevant evidence that COI is indeed a highly variable and subsequently saturated marker, and rather it is the case that some species of Aeshnidae can only be distinguished by more variable markers than COI, in this case ITS. COI is the standard marker used in insects for species determination. I'm not saying it's perfect, but it has been shown many times that it doesn't work as well in Odonata as Odonata-experts would like to imagine. However, this cannot be resolved by the unsubstantiated claim that COI is a highly variable and saturated marker.

4. I am not an expert on Odonata, but if someone were to propose taxonomic changes to a group I am involved with based on such poorly supported and often conflicting DNA data, and furthermore support those changes with a large number of shared characters, I would suggest not publishing those changes until they are unequivocally supported. However, insect taxonomy does have some common rules in my view, and I would recommend the same in this case. After all, it is not impossible that the proposed changes (weakly supported) will cause more chaos than improvements. Obviously Odonata need taxonomic studies and it is very appropriate to apply DNA data as well. Datasets and analyses should be considerably more robust than in this case. Today, it is no longer a problem to obtain information on the entire mitogenome and substantial portions of nuclear DNA, and it seems that without such data it is virtually impossible to reliably reconstruct the taxonomy and phylogeny of Odonata.

Based on the mentioned reservations, I believe that the manuscript contains many weak points (e.g. unnecessarily long description of partial results, many weakly supported claims, insufficiently supported taxonomic actions). Therefore, I do not recommend accepting it for publication.

Comments for author File: Comments.pdf

Author Response

Please see attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

The analytical methods applied look sound, and it appears fair to use the presented dataset to provide confirmatory evidence to support conclusions that were reached using morphological characters.   Not being familiar enough with methods and techniques of molecular systematics I cannot comment on the strength of the data as base for addressing deeper level relationships.

Suggestions to improve the text are given as comments plus hihlighting on the manuscript. They are only suggestions.

Just some thoughts:

An obituary for Prof. Dr.  G. Peters recently published in Odonatologica ended with:

Unfortunately, Günther Peters’ hope to find an appropriate placement genus for the European Aeshna isoceles (O.F. Müller, 1764) died with him. In his opinion, this species is neither an Aeshna nor Anaciaeschna, and he said that it might end up as a new genus. He may have already started a file under the name of Anisoaeschna. This will, however, not be the only problem in aeshnid systematics in which Günther Peters’ experience will be badly missed.

In my (GT) opinion Anisoaeschna, even Anisoaeshna ( άνισος =Greek for unequal, not the same as), may be a more appropriate name for the new genus than Isoaeschna as it is found in this study definitely not to be the same as Aeshna.

Anyway, it would be nice - if possible - to dedicate the paper to Günther Peters who, after intense engagement with the same systematic problems already 50 years ago, passed away on 15th January 2023 in Panketal near Berlin at the age of 90.

Comments for author File: Comments.pdf

Author Response

Please see attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

In this manuscript, the authors try to disentangle the phylogenetic relationships of taxa within the family Aeshnidae across the entire Holarctic with a special focus on Western Palearctic species by using previously published sequences and providing some new sequences of their own.

General comments: The general structure of the manuscript is fine, however, I do have several issues with this manuscript. While the language in some parts of the manuscript is ok, other parts are either grammatically incorrect or flawed (I’ve indicated several sections in the specific comments), lacking verbs or other words or deviate from the generally acknowledged terminology of this particular field. These linguistic flaws also sometimes make it hard to comprehend what the authors actually want to express. The entire manuscript needs a thorough language check and rephrasing/rewording of some sections.

The lack of line numbers makes the manuscript harder to review, this should be changed in future versions.

The appearance of the trees should be improved. At the moment, some branches appear thicker than others sometimes making it impossible to deduce branching patterns or read posterior probability values. The trees should also rather be shown as phylograms and not dendrograms and scale bars should be included. The authors may also want to consider changing the appearance of their trees (collapsing to haplotypes, changing support values to symbols,…) in order to increase readability.

The authors sometimes use ‘West’ and sometimes ‘Western’ Palearctic. This should be unitized throughout the manuscript. I assume, the authors used MrBayes for Bayesian Inference and not -as the stated throughout their manuscript- Maximum Likelihood analysis. If so, this should also be changed throughout the manuscript.

Specific comments

Line 25: Please state the exact number rather than a vague approximation.

Line 36: . is missing after analysis

Line 42: we did not subsume…

Figure 1: I don’t really see the point in showing three pages full of pictures. Yes, Aeshnidae have diverse morphologies and colorations (although the general body shape and size are conserved to a certain extent) but in my opinion it would be enough to only show the most divergent morphologies...Also, this way the figure itself and the figure legend don’t fit on one page.

Line 141: Here (as well as in table 1), I would strongly suggest to cite the respective studies the sequences were originally published in (Galimberti et al. 2021, Geiger et al. 2021, Haring et al. 2020, etc…). So far, only Galimberti et al. (2021) are cited but not in this particular context…

Table 1: Include references to original studies where additional sequences were published in. Remove residual vertical bars from the table.

Figure 2: Is sample 17047 really ‘Wast’ Caucasus? Why are some branches appearing thicker than others?

Figure 4: Thick branches appear unreadable. Related- or distinctness of taxa cannot be deduced from the figure.

Page 6 of M&M, last paragraph: grammatically, this sentence appears wrong to me… ‘…were analyzed…using…was applied for both genes.’ I think it’s the last part…

Same paragraph: with M=-5 and S=1.2

Same paragraph: I would suggest: ‘…are always present in some species…’

Figure 5: In my opinion, the authors should only show support values higher than 75 or 80 and don’t really need to show the color-coded branches.

2.5. haplotype network analysis: inference method…intermediate molecules.

3.1. ‘Most of the specimens’ should be changed to the exact numbers. …some derived from the same group and published previously should be rephrased, in my opinion.

It corresponds well…

Clade 1: I do not fully agree with the authors here. A. juncea shows a lot of structure, although this structure is not directly linked to geography as Russian samples turn up in all clades/branches of the tree. It is however, monophyletic with respect to A. subarctica.

Clade 3: cyanea-clade includes

Clade 4: identical with A. mixta

3.2: …so that we were able to include

…specimens of A. cyanea from North Africa and Europe

Clade 7’: last sentence of the paragraph is missing a verb

Page 33: What do the authors mean by ‘(even one uniting two species of the same species)’?

…and in fact…

…the haplotype network indicates… and here, I would disagree with the authors as the high number of mutations rather shows the opposite of a relatively close relationship. Even the different species are fairly far apart from each other in terms of mutations…

Page 34: …The second consists of Anax….

…same relationships between Anax species is seen in the..

Page 35: The second paragraph is very difficult to read for me. First, because the language sometimes seems grammatically incorrect or does not resemble the terminology of the field. This section as well as others in the manuscript should be checked by a native speaker. Secondly, I don’t fully agree with the content. The authors write about the lack of genetic differentiation between A. imperator and A. parthenope. In figure 3, however, A. imperator from Liberia and Germany turn up in a totally distinct clade from A. parthenope with the only exception of one A. imperator sequence from Germany which turns up in the A. parthenope clade. However, to me, this seems more like a misidentification or sample mix-up rather than a general fact (although I am aware of this topic but also see discussion in Geiger et al. 2021). Furthermore, in figure 4, A. parthenope and A. imperator are even separated from each other, although, this clade is not resolved but resembles a polytomy. I would advise the authors to rephrase this section language- and content-wise. I also don’t understand what the authors want to express with the BLAST statement. Blasting A. immaculifrons sequences will hopefully will reveal Anax spp. as the next closest hits.

Page 36: I don’t understand what the authors want to express with the ‘Some restoration….’ paragraph. Please clarify and rephrase.

Page 37: The first paragraph is very hard to read and understand and the grammar and construction of the sentences seems wrong.

4. Overall discussion: …substitutions are accumulated…

‘This well-known concerns… this sentence is missing a verb.

Page 38: First sentence seems grammatically wrong.

…the name of the genus to which the type species has been attributed for a long time;

References: The entire reference section needs reformatting. Some references are written in bold, some in hyperlink, some only partially in the first or latter style. Sometimes, journal names are abbreviated, most of the times journal names are written out fully. All of these issues need to be changed and unitized according to the journals requirements.

Please refer to my general comments for the quality of the English language. Again, I would recommend having a native speaker going over the manuscript before resubmission.

Author Response

Please see attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

I am sorry, but I do not feel the manuscript was significantly improved, and the authors did not convinced me by their corrections. But I leave the final decision on the Editor. Beside that, there is a lot of smaller mistakes in the corrected MS, like missing spaces, missing numbering of references in the list of references, so the numbers in the text cannot be checked. The authors for example use COI as "mitochondrial cytochrom coxidase subunit I", but in fact it is only a small part of the whole gene, and it is cytochrome, not cytochrom. Also ITS is described only as "internal transcribed ITS", but it is internal transcribed spacer and there are more spacers used in phylogeny or systematics. Some parts of the MS are included twice. I still think it is useless to use haplotype networks for more species together. Some Figure legends are not selfexplanatory, so it is not clear from them what exactly is there (Fig. 2), moreover I do not understand why to include same thing mirrored in onen Figure. I also still do not agree that COI is faster evolving that ITS. The authors did not provide any evidence for it, so it seems to me rather that the results from COI did not suit them very well, so they attributed it to COI.

Overall, in my opinion, the manuscript is not of very high quality, the authors tried to correct it as quickly as possible, and their corrections did not completely dispel my doubts about whether it is appropriate to publish it. It is certainly only my subjective opinion and therefore I leave the final decision to the Editor.

Author Response

please see attached file.

Author Response File: Author Response.docx