In the manuscript “Revisiting the phylogeny of Wolbachia in Collembola”, Ma and coworkers sample Wolbachia strains from various springtail host species and investigate their phylogeny. Wolbachia are the most common invertebrate symbionts and are found in most hexapod orders. However, the Wolbachia found in Collembola are genetically very distinct from all other Wolbachia strains, and are thought to be involved in creating parthenogenesis in many collembolan taxa. Therefore, this study is relevant to the field of endosymbiont research and entomologists alike. The main findings are 1) All Wolbachia from springtails form a monophyletic group, hereby correcting a previous study; and 2) Wolbachia has multiple origins in springtails. The manuscript is well written, the figures are clear and visually appealing, and the phylogenetic analyses are detailed and thorough. I believe the paper will be a valuable contribution to the Wolbachia literature. One critical point would be how the authors show that Wolbachia from Nelipleona are supergroup E strains and not a distinct strain. In the 16S+ftsz analyses, 3 species of Nelipleona are included, but in the MLST analysis, only a single species is analysed. To exclude this is simply a sampling artifact, the authors should perform an additional 16S+ftsz analysis using only the taxa used for the MLST analysis. If the single Nelipleona Wolbachia is distinct from supergroup E in there as well, this would support the author's argument even further. I also see a few problems and errors with how phylogenetic terms are employed and think that the paper could be improved by discussing not only Wolbachia phylogeny, but also its implications for Wolbachia evolution in springtails. Please find more details on these and other points below, hopefully constructive.
Line 54–55 There is no direct evidence for Wolbachia– mediated speciation. This makes it sound as if there is.
Line 90–91 This sentence, and other similar ones in the manuscript, is very problematic. Collembola are not an “ancient group of arthropods”. Arguably, all recent taxa have evolved for the same amount of evolutionary time. Also, collembolans did not “branch off” anything. What you may want to say is that the last common ancestor of Collembola and Insecta is very old.
Line 116–117 I would advise against using terms such as “intermediate”, “basal”, “in between”, etc when describing phylogenies, this is simply wrong. Trees show sistergroup relationships and should also be described as such. For example, instead of saying “taxon x” is the “most basal” taxon in this tree, one should use “taxon x is the sistergroup to all other taxa in this phylogeny”. Please see Krell & Cranston (2004: Which side of the tree is more basal? Systematic Entomology 29, 279-281) for more details.
Line 303 I don't see the point of “data available on request” statements. Including the results of recombination analyses in the supplementary files does not require a lot of work.
Line 311 Please rephrase problematic formulation: “were separated by wThou“
Line 313 Please rephrase “outermost taxon in clade E”
Line 337–348 I don't think it is necessary to mention this in the manuscript. Maybe just drop an email to the database curators?
Line 368–370 How can non-matching phylogenies be “intensified” by long branches? To match phylogenies, one only looks at branching patterns and not at branch lengths.
Line 371–375 Here you use the 16S+ftsz phylogeny to argue that geography has played a role in diversifying Wolbachia from springtails. However, just in the paragraph before you make the point that this phylogeny is “insufficient” for classification of Wolbachia strains.
Line 377–379 It would be interesting to read a little more on how you would interpret the non-matching phylogenies of Wolbachia and Collembolans.
Line 393–394 In your Figure 1, Supergroup E is sister to Supergroups A+H; Supergroup B is the sistergroup to the former 3! You should also mention that this phylogeny is not very well supported, except for the monophyly of most supergroups.
Line 399–404 Gerth et al did not claim E is “basal” to all other supergroups, but rather that it is the sister to all other >investigated< supergroups. Also, the LBA problem has been thoroughly addressed in the cited study. I am not saying that there may not have been a problem with LBA in this study, but if you make this argument here, it should be explained better.
Figure 1 shows an unrooted tree, which I think is fine. However, I wonder why you displayed in this way, and not as a truly unrooted phylogeny. The way the reader perceives this tree is that it is rooted with supergroup C. I think it is always best to show unrooted phylogenies in an unrooted manner.
Table A3 is huge and only repeats information from PubMLST database. Maybe think about omitting it.