The soft underbelly of a phylogenetic analysis

Here’s another rejection letter
that sees things the way the editors want to see it, not the way things need to be seen. I post these reviews and replies because someday you may want to publish a paper yourself and you need to see what editors are willing to say and do to keep, in this case, the origin of pterosaurs a mystery, and to keep amateurs from embarrassing the academic community by reporting that all they need do is add a few more relevant taxa.

Associate Editor Comments:

The intro: “The manuscript argues for a hypothesis that places Longisquama, Sharovipteryx, Cosesaurus, and a small number of other taxa as being especially closely related to pterosaurs, with their respective morphologies informing the long sought-after origin of characteristic pterosaur traits. The manuscript quite correctly identifies its preferred hypothesis as a minority viewpoint, lamenting the fact that its previous versions continue to be widely ignored by other pterosaur and archosaur workers. The hypothesis is based on a phylogenetic analysis of 231 multistate characters and 1090 taxa — from which the tree topology of 24 taxa relevant to the arguments of the paper were figured and discussed. The results of the entire analysis are available only on the author’s website and to my knowledge have never been published in a peer-reviewed platform.” That will change when one editor and one reviewer let the work see publication, but apparently not on this editor’s watch. 

Where the teeth are bared: “It is an understatement to say that these results differ significantly from those of other studies, with every major reptile clade, as typically recognized, being extensively paraphyletic.” All clades presented in the LRT are monophyletic and fully resolved. This editor is working from an old textbook. Taxon exclusion led to errors in prior studies. This can be readily checked by simply adding taxa and checking that all taxa document a gradual accumulation of derived traits in competing cladograms. That’s why the large reptile tree is so large, to minimize taxon exclusion problems that plague smaller studies. 

There is only one way to get published: “The author states that his character matrix is not really drawn from existing studies but rather was “largely built from scratch.” One could interpret this as an admirable attempt to shed the existing assumptions that burden other studies, but ultimately this hypothesis will never overturn existing paradigms until it demonstrates that it better explains the totality of the existing data. This study certainly does not do that. Simply disregarding a large percentage of the characters that the larger community of workers has decided are important for resolving reptile phylogeny in favor of those the author deems relevant is never going to accomplish this goal.” See below.

Sidenote: “The criticisms of the author’s approach by Hone and Benton 2007 still appear to be relevant – at least they have never been directly addressed, nor are they addressed in this manuscript.” Yes, they are. Hone and Benton excluded the fenestrasaur taxa that overturn the pterosaur origin question in Peters 2000, after promising to test them. Why is every paleo colleague afraid of Cosesaurus (the subject of the submitted manuscript)?

Case closed: “Until the author explicitly demonstrates why the characters he omits should not be included or how adding taxa and characters to an existing matrix, such as that found in the Nesbitt (2011) study, produces the promoted tree topology, I cannot recommend publishing this work or sending it out for further review.”  Problem 1: Nesbitt’s 2011 paper was on archosauriforms. Pterosaurs were thrown into that study, but they are not archosauriforms, as documented 11 years earlier. Problem 2: Adding taxa or characters to an existing matrix assumes the existing matrix is faultless. It is not. Nesbitt 2011 suffers from some inappropriate taxon inclusion and a great deal of taxon exclusion, along with some bad scoring that we looked at in a nine-part series ending here. Problem 3: Peters 2000 added taxa to four prior phylogenetic analyses and recovered the same results each time. None of those four were built on prior analyses. Now let’s move forward 11 years. Why was Nesbitt 2011 published when it mentioned, but did not include relevant taxa reported by Peters 2000? Editors and referees let that pass. Why? Have the rules changed?

Evidently it matters who the author is, and how well they are connected in the academic community, not how well a project is researched.

Editors and referees are only human. They have an agenda and a world view, like everyone does. They see what they want to see, comment on what they want to comment on and maintain whatever status quo they currently follow. How do we know this in this case? Note how little was said in this review (not one sentence) about the new pterosaur traits found in Cosesaurus, which formed the subject of this paper.

My reply:

Dear [Editors]:

Thank you for your kind reply and review.

Ultimately the number of characters or their publication history means little, since one set of two hundred characters will result in the same tree topology as another set of two hundred characters. The character list is the soft underbelly of any analysis, the part editors and reviewers go to when they cannot argue against the demonstrated gradual accumulation of derived traits shown by the included taxa, universally excluded from other studies that include pterosaurs. 

Maintaining the majority view will only keep the origin of pterosaurs in the dark. 

Best regards,

Nesbitt SJ 2011. The early evolution of archosaurs: relationships and the origin of major clades. Bulletin of the American Museum of Natural History 352: 292 pp.
Peters D 2000b. A reexamination of four prolacertiforms with implications for pterosaur phylogenesis. Rivista Italiana di Paleontologia e Stratigrafia 106: 293–336.

1 thought on “The soft underbelly of a phylogenetic analysis

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.