[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

RE: Pterosaurs as basal archosauriforms

And yet Bennett still didn't add Longisquama, Cosesaurus or Sharovipteryx, 
despite citing Peters (2000) which argued they were key taxa for placing 
pterosaurs outside Archosauriformes.  So Peters will justifiably say the 
analysis is pointless, even ignoring the many suprageneric OTUs and use of 
Lepidosauromorpha as an outgroup.

I was greatly amused by- 

"Hone kindly sent me their Nexus files, and
comparing them with the published data sets of Bennett (1996)
and Peters (2000) revealed that the inability of Hone and Benton
(2007) to replicate the analysis of Bennett (1996) was not the
result of differences in the coding of Scleromochlus for Char. 43,
but rather was the result of (1) four separate coding errors in retyping
the data matrix and (2) leaving eight of the 9s that Bennett
(1996) used to code for missing data in the matrix unchanged
while using ? as the symbol for missing data and then including 9
in the Symbols statement of the Nexus file so that PAUP treated 9
as a distinct character state rather than as missing data. Similarly,
their inability to replicate Peters' (2000) analysis of a modified
Bennett (1996) data set was the result of seven coding errors in
retyping Peters' published data matrix plus two 9s treated as a
distinct character state."

Coding errors due to retyping is one thing, but leaving 9 as a separate state 
instead of unknown?  Twice?!  First I'm appalled Hone and Benton apparently 
tried so little to determine why their versions of each matrix didn't produce 
the same result.  Second, I'm appalled none of the peer reviewers thought to 
check why this happened.  Surely one would notice the matrix contains both 9s 
and ?s but otherwise only 0s, 1s and 2s.  Third it makes Hone and Benton's 
statement below particularly ironic-

"The unexpected results produced by Peters (2000) and the reanalysis here, 
suggest that his codings should be examined more closely, particularly with 
reference to the pterosaurs. There are numerous methodological errors 
throughout the paper as well as errors in the interpretations of some specimens 
and the resulting codings."

Of course Hone and Benton were also the ones to leave Lepidosauromorpha as an 
OTU in a supermatrix that also included Gephyrosaurus, Sphenodontia nd Squamata 
(and an identical mistake with Choristodera and three choristodere genera), so 
maybe I shouldn't be surprised.

Mickey Mortimer

> Date: Mon, 24 Sep 2012 08:37:41 -0700
> From: bcreisler@gmail.com
> To: dinosaur@usc.edu
> Subject: Pterosaurs as basal archosauriforms
> From: Ben Creisler
> bcreisler@gmail.com
> A new online paper:
> S. Christopher Bennett (2012)
> The phylogenetic position of the Pterosauria within the
> Archosauromorpha re-examined.
> Historical Biology (advance online publication)
> DOI:10.1080/08912963.2012.725727
> http://www.tandfonline.com/doi/full/10.1080/08912963.2012.725727
> A previous analysis of the phylogenetic position of the Pterosauria
> argued that pterosaurs were not closely related to dinosaurs as is
> generally accepted, but rather were outside the crown group
> Archosauria. However, that study was dismissed for the use of
> inappropriate methods. Here, the data set from that analysis was
> divided into five partitions: one with characters associated with
> cursorial digitigrade bipedal locomotion and the other four with
> characters from the skull and mandible, postcranial axial skeleton,
> forelimb and hindlimb, respectively. The partitions were subjected to
> homogeneity testing, and the Cursorial partition was found to be
> incongruent with other partitions and all other characters at the α =
> 0.01 probability level. Deletion of the Pterosauria removed all
> significant incongruence, demonstrating that the incongruence results
> from the coding of pterosaurs for the cursorial characters. The cause
> of the incongruence was interpreted as homoplasy in hindlimb
> morphology, and after re-evaluating and reformulating the characters
> of the Cursorial partition, the revised data set was tested for
> homogeneity and no significant incongruence was found. Lastly, the
> data set was updated with additional characters and taxa from recent
> analyses, tested as before, and when analysed suggested that the
> Pterosauria were basal archosauriforms well outside the crown group
> Archosauria.