E&P of Pterosaurs, Unwin - part II

[david peters <davidrpeters@earthlink.net>]

On the phylogeny and evolutionary history of pterosaurs - by D. Unwin 2003

This is part II of an earlier discussion.

In the â??80s and â??90s pterosaurs were nested tenuously within â??the 
Ornithodiraâ?? even though dozens of the characters of that clade did not 
match. Here in the â??00s, it would appear that some of the same sorts of 
problems are appearing. 

After discussing some of the current pterosaur origin hypotheses, D.Unwin 
(2003) wrote: â??At present, it is not clear which, if any, of these hypotheses 
is correct. Happily, however, with regard to the polarization of characters 
used for establishing ingroup relationships with pterosaurs, this is largely 
irrelevant, because pterosaur skeletal anatomy is so derived that in almost all 
cases the plesiomorphic condition is common to each of the three outgroups used 
in this study: basal ornithodirans, basal archosaurifors and prolacertiforms.â??

Clearly D. Unwin did not want to invite the higher prolacertiforms to his 
matrix. 

The outgroup in Davidâ??s matrix, known as â??outgroupâ?? was scored with all 
zeroes, but note the following, which should have been scored (1), not (0), if 
these taxa were chosen: 

>From D. Unwinâ??s Appendix 2:

1. Dentary: more than 75% length of lower jaw. Longisquama {teeth below the 
orbit}.

6. Quadrate: inclined anteriorly. Longisquama.

9. Rostrum: low with straight or concave dorsal outline. Cosesaurus, Longisquama

11. External nasal opening: low and elongate. Cosesaurus, Longisquama

12. Nasal process of maxilla: inclined backwards. Cosesaurus, Longisquama, 
Sharovipteryx

13. Maxilla-nasal contact: broad. Cosesaurus, Longisquama

14. Orbit: larger than antorbital opening. Cosesaurus, Longisquama.

15. Ventral margin of skull: curved downwards caudally. Longisquama, 
Sharovipteryx.

19. Two, large, fang-like mandibular teeth: absent. Cosesaurus. 

48. Sagittal cranial crest...soft tissue...: present. Like crazy in 
Longisquama. Less so in Cosesaurus.

Many of the above characters might also score (1) for Sharovipteryx, but the 
top half of the skull is unavailable and other parts are difficult to judge. 

I did not attempt to check characters among ornithodirans or basal 
archosauriforms mostly because Iâ??m not paid by their lobby and previous 
experience has taught me that they are not germane to the issue. Also, whether 
the results change or not with these changes is irrelevant to this discussion. 
Also, I have not combed the rest of the cladogram as I have the "outgroup". So 
whether the rest of the cladogram is spotless or similarly corrupt is unknown 
to me at present.

Precision will make or break a cladogram. As one of the keepers of the 
pterosaur torch, Dr. Unwin needs to be more precise. Pterosaurs are not 
orphans. They belong to the Prolacertiformes. They are not â??so derivedâ?? 
from ancestral forms. Pterosaurs are baby longisquamids with minor 
modifications.

Additionally, D. Unwin did not make note of pedal digit V -- which is elongated 
in higher prolacertiforms and pterosaurs -- and almost dissolved in the other 
two putative outgroups. Also various issues surrounding pectoral keel 
possession, pelvic shape, hemal arch and caudal transverse process 
retention/loss were ignored.

Hereâ??s where the cladistic problems start. If the characters are not keyed in 
correctly and without prejudice, we can have little faith in the outcome of the 
computations.

If Iâ??ve made any errors above, please bring them to my attention. 

Weâ??ve got to do better next time.

David Peters
St. Louis