Re: something's wrong here: Qianosuchus phylogeny

[Mike Taylor <mike@indexdata.com>]

David Marjanovic writes:
 > > Really?  What is the reason for knowing a priori that scleromochlids
 > > could not possibly be paraphyletic with respect to pterosaurs?
 > 
 > There is only one known scleromochlid, *Scleromochlus* itself, and
 > it has autapomorphies, so the most parsimonious assumption is that
 > it's not an ancestor of anything else we know -- never mind its
 > geological age, which IIRC isn't older than the oldest known
 > pterosaurs.

OK, but that's an accident of what we currently happen to know about
scleromochlids.  I guess you were making a point about that specific
taxon; I misread your comment as meaning that no taxon ever can be
directly ancestral to any other -- which of course is what I was
disagreeing with.

 > And then there's the widely acknowledged principle that only
 > monophyletic taxa should be named above the species level.

Widely acknowledged, no doubt; but certainly not universally
implemented!

 > If there were two scleromochlids known, and the group as a whole
 > turned out to be paraphyletic, that would mean one of the
 > scleromochlids had turned out not to be a scleromochlid.

If a cladistic definition is in use for Scleromochlidae, yes.

 _/|_    ___________________________________________________________________
/o ) \/  Mike Taylor    <mike@indexdata.com>    http://www.miketaylor.org.uk
)_v__/\  "The shell inparts commands, addresses memory for programs and
         executes via the Unix pipe" -- ix MAGAZINE, April 1985