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

Re: Pteromimus and pterosaur origins



----- Original Message ----- From: "David Peters" <davidpeters@att.net>
Sent: Saturday, January 24, 2009 4:50 PM


Note: only three generic taxa are represented here.  All the rest are
suprageneric. Which leaves open the question, which specimens/genera
served as poster child for each of the suprageneric taxa?

Probably none -- or he'd have said so. If he had used *Prolacerta* instead of all of Prolacertiformes, he'd have said so.


Two possibilities: Either he coded the whole known morphological diversity. In that case, expect to find polymorphisms in the matrix. (It's always good to have a look at data matrices. They're not as boring as one might think.) Or he reconstructed the MRCA of the group and coded that -- which requires of course a phylogeny of the group.

It goes without saying that both approaches can still go wrong. If the OTU is para- or polyphyletic, and both are serious and underresearched possibilities for Prolacertiformes, the data matrix will contain a fictitious taxon, and that leads to "garbage in, garbage out". So, I agree Atanassov didn't adequately test the phylogenetic position of the pterosaurs and his two new taxa.

Atanassov referenced Peters 2000, so why he chose not to include
Longisquama, Sharovipteryx and the others can only be attributed to
not being able to personally observe them.

At least *Longisquama* is clearly supposed to be included in Prolacertiformes. (Which is a bad idea -- Senter 2004.)


There is also the question of cladistic _analysis_, which MacClade is great at. After thePAUP* tree(s) is(are) created it's time to evaluate it(them). MacClade is great at showing exceptions and mistakes in coding.

Yes, except that that's not called "cladistic analysis"; that's what PAUP* does.


For instance, note that all of the sister and predecessor taxa of pterosaurs in Atanassov's phylogeny have a vestigial or absent pedal digit V. Pterosaurs don't. That's a red flag that should tell any reasonable person, evolution doesn't work that way

Wait, wait, wait, wait, wait. Be very careful with this kind of assumptions. Natural selection is not constrained by history; if a population with a reduced 5th toe that only consists of 2 phalanges anymore finds itself under selective pressure for elongation of that toe, that's what's going to happen. Even lost phalanges can come back; how many phalanges a digit contains depends on the way it grows in the embryo -- a metapodial and its phalanges start as a cartilaginous rod that grows from proximal to distal and segments as it grows --, and, according to an SVP presentation I saw last year, the basal anomodont *Suminia* did reverse from 2-3-3-3-3 to 2-3-4-5-3.


The same for manual digit IV.

Quite so. As long as it's there, it can grow.