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

Re: Sobral and Langer 2008 (was pteros have lift-off)

Tim Williams wrote:

However, I can only re-iterate David M.'s message that supertrees do
not contain any new data.  They pool together pre-existing trees.<<<

(sorry, in my haste to clear my in-box yesterday I had already deleleted David's email, which I'm sort of paraphrasing from here)

I'm not sure I agree with the definitions of "data" and "testing" implicit in the statement(s) that a supertree has/does niether. I'm sure most phylogenetic workers have rerun published trees to check and make sure the original authors did it correctly, and by any working definition I use checking results would be "testing" a dataset. Moreover, if you achieved a different result, by any useful definition that new result would be new data, even though the input dataset was itself (presumably) not new. Of course if you were the one in error it may turn out to be useless data, but that doesn't change the fact that new results from running the "same" data is very much itself data.

To be sure, a supertree may not be terribly convincing by itself (and all of the other problems with possible result bias mentioned by David and Tim are real) but we shouldn't dismiss the results of a meta-analysis a priori on methodological grounds, we need to look at the matrix and see where bias (e.g. resulting from mixing incongruous/largely non-overlapping data matrices) exists.

Note this isn't for or against any of the phylogenetic hypotheses being bandied about for pterosaurs, just a quick observation on methodology.

Scott Hartman
Science Director
Wyoming Dinosaur Center
110 Carter Ranch Rd.
Thermopolis, WY 82443
(800) 455-3466 ext. 230
Cell: (307) 921-8333


-----Original Message-----
From: Tim Williams <tijawi@yahoo.com>
To: dinosaur@usc.edu
Cc: davidpeters@att.net; tijawi@yahoo.com
Sent: Thu, 22 Jan 2009 5:40 pm
Subject: Re: Sobral and Langer 2008 (was pteros have lift-off)

David Peters wrote:

Sobral, G. and Langer, M. 2008 JVP 28: supplement to (3):
145A A supertree approach to prolacertiform phylogeny

Thanks. However, I can only re-iterate David M.'s message that supertrees do
not contain any new data. They pool together pre-existing trees.

No taxon appears suddenly in the fossil record!

Sure they do. It happens all the time. How many basal bats (chiropterans) do
we know of, for example?

There are always clues to ancestry.

I agree. But in the absence of basal forms, these clues often arise indirectly
- such as by examining the character states that derived members of a clade
share with members of other clades. This is what we're forced to do with
pterosaurs. (And bats as well.)

That's a very bad
sign when the 'preferred' cladogram can't nail
down a sister taxon.

It does indeed give us a sister taxon. Pterosauria (or a _Scleromochlus_ +
Pterosauria clade) comes up as the sister taxon to Dinosauromorpha. A cladogram
will always provide a sister taxon (more than one in the case of a polytomy).
The issue here is the degree to which the recovered sister taxon helps elucidate
the ancestral morphology or ecology of pterosaurs. So far, it hasn't helped a
great deal. But that's not the fault of cladistics, or of Hone and Benton; it's
the fault of the fossil record, which hasn't yet yielded basal pterosaurs.

Yes! Tim, good work!
If you're referring to Pteromimus, it is indeed related
to pterosaurs. It's a langobardisaur The first one
recorded for North America, but cladistically its
pre-Cosesaurus but it shows they too had an antorbital
fenestra. Very important.

If Procoelosaurus is also on your Tecovas list, it's a
basal croc. Both interesting specimens.

Yes, I was referring to 'Pteromimus' and 'Procoelosaurus'. Interesting stuff
regarding their respective affinities - though I look forward to something on
this in the published literature. From what you say of _Pteromimus_, I wonder
how much material potentially overlaps with _Protoavis_.

Yes, you're right, but let's take it up a notch.
Forelimb length is about 100% of the hindlimb length in
Archaeopteryx. i'm still wondering, what is the
explanation for such an extension on this taxon -- or any
pre-Archaeopteryx taxon -- with such elongated forelimbs,
non-supinating/pronating antebrachia and trenchant manual
unguals if not for arboreality?

Two other alternatives:

(1) Prey capture. Longer forelimbs allowed increased reach. Targeting large
prey that was grasped with both hands allowed a decrease in supination/pronation.

(2) Gliding (or some other form of non-powered aerial locomotion). Longer
forelimbs allowed a larger flight surface, and the arms were used for climbing
vegetation. This is partly concordant with your scenario.

I'm not necessarily advocating either of these scenarios. I'm only saying that
alternate hypotheses to arboreality/brachiation do exist as an explanation for
further forelimb elongation in the line leading to birds.