Re: Morpho v molecular (was Re: Tinamous: living dinosaurs)

[David Marjanovic <david.marjanovic@gmx.at>]

>  As for the content and position of Whippomorpha within Artiodactyla,
>  well that's a different matter. A whippomorph-ruminant clade is
>  pretty unparsimonious in Spaulding et al.'s data- it takes 31 more
>  steps. But there are lots of ways molecules conflict with morphology
>  in current analyses. In Squamata, molecular phylogenies don't have
>  Scleroglossa.

There is no morphological analysis of squamate phylogeny that is worth 
mention in this respect. There simply isn't. The molecules can say 
whatever they want without opposition.

Even Conrad's impressive attempt -- a big step in the right direction!!! 
-- has _way_ too few characters for its taxon sample, and _way_ too few 
fossils in its enormous taxon sample, which probably explains the 
_completely_ novel results such as snakes being just another clade of 
limbless skinks. Somebody should do their PhD on just the things that 
Conrad didn't have time to do.

Personally, I like the molecular topology. Lacertid-amphisbaenid 
transitional forms are turning up now, and Iguania appears in the fossil 
record in the Late Cretaceous (the much older fragments from India that 
could, I hear, just as well be sphenodontians notwithstanding), while we 
have gekkotans, scincomorphs and anguimorphs-or-something in the Middle 
Jurassic. My unqualified impression is that the traditional 
classification which puts Iguania as the sister-group to all other 
squamates is simply based on very few characters.

>  In Mammalia, they don't have Ungulata.

Neither does any morphological analysis (too small as they all are at 
the moment). Only romerograms have Ungulata.

>  Similarly, there are many molecular relationships that aren't backed
>  up by morphological evidence (like you say for archosauromorph
>  turtles),

Or crurotarsan turtles.

>  You never see more genes and sequenced taxa leading to molecular
>  results that resemble the traditional morphological results after
>  all. The changes always seem to go one way.

Since 2001, yes, with a few minor exceptions perhaps.