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

Re: Turtles and Crocodylians are not Reptiles - no? What are they?

On Wed, Jul 23, 2008 at 12:27 AM, Tim Williams
<twilliams_alpha@hotmail.com> wrote:
> I would say that the molecular data is even more problematic, given that the 
> vast phylogenetic diversity of the Sauropsida is so poorly
> sampled among extant taxa.  Aside from turtles, we only have two surviving 
> archosauromorph groups (Crocodylia, Aves) and two
> surviving lepidosauromorph groups (Squamata, Sphenodontia); and all four are 
> quite derived within their respective clades.  When it
> comes to inferring the affinities of turtles, I tend to put more faith in the 
> morpho-based analyses.

The number of *clades* is not so quite a limiting issue - and here, we
are take into account just the extant taxa and their relative position
on the tree. It would be some problem of sequence alignment, but one
could use just the parts that align well (and have informative sites).

Of course that I'm not saying that when there is conflict between
molecular and morphological data the molecular side always wins. But
the turtle case is very similar to the cetacean case - the then
available morphological/fossil data put the whales besides
mesonychids, while the mitochondrial DNA analysis said that they would
join with hyppos (Artiodactyla).

Probably the turtle position issue will be settled down - as occurred
in the Cetacea case - when a new informative fossil (or a collection
of informative fossils) was found and/or descripted (or eventually
fossil DNA).

Here the molecular data advantages are twofold:
1) We could be very confident about the heritability of molecular
characters (yes, sometimes we will found horizontal transmission and
hybridisation, but hardly we would find a hybrid of turtle and crocs);
2) We generally find many more informative characters in DNA segments
than in a bunch of fragmented fossil bones.

With few morphological information, the putative synapomorphy could be
interpreted as convergence/homoplasy. (The same would be said if we
have few informative molecular data.)

> Yes, both true.  Although (b) is true only if turtles are diapsids, rather 
> than parareptiles (with Testudinata as sister taxon to pareiasaurs).

But there is a lack of less derived turtle fossil that show this
affinity - i.e. fossil that is more similar to a putative common
exclusive ancestral between Testudines and pareiasaurs.


Roberto Takata