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

Re: Chelonians and Small et al's Archosauromorph

> Molecular data seem to give an emphatic support to the
> position of turtles with extant archosaurs rather than
> with the living lepidosaurs.

Sometimes they do find them within lepidosaurs, though. :-) The
morphological position is neither in archosaurs nor in lepidosaurs but far
outside both, as the basalmost living sauropsids, right in Pareiasauria (in
a dwarfed subclade), with many supporting characters.

I've seen a paper that tried to use molecular clocks to date divergences
among birds. (I've lamented about it to the list long ago.) They used crocs
(*Alligator mississippiensis*) and turtles (*Trachemys scripta*) as
outgroups and the croc-bird split, which can be guessed pretty well from
fossils, for calibrating the tree. Fair enough, but turtles came out as the
sistergroup of crocs, and this divergence was dated to the end of the
Triassic when both crocodylomorphs and turtles already existed! Means, their
sequences are too similar for them to be close relatives. :-P But see below.

> With less confidence it
> also suggests that the sphenodon is an archosaur
> rather than a lepidosaur.

Never seen that. Weird.

> we have to accept this findings

Do we? :-)

> Also one has to keep in mind that the times are
> changing. Long ago the amount of sequence data was
> relatively small and

today it still is. There are extremely few turtle sequences. The only 18S
rDNA sequence from a turtle (*Trachemys scripta*) is short and full of
mistakes, or so I've heard.

> Hence any day sequence data is superior
> to morphological data in phylogenetic reconstruction.

If the alignment was correct... and if there are enough taxa in the
matrix... now a basal pareiasaur genome would be nice...

> but he was clearly missing the
> point that convergence over hundreds of sites is not
> probable in the current models of sequence evolution.

Sure it is. :-) And of course things like misalignment guarantee a wrong
outcome -- such things can't happen with morphology.

> So we need to chase the origins of turtles within
> archosauromorphs rather than elsewhere as Reisz and
> Lee had done in the past.

Lee still does.

> Like tyrannosaurs,