Re: No Cretaceous placental mammals?

[gerarus@westnet.com.au]

>>> The new eutherian phylogeny, produced with the inclusion of
>_Maelestes 
>>> gobiensis_ of the Djadokhta, and using a considerable spread
>>> of Cretaceous eutherians and Paleogene mammals finds:
>>> * monophyletic Euarchontaglires, Laurasiatheria, Xenarthra, and
>Afrotheria 
>>> (the latter two as sister taxa, and together the sister
>>> taxon to laurasiatheres)
>>
>>Wow! As soon as one takes a serious look at the morphological data,
>most of 
>>the molecular topology comes out! I'm impressed. :-)
>>

The Wible et al. paper didn't recover a monophyletic Laurasiatheria
or complete Afrotheria. The problem taxa were (surprising at it may
sound :-P) Lipotyphla. Once again, the afrosoricids joined with
Eulipotyphla rather than other afrotheres - the topology was
((Blarina, Erinaceus) (Solenodon (Eoryctes, Potamogale))). Afrotheres
except Afrosoricida did form a clade. The Lipotyphla + Afrosoricida
clade was sister to the Xenarthra + core Afrotheria clade rather than
to the other Laurasiatheria used (Carnivora + Cetartiodactyla). The
Templeton tests in the supplementary data don't appear to have
included tests of any more 'molecule-friendly' positions for
lipotyphlans.

The analysis didn't include Perissodactyla, Chiroptera, Pholidota or
Sirenia. That minor quibbling aside, it's still a vast improvement on
its predecessors.

>>> * the sister group to Placentalia is a clade comprised of
>_Purgatorius_, 
>>> _Protungulatum_, and _Oxyprimus_
>>
>>Quite unexpected.
>>

As I said yesterday, though
(http://dml.cmnh.org/2007Jun/msg00470.html), more traditional
positions for these taxa were not statistically rejected, so this
result should be treated with a little caution. Very interesting
though.

    Cheers,

        Christopher Taylor