Re: No Cretaceous placental mammals?

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

> Why, oh why does this keep coming up?  The fact that placentals
> divurged genetically from marsupials in the Jurassic

How is this a fact -- for one thing, it depends on the calibration --, and 
how does it fit the fact that we find the basalmost and very plesiomorphic 
meta- and eutherians in the Barremian?

> is in no way
> contradictory to the fact that crown group placentals have a more
> recent divurgence time (possibly even Cenozoic),

Of course not, but most molecular divergence date estimates _for the origin 
of the crown-group_ are deep in the Cretaceous as well (often at the end of 
the Early Cretaceous), though this again depends on the calibration.

> nor is it in
> opposition to the fact that we don't see fossils with "key" placental
> synapomorphies.  Genetic divurgence has to occur prior to phenotypic
> divurgence, and since we tend to pick "important" characters by
> reference to what we can observe, we should expect genetic divurgence
> to occur well before the evolution of the crown-group synapomorphies
> selected via post-hoc observation of extant groups as "important".

But synapomorphies of two clades have to evolve, by definition, before these 
two clades split from each other. The lack of Cretaceous fossils with 
placental autapomorphies ( = synapomorphies of Supraprimates + 
Laurasiatheria-Afrotheria-Xenarthra under the topology of the new paper), in 
the presence of plenty of Cretaceous mammals that lack these features, is 
strongly suggestive, probably even statistically significant. Likewise, the 
lack of Jurassic fossils with therian autapomorphies ( = synapomorphies of 
Metatheria + Eutheria) in the presence of plenty of Jurassic mammals that 
lack these features tells us something.

> That is, the sibling populations that were (genetically speaking) never to 
> mix again and give rise to groups that we would call placental and 
> marsupial mammals did not utilize the derived form of reproduction seen in 
> either group today.  That evolved later, most likely several millions of 
> years after the speciation event that lead to the divurgence of these 
> linneages.

Of course.

To toot my own horn, a big paper on (among other things) the strong 
dependence of molecular divergence date estimates on the choice and 
treatment of calibration points is coming out in the current issue of 
Systematic Biology: 
http://www.informaworld.com/smpp/content~content=a778976765~db=all~jumptype=rss