Gondwana Split May Complicate Mammalian Evolution

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

> Wow!  If most modern mammalian groups are Gondwanan, and
> most (or all) modern _birds_ are also Gondwanan (see article in
> recent, one begins to wonder whether the Chixulab left much alive
> in the Northern hemisphere at all.

It's not that simple. The Afrotheria-rest split is put at 103 Ma ago by
molecular clocks, which indeed happens to coincide very precisely with the
Africa-SA split. The Xenarthra-Boreoeutheria split is put at 94 to 95 Ma
ago... this predicts Boreoeutheria must have reached NA at that time via
island-hopping or so. The Laurasiatheria-Euarchontoglires split lies between
79 and 88 Ma ago.
        Fossil evidence: A new study puts Zalambdalestidae and Glires
together. Would fit this model. The same study supports (more weakly) the
inclusion of Zhelestidae (paraphyletic or not) in Laurasiatheria. I cited an
SVP meeting abstract long ago which argues that Zhelestidae is related to
Zalambdalestidae instead. Would be less evidence for, but still no evidence
against the model. Practically no Cretaceous mammals from Africa are
known -- absence of evidence. :.-( Must have had a very interesting, because
isolated, fauna. Late Cretaceous SA, however, appears to be pretty well
known. There we have an isolated fauna composed of AFAIK peculiar
multituberculates, derived dryolestids, the enigmatic gondwanatheres and
even docodonts, not to forget *Vincelestes*, a close relative of Theria, but
so far not a single therian sensu stricto (even though one metatherian tooth
has turned up in Madagascar and an astragalus of a eutherian, called
*Deccanolestes*, is known from the end-Maastrichtian of India). Dryolestids,
gondwanatheres and the only known SA monotreme survived into the Paleocene
when therians appeared, apparently quite suddenly, just as if immigrated
from NA a very short time before the K-T.
        Eutherians in Australia in the EK would add great support to the
molecular scenario. Claims of such discoveries exist; *Ausktribosphenos* and
*Bishops* are often interpreted as such. At the moment it seems much more
convincing to me that they, like monotremes and *Ambondro*, are members of
Australosphenida, the basalmost branch of crown-group mammals. As I hear a
JVP paper defending their eutherian status is in the works... let's wait for
it :-)
        The paper cites a molecular clock date for the split between Meta-
and Eutheria... at 173 to 176 Ma ago. I refuse to accept that. I hope the
authors of that didn't accept Marsupionta. :->

William J. Murphy, Eduardo Eizirik, Stephen J. O'Brien, Ole Madsen, Mark
Scally, Christophe J. Douady, Emma Teeling, Oliver A. Ryder, Michael J.
Stanhope, Wilfried W. de Jong, Mark S. Springer: Resolution of the Early
Placental [ = crown-group eutherian] Mammal Radiation Using Bayesian
Phylogenetics, Science 294, 2348 -- 2350 (14 December 2001)

BTW, as I've said before, formal informal names like "boreoeutherians" very,
very strongly imply that there is a formal name Boreoeutheria somewhere.
Let's simply forget Altungulata, like Pachydermata or Cotylosauria*! :-)

* Which, unfortunately, has recently got a definition... it's a node-based
clade which includes Diadectomorpha and Amniota. Why did they (Laurin &
Reisz) have to bring that old discarded rhizome name back? :.-( (The
PhyloCode dislikes such definitions, too.)