Re: mammalian lineages (was Re: Cretaceous taeniodont)

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

> I wonder if someone might clarify this area for me.  I've read that there
> was a third (now extinct) lineage of mammals, but I've never been able to
> figure out just _how_ they were different.

Marsupials and placentals _together_ constitute one tiny end branch on the
huge mammalian tree. Let's see what I remember of the analysis by Luo et al.
(2002) (shortened):

Mammaliomorpha
|--+--Haramiyidae (little gnawing herbivores, LTr to LJ)
|  `--Tritylodontidae (specialized l. g. h., LTr to EK)
`--the mammalian jaw joint appears here
     |--Tritheledontidae (little insectivores, LTr + EJ)
     `--Mammalia in the usual sense
          |--*Adelobasileus* (a LTr braincase)
          `--+--Docodonta (seemingly mole-like, ?LTr to start-K)
             `--Mammalia sensu Luo et al.
                  |--*Sinoconodon* (shrewlike, EJ)
                  `--Mammaliformes (have mammalian tooth replacement)
                       |--*Morganucodon* (shrewlike, EJ)
                       `--+--*Hadrocodium* (tiny, shrewlike, EJ)
                          `--crown-group Mammalia
                               |--+--*Shuotherium* (shrewlike, LJ)
                               |  `--Australosphenida
                               |       |=weirdos from MJ to EK
                               |       `--MONOTREMATA (since EK)
                               `--+--the rest of Triconodonta (MJ to ?LK)
                                  `--Theriiformes (see below)

Theriiformes
  |--Multituberculata (gnawing, LJ to Oligocene)
  `--Trechnotheria
       |--Spalacotheriidae (see below)
       `--Cladotheria
            |--Dryolestoidea (+-shrewlike, LJ to Paleocene)
            `--Zatheria
                 |=lots of MJ to EK teeth & jaws
                 `--Boreosphenida
                      |=lots of K teeth
                      `--Theria
                           |--Metatheria
                           |    |=various (LK to Miocene)
                           |    `--MARSUPIALIA
                           `--Eutheria
                                |=various (LK to Oligocene?)
                                `--PLACENTALIA

Spalacotheriidae includes *Zhangheotherium* and a lot of isolated teeth from
the K and IIRC LJ. Dryolestoidea was quite diverse, especially in the LK of
South America, where they practically constitute the only known mammals.

Only Placentalia and a few ?other placentals have the placental mode of
reproduction. All other mammals (including basal eutherians like
*Zalambdalestes* and *Ukhaatherium*) retain epipubes, the "pouch bones",
which, in female marsupials, support the pouch. These bones constrain the
belly volume and seemingly make long gestation times impossible.
        Therefore it can be assumed that the marsupial mode of
reproduction -- though not necessarily with a pouch! -- is normal for
mammals more closely related to us than to monotremes. Nothing is known,
however, on when the transition to this from egg-laying happened. It has
been suggested that the multituberculates have such narrow hips that they
can't have laid eggs... I have no idea if this means anything (unlike
archosaur and some turtle eggs, mammal eggs are soft-shelled).

>  The name seems to indicate dentition, but that seems
> too minor to declare such a radical separation.

Indeed it is -- for a long time all Mesozoic mammals whose molars had 3
cusps arranged in a straight line were lumped into Triconodonta. It has
turned out that this pattern is merely the normal state for mammals and
their relatives. Most "triconodonts" have been moved elsewhere, but a
smaller clade that consists of Triconodontidae (whose 3 molar cusps are of
identical height), Gobiconodontidae (e. g. *Gobiconodon* and *Repenomamus*)
and *Jeholodens* seems to be real.