Re: Apomorphy-based definitions

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

> The apomorphy-based Mammalia continues (after many decades)
> to provide a clearcut boundary based on osteology (jaw and ear ossicles).

There are _two_ osteological apomorphies that have been used: 1. appearance
of the dentary-squamosal jaw joint; 2. Meckelian cartilage rips through in
early ontogeny, leaving the postdentary bones ( = hammer) disconnected from
the lower jaw (so that they are involved in the middle ear only). The first
apparently unites Tritheledontidae with traditional mammals* (the literature
is remarkably silent, I interpret the name *Diarthrognathus* as indicating
this; Tritylodontidae only have the old jaw joint). The second looks to be a
synapomorphy of (*Hadrocodium* + crown group). So which apomorphy do you
mean?

* Don't be confused by *Probainognathus* which has also developed an
additional jaw joint, but between the _surangular_ and the squamosal.

> When we do find transitional forms that fill in the remaining gap, we
simply
> take the very same characters and refine them a bit in light of the new
> evidence.  It's a very stable strategy.

Constantly changing the characters is a stable strategy, you think?

>       Class Aves on the other hand has not enjoyed this same kind of
precise
> osteological definition, and we are suffering the consequences.  The
"true"
> semilunate carpal block might not be quite as precise as mammalian ear
> ossicles,

Especially since it may have reversed several times (Tyrannosauroidea,
Ornithomimosauria... depends on the phylogeny).

> And eventually we may
> learn that the "true" semilunate was a key step in the development of
> powered flight.

It was a _prerequisite_ that might have developed long, long before any sort
of flight.

> And it gives a little bit of "wiggle room" to refine the definition, and
> still maintain stability without the "straight-jacketing effect" of node
and
> stem-based definitions (which I think are even more arbitrary).  All
> classifications are arbitrary, and the best we can do is to minimize it
the
> best we can.

I think when the definition is not fixed, then we'll get a headache every
time a fossil with a transitional state of the key character is discovered:
Should we apply the current definition? Should we refine it? How much? Why
or not? This is IMHO more arbitrary (and hurts more in the head) than
necessary.

>      Mammalia has demonstrated the superiority of such an apomorphy-based
> strategy,

Nope. Suppose you use the first apomorphy above and then find a fossil that
is the sister-group to (Tritheledontidae + rest) but doesn't have a
preserved jaw. You can't tell whether it is a mammal or not.

> and clearly shows how confusing and imprecise crown groups (in
> particular) are to deal with.  We don't know if multituberculates belong
to
> the crown group, but we do know they have the mammalian jaw and three ear
> ossicles.

Lately they have often been found as the sister group of Monotremata
(ignoring Australosphenida). Would put them into the crown-group. Characters
of the foot (no ref, I've seen this once in the library, I'll look for it
sometime) put them next to Triconodonta, which is also happily in the
crown-group.
        If Multituberculata and Haramiyida are closely related, as used to
be assumed and is uncertain now, then we are faced with considerable
convergence -- *Haramiyavia* has the first apomorphy, but not the second
(see above).

> but the whole eutherian classification is going to crumble when their
> polyphyletic Grandorder Ungulata is split up [...]

So what? I mean, if we believe the geneticists, then it has crumbled long
ago. Where's any problem with that?

> Clade Altungulata seems to be particularly unnatural.

There are still paleontologists who find no problem with it. The following
paper never mentions the name, but...

Emmanuel Gheerbrant, Jean Sudre, Mohamed Iarochene & Abdelkader Moumni:
First ascertained African "condylarth" mammals (primitive ungulates: cf.
Bulbulodentata and cf. Phenacodonta) from the earliest Ypresian of the Ouled
Abdoun Basin, Morocco, JVP 21(1), 107 -- 118 (26 March 2001)

"ABSTRACT---We report here the discovery of the first well identified
"condylarths" from Africa, from the phosphatic beds [...] which have also
yielded the oldest known proboscidean [*Phosphatherium escuilliei*]. [...]
show closest relationships with Mioclaenidae and Phenacodonta respectively.
[...] indicating sister-taxa relationships within Taxeopoda (Phenacodonta,
Pantomesaxonia). Moreover, [...] shares a remarkable derived feature with
more advanced pantomesaxonian ungulates (Perissodactyla, Hyracoidea,
Tethytheria and extinct relatives): the development of an entolophid. This
raises the altiernative question of their sister-taxa relationships within
Taxeopoda and indeed the question of an African origin of Pantomesaxonia,
which is congruent with the Paenungulata hypothesis [but not Laurasiatheria
and Boreoeutheria]. [...]"

Afrotheria is never mentioned in the paper. Only the conclusions end with
"This in turn would raise the question of an African origin of
pantomesaxonians stressed here, which is also in accordance with the current
hypothesis of an African origin of Paenungulata (e.g., molecular data).".

> And
> supersplitting Linnean ranks like that is why such ranks now have such a
> terrible reputation among cladists.

It's not only that. It's also the fact that nobody seems to have got an
answer to the questions I posed yesterday in response to a post by HP Tracy
Ford -- such as, what is an order?

> But total
> cladification ultimately fails because it is based ONLY on branching order
> (and fails to reflect the reality of paraphyly).

Why isn't it enough to acknowledge that _species_ can be paraphyletic?

> P.S.  If the Feducciaries are right, and the "true" semilunate evolved
> twice, then we are ALL in very very deep (up to our necks) in trouble.

The PhyloCode even makes a proposal to get around that problem. Just word
the definition as "the first organism which had a 'true semilunate'
[detailed descriptions and illustrations!] homologous to that of *Passer
domesticus* [or whatever bird] and all its descendants". :-)