Re: (Paleognath monophyly)

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

> >Pygostylia (misnomer)
> >   |--Confuciusornithidae
> >   `--+--?*Jibeinia*
> >      `--+--?*Longipteryx*
> >         `--+--?*Protopteryx*
> >            `--+--?*Eocathayornis*
> >               `--Ornithothoraces
> >                    |--the diversity of Enantiornithes (paraphyletic?)
> >                    `--Euornithes
> >                         |--*Patagopteryx*
> >                         `--+--*Yanornis*
> >                            `--+--?*Yixianornis*
> >                               `--+--*Apsaravis*
> >                                  `--Ornithurae sensu strictissimo
> >                                       |--Hesperornith(iform)es
> >                                       `--+--*Ichthyornis*
> >                                          `--Neornithes
> >
> > [...]
>
> The strongly forked dentary, if it is a character reversal, is readily
> explicable within a neotenic context, during which see reversal and the
> expression of primitive characters.

Of _so_ primitive characters? I mean, the phylogenetic distance between
Confuciusornithidae and Neornithes is considerable... BTW, there is little
evidence about whether the condition in Confuciusornithidae is plesiomorphic
or autapomorphic. Is that region of the dentary known for *Sapeornis*, the
sistergroup of Pygostylia?

(To me, the strongly forked dentary looks like an adaptation to some way of
biting. But that's speculation.)

> Cranial pneumaticity caudal to the
> quadrate and the inflated alaparasphenoid are the strongest of the
> characters listed, and the only two which Cracraft & Clarke produced that
I
> would call "unambiguous,"

("Unambiguous" on a cladogram just means "couldn't be a synapomorphy of a
bigger or smaller clade on this one cladogram".)

> however, two such characters do not amount to an
> overwhelming advantage to the paleognaths-are-holophyletic concept,

:-) How many characters argue for any alternative?

> and at the very least, it that seems
> that multiple possible phylogenies of
> Paleognathae are equally viable.

How do we measure this, if not by counting which phylogeny explains the data
in the simplest way?

> > >From _where_ within Neognathae?
>
> As mentioned tinamous seem to be more closely related to the gallinaceous
> birds than they are to other ratites.

OK. So you seem to propose

Neornithes
|--Palaeognathae
|   |--Lithornithiformes
|   `--Struthionidae
|--Aepyornithidae
|--Dinornithidae
|--Apterygidae
|--Casuariidae
|--Rheidae
`--Neognathae
     |--Neoaves
     |--Anseriformes
     `--+--Galliformes
        `--Tinamidae

I think you resolve the big basal polytomy? If so, please tell us how.

> Yes, Houde (1988) which I think Mortimer derided as out of date, presented
> multiple ppssible phylogenies of the Paleognathae, which Feduccia
concurred
> with in 1996.

I won't have access to Houde (1988) soon, so I turned to Feduccia (1996).
        On p. 263 he reproduces a figure from Houde, namely the "Sterna,
coracoids, and humeri of a pheasant (*Phasianus*) and a paleognathous bird,
a tinamou (*Tinamus*), showing the striking similarities in the deeply
incised posterior region of the sternum and in the morphology of the
coracoids." I know far too little about neornithine humeri and coracoids to
judge their similarity. The sterna, however, seem to share only one
character: poor ossification which leaves only the cranial end, the keel and
the lateral processes. (Reminds one of *Eoalulavis*.) This poor ossification
may be related to the lifestyle of these rarely but strongly flying birds;
that lifestyle may well be a synapomorphy for them (it's a "primary
homology", means, it would be coded as the same character in a cladistic
data matrix), but it's just one.
        p. 273: "As Houde and Olson concluded [in the description of
Lithornithidae in 1981], 'The new fossil birds reported here are probably
remnants of what may have been a diverse radiation of paleognathous
carinates. [...] Tinamous and ratites may have descended independently from
various families or orders within this radiation of paleognaths, or some
ratites may have evolved secondarily from neognathous birds through neoteny'
(1237)." In other words, Houde & Olson (1981) proposed two contradicting
hypothesis (paleognath holophyly and polyphyly), and didn't even try to
decide.

What do you think of Mlíkovský's opinion, which says that lithornithids are
simply tinamous? He lumps them all into Tinamidae. Feduccia (p. 274)
mentions that lithornithids "are in fact very like tinamous in many
features", mentions only their [well ossified] sterna and their curved toe
claws as differences to tinamous, and quotes Houde (1988, 107): "I cannot
overemphasize the similarity of the Lithornithiformes and the Tinamiformes";
"analysis of the Lithornithiformes is most appropriately made [...] with the
Tinamiformes, the group to which they are phenetically most similar and to
which they presumably diverged", which should imply that Houde interpreted
Lithornithiformes as the sistergroup or as part of Tinamiformes.
        But then it comes (still on p. 274): "Houde denies that any derived
characters unite the monophyletic Palaeognathae" -- a testable hypothesis
that has been tested by Cracraft & Clarke (2001) -- "and proposes that the
living paleognathouse birds are paraphyletic, with the tinamous and
neognathous birds being sister-groups; this would be equivalent to figure
_b_ in his diagram (see p. 276)."
        This diagram is:
Neornithes
|--Palaeognathae
|    |--Lithornithidae
|    `--Ratitae
`--Neognathae
     |--Tinamidae
     `--neognathous birds
        And this explicitely says that ratites are holophyletic. All that
happens is that the tinamous migrate from the base of Palaeognathae to that
of Neognathae, making the traditional contents of Palaeognathae
paraphyletic. Would be interesting to add the 5 characters of Cracraft &
Clarke (2001) to Houde's analysis... oh wait, he didn't conduct an analysis,
he just made educated guesses, one would first have to compile a matrix from
the characters he used.

> Under Houde's hypothesis, some paleognaths are more closely
> related to and independently derived from a grade of basal, paleognathous
> carinates such as the lithornithids, while others are more closely related
> to and perhaps secondarily derived from neognaths (e.g., tinamous).

Is Houde's text more vague (...vaguer?) than his figures? -- Oops, probably,
because in the figure lithornithids are holophyletic... and on p. 277
Feduccia continues: "Among the strangest of all living birds, the New
Zealand kiwis are thought by Peter Houde (1988) to be derived from the
*Paracathartes* grade of the lithornithids." Now what does this mean?
Perhaps (including information from a figure on p. 275)

Palaeognathae (misnomer)
|--*Lithornis celetius*
|--*Pseudocrypturus*
|--Aepyornithidae
|--Dinornithidae
|--Casuariidae
|--Rheidae
`--Ratitae (misnomer)
     |--Struthionidae
     `--+--*Paracathartes*
        |--*Lithornis promiscuus*
        `--*Apteryx*?

Still quite a hefty polytomy...

> > Trivial characters?
> >
> >Of course not. But they are _fewer_ than the apomorphies of Palaeognathae
> >plus the apomorphies of Neognathae. :-) All these would have to be
> >explained as homoplasies in order to make tinamous and
> >galliforms sistergroups, which means a higher number
> >of ad hoc assumptions than the alternative.
>
> It means a stringent application of parimsony analysis, it does not equate
> to just-so ad-hoc hypothesis formulation.

Ad hoc or not, each additional assumption of homoplasy is one additional
assumption. And Occam's Razor means we should minimize the number of
additional assumptions.

> After all, convergence seems to be rampant in birds.

Then we could just as well give up entirely all phylogenetic work on birds.

> >No, it doesn't -- evolution happens. Is there reason to think it suggests
> >more than different locomotory adaptations -- indeed some are cursorial
> >while others are mediportal --, or perhaps multiple losses of flight
(which
> >are suggested anyway by geography and a few ontogenetic data)?
>
> It is possible, but what uniquely derived characters are present in the
> pelves to unite ratites?  To my knowledge there are none.

Then so what, as Mickey wrote. As long as the rest of the skeleton has
enough ratite apomorphies to counter the number of similarities of certain
ratite pelves to those of... which neognaths?, the simplest assumption will
continue to be ratite monophyly.

> Given this, why should we presume that the osteology
> of the pelvic girdle in these forms is a function of common ancestry,
>  and not reflective of polyphyletic derivation of these birds?

Because these bird don't consist of pelves only. :-) (Unlike many Mesozoic
mammals, which seem to consist of "the tooth, the whole tooth, and nothing
but the tooth"...)

> >Just how tremendous is it? How tremendous does it
> >have to be that you think it couldn't have evolved from
> >the structure of an exclusive common ancestor?
>
> The differences are sufficient to delineate four palatal configurations
> within the paleognathous palate, which McDowell did in the late 1940s (48,
> IIRC), and possibly five if one considers the Malagasay aepyornithids.

Fine, do some of these configurations have synapomorphies with certain
neognathous configurations? Houde seems to think (according to his figure
mentioned above) that the configurations of both kiwis and ostriches evolved
from the configuration of *Pseudocrypturus*... do you agree?

> I have already outlined the principal case for the neotenic and
polyphyletic
> status of Paleognathae, and ratites.  If you desire further enumeration, I
> would suggest checking the far more detailed sources from which I have
> derived this standpoint (e.g., Houde 1988, Feduccia 1996, etc.)

OK, Feduccia lists those, but they (such as poorer skeletal fusion) could be
plesiomorphic just as well as neotenic.

> I am not aware of any extensive cladistic analysis of the
> Tinamidae, but it would appear that _Tinamus_ is the most basal member of
> the family.  IIRC, incubation is carried out by the male in Tinamidae,
with
> nest-building and parental care subsequent to hatching, shared.

Does this hold for *Tinamus*? If so, it would appear that "parental care
[sensu lato] primarily by males" is plesiomorphic within Tinamidae, which in
turn would be consistent with Cracraft & Clarke's optimization of this
character as a paleognath apomorphy.