Re: Molecular clocks and avian diversification

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

----- Original Message -----
From: "evelyn sobielski" <koreke77@yahoo.de>
Sent: Tuesday, October 24, 2006 11:18 PM

> > > A nice way of putting what a biomedical researcher
> > > seeking major founding would without doubt have
> > > called "...will revolutionize our understanding of..."
> >
> > Where's the revolution?
>
> Molecular dating finally being based on sound science.
> birds are quite state-of-the-art in the matter; other
> branches of evo-bio are not really arrived at relaxed
> clocks yet.

I hope you're kidding... :-S Are you saying fixed clocks are still used 
anywhere? :-o

> > An Early Cretaceous date for the basal
> > divergence of Neornithes is highly suspect. And so
>
> Not necessarily as deep as suggested, true. But
> possible (though not really likely).

Judging from the Yixian Fm... is it possible that they were all hiding in 
Gondwana...?

> > What's so "very distinct" about it?
>
> Vegavis was a full-fledged member of a crown-group
> family (which later became extinct due to other
> reasons)

OK, but this doesn't require a great distance in time from the origin of 
Neornithes.

> > So you think there was a trichotomous split?!?
>
> Not really, I think that trichotomous vs dichotomous
> is a matter of perspective. Evolutionary lineages do
> not necessarily have a thickness of 1 as cladograms
> might suggest.

What do you mean?

> The Galliformes are particularly interesting: anything
> with a gamefowl lifestyle seems to be so much prey in
> the Late Cretaceous ecosystem. There were terrestrial
> theropod predators, airborne enantiornithine predators
> aplenty.

That's no worse than today, IMHO.

> They were not a monophyletic group. And "graculavid"
> (as form taxon) birds do not come just form NJ, only
> the ones that were placed in the presumed "family"
> were (or at least from the general area). Vegavis
> would have been placed in the Graculavide (Presbyornis
> was).

Oh, that's what you mean. But if it's _so_ broad, "Graculavidae" is useless 
even as a form taxon.

> Paleogene radiation suggests a Late Cretaceous split
> as more likely.

Or a split in the ecological vacuum of the early Danian...

> Neornithine seabirds from the Maastrichtian ought to
> have been found if there was appreciable diversity.

I agree.

> Landbirds, ditto.

Not necessarily. The Maastrichtian fossil record of badly preservable 
terrestrial animals is pretty bad.

> > Don't confuse cladogenesis ( = any halfway permanent
> > split of a population
> > in two) and speciation. Those aren't the same...
> > except under some species concepts...
>
> I don't, but speciation is necessary for cladogenesis.

Then you are confusing them. :-) Speciation _as understood by the Biological 
Species Concept_ is necessary for _irreversible_ cladogenesis, but I'm not 
restricting myself to that. Browse the archives, and you'll find lists of 
fertile hybrids between extant "genera" -- clearly unreversed cladogenesis 
has occurred without being irrevers_ible_.

> Any lineage should have passed through a point where
> it was a lower-level taxon in the Linnean system, e.g.
> a genus or species.

The Linnean system only serves to confuse such issues. It's a smokescreen. 
:-)

> Hmmm... a drawing of the evolutionary relationships of
> these ducks would help. The result is something that
> cannot be represented by a dichotomous tree, not if
> you permit trichotomy either. A result that cannot
> come from a cladistic analysis, hence: cladistics does
> not help in this case.

Strictly speaking that's true, but the fact that introgression will be shown 
as systematic convergence on a cladogram should certainly help in 
recognizing possible introgression. Only clade origins by hybridization or 
outright lineage fusion (shown in the figure of 
http://www.ohiou.edu/phylocode/art1-3.html) will really throw cladistics off 
track -- the cladogram will then contain a trichotomy.

> This is not cladogenesis, granted - but it is a case
> for what may happen at nodes because (without further
> research) all active speciation events are possible
> nodes in a phylogeny.

I don't understand "active speciation event".

> Or: what were the first-diverged (as opposed to
> last-common) ancestors of 2 sister clades? Certainly
> living, breathing animals. Sister species? Usually.

Not even necessarily different species. They could have been populations of 
the same species -- depending on the species concept. (For example, the 
Hennigian species concept takes cladogenesis and calls it "speciation"; 
under that concept those ancestors automatically belong to different 
species.)

> (That's what I
> meant in the "trichotomy" bit on Galloanseres above:
> gamfowl and waterfowl are apparently early
> divergences).

Oh. So just two dichotomies separated by a short internal branch?