[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

AW: Shaking up the bird family tree



I won't discuss the article, but 
http://bwfov.typepad.com/.a/6a00d8341cda4153ef00e5538e6ce88834-800wi

Left, reading downwards:

clade H: Very good. Many isolated lineages are liable to reshuffling, but 
overall a nice phylogeny.

turacos and "clade" I: Look how weakly supported they are! These are not 
clades, man! Cuculiform and Gruiformes s.str. resolution looks nice tho.

Hoatzin: yeah right. Well it hangs out with cuculiforms and turacos, which may 
be relevant.

Clade L: cypselomorphs. Oilbird goes mit Caprimulgiformes s.str., which is 
probably relvant.

Clade M: Good.

The pigeon-and-so-on stuff: The weird thing is that it would perfectly resolve 
most of what we known (which is not enough and hitherto contradictory). Huh.

The nodes the news tout here are VERY weak. The Mirandornithes are however a 
group that would need to be discussed on their own, like clade M. Distinctness 
of tropicbirds is assured. Mesites and sandgrouse, maybe, and pigeons WTF? But 
I do not think it can be falsified (it's more reasonable than the proposed 
"Cretaceous" origin of Columbiformes). The tropicbirds have a deep fossil 
record, and it shows that they are very VERY deep. The later radiation has a 
very lousy fossil record.

What happened to sandgrouse as charadriimorphs?

Galliformes: Yeah, that's how they look. Nice taxon sampling BTW. Only thing 
missing might be a grouse, just to flaunt them (as Neogene pheasants in arctic 
gear).

Anseriformes: _Thalassornis_, _Cyanochen_, _Pteronetta_, _Plectopterus_ would 
be reasonable additions that come to mind. _Biziura_ and _Oxyura_ are so weakly 
supported (they are probably much closer; the signal of _Oxyura_ seems skewed). 
Putting at least 1 of the above and a shelduck in would probably improve the 
resolution immensely.

paleognaths: WTF?! But the internal relationships are correct *relative* to 
each other, if you put the base a bit crownwards of the base of the tinamous. 
What to make of that IONO.

(right side read upwards)

Charadriiformes are too advanced. The fossil record indicates *quite* strongly 
that they should precede Ciconiiformes et al. (Turnix is far far lower; IIRC 
the fossil record of such birds much predates anything that is like a gull or 
an auk.)

Clade E: I would not trust this. Support is not particularly strong, and one 
major branch *might* be genomically aberrant (they are notorious for "wild" 
behavior in molecular phylogenetics) Accipitrid resolution in as good as it 
gets with 4 taxa. But accipitrids are OLD as far as anyone knows. (Where would 
horusrnithids go?)


Coliiformes + Strigiformes? Why not. But *what* about the Strigiformes and 
cypselomorphs?

(What happens if you removed the whole of the Charadriiformes? Look how weak 
the support is in a whole section of the tree!)

Clade D: _Leptosomus_ is messing this up. What happens if you remove it?
Trogons with the other guys? Perfect!

Clade C: According to what I have seen, Passeriformes would attach near the 
internal node that has 99% support; go figure. "Woodpeckers evolved from 
Coraciiformes", well, as if hoopoes and hornbills were ever especially 
representative of Coraciiformes... Glad we're rid of that. _Brachypteracias_: 
Heh. The whole thing looks MUCH more beautiful if you map in the fossils from 
Messel et al. The clade from the trogons onwards, has a beautifully diverse 
stem record.

seriemas: What happens if you prune them? They may be close to Falconiformes 
(s.l.), but how do THEY get to be THERE?

Nodes B+A: Though Shalt Not Discuss "clades" With Less Than 80% ML Support. At 
least not in mol-phyl.

It's not the first time something vaguely resembling this is found, but 
everything else suggests that it's a glorious case of long branch attraction. 
In some studies Passeriformes go near the BASE of the entire neoavian tree; 
they are notoriously jumpy but morphologically *rather* distinctive.
Falconiformes are liable to lose their partner (hawks) and go wild.
What would happen if the Falconiformes were removed? Would the NW vultures and 
accipitrids pop towards clade H perchance? And the Passeriformes and 
psittaciformes dislocate in and around clade C?
Parrots are also "jumpy" but there is a pattern. They are not seldom are 
resolved in the general vicinity of the near passerines.

Acanthisitti: There you have it again. Nice.

Suboscines: That's basically how they look like, yes. Fossils are scarce, but 
the crown biogeography says a lot.

Basal oscines: Nice taxon sampling makes the "Corvida" of DNA-DNA hybridization 
fame a neat Henningian comb.

_Picahartes_ would immensely benefit form adding any 2 of:
* a petroica
* a fairy-bluebird
* a hyliota

Bad taxon sampling makes the non-Passeriodea Passerida suck. Add at least 4 of 
the following to sort it out:
* Bearded Reedling
* dipper
* flycatcher-tit
* mockingbird
* Palmchat
* Philippines "treecreeper"
* reed-warbler
* silky-flycatcher
* starling
* sugarbird
* sunbird
* tit
* (Eurasian) treecreeper
* wren
* Wrentit


In summary, the "backbone" of the tree looks flawed. There are clear and good 
clades recognizable (superorders if thats your game), but the position of about 
1/3 of these relative to each other runs contrary to what other information we 
have. It's not the first time something like this happens.

The actual branch lengths will probably show immediately what the problem is.

Cracraft et al (2004) have a very nice collection of trees from sundry data to 
look up against this.

The K-Pg boundary is someplace in this "uncertainty region" that resists 
resolution.


Regards,

Eike


      __________________________________________________________
Gesendet von Yahoo! Mail.
Dem pfiffigeren Posteingang.
http://de.overview.mail.yahoo.com