AW: To Wong Foo, Thanks for Everything! New Papers

[evelyn sobielski <koreke77@yahoo.de>]

> Morgan-Richards, M., Trewick, S.A., Bartosch-Harlid, A.,
> Kardailsky, O.,
> Phillips, M.J., McLenachan, P.A., and Penny, D. 2008. Bird
> evolution:
> testing the Metaves clade with six new mitochondrial
> genomes. BMC
> Evolutionary Biology 8:20. doi: 10.1186/1471-2148-8-20.
> 
> ABSTRACT:
> Background
> Evolutionary biologists are often misled by convergence of
> morphology and
> this has been common in the study of bird evolution.
> However, the use of
> molecular data sets have their own problems and phylogenies
> based on short
> DNA sequences have the potential to mislead us too. The
> relationships among
> clades and timing of the evolution of modern birds
> (Neoaves) has not yet
> been well resolved. Evidence of convergence of morphology
> remain
> controversial. With six new bird mitochondrial genomes
> (hummingbird, swift,
> kagu, rail, flamingo and grebe) we test the proposed
> Metaves/Coronaves
> division within Neoaves and the parallel radiations in this
> primary avian
> clade.
> Results
> Our mitochondrial trees did not return the Metaves clade
> that had been
> proposed based on one nuclear intron sequence. We suggest
> that the high
> number of indels within the seventh intron of the
> ß-fibrinogen gene at this
> phylogenetic level, which left a dataset with not a single
> site across the
> alignment shared by all taxa, resulted in artifacts during
> analysis. With
> respect to the overall avian tree, we find the flamingo and
> grebe are sister
> taxa and basal to the shorebirds (Charadriiformes). Using a
> novel
> site-stripping technique for noise-reduction we found this
> relationship to
> be stable. The hummingbird/swift clade is outside the large
> and very diverse
> group of raptors, shore and sea birds. Unexpectedly the
> kagu is not closely
> related to the rail in our analysis, but because neither
> the kagu nor the
> rail have close affinity to any taxa within this dataset of
> 41 birds, their
> placement is not yet resolved.
> Conclusion
> Our phylogenetic hypothesis based on 41 avian mitochondrial
> genomes (13,229
> bp) rejects monophyly of seven Metaves species and we
> therefore conclude
> that the members of Metaves do not share a common
> evolutionary history
> within the Neoaves.

I have been waiting for this.

I did a little background research on my own and as it seems, indel/transposon 
data is not the silver bullet it was presumed to be. But we're getting data on 
its restrictions, which is good.

Insertion data seems to be more reliable than many sequences for periods of 
about 10 Ma, as long as the (theoretically) shared transpositions are "hot". If 
they "cool", reliability drops. One study in mammals (checking ungulate 
phylogeny) found a homoplasy rate of (at least) 25%. Apparently the integration 
of transposons is not *that* restricted apart from disruptive effects, but 
their maintenance over dozens of Ma is only possible in a few selected 
positions. About excision of transposons, les is known ATM (I think). For 
non-transposon indels, the same would seem to hold true, only these events are 
generally rarer.

Indels/transposons give good signal though for the Late Miocene to Recent, and 
maybe even for more ancient times, but in the latter case they should only be 
used to decide between hypotheses supported by other data, not to support or 
falsify hypotheses (largely) on their own. There is a galliform paper which I 
need to check out; perhaps the signal can be improved by using A LOT of such 
data.

I would not call the kagu's place unexpected. For although the "Metaves" are 
not good (incidentially cleaning up a whole load of fossil record problems), 
several individual clades that were thought to be in them are good. The 
Mirandornithes an the cypselomorphs are supported by a broad range of data. So 
this may well hold true for the "Gondwanan pseudo-/para-gruiforms" as well. But 
no member of this putative clade is in the present study except the kagu.

Columbiform evolution continues to be extremely enigmatic. Given that 
didines/raphids are embedded in a (more or less?) rather basal position in the 
columbids, no non-columbid stem clades are known. IONO why we don't have a 
complete mt genome from "Volantirattus urbanus" (i.e. _Columba livia 
domestica_) already; I would have thought we had.

The proposed position of the Mirandornithes is very satisfying. Insofar, the 
comments I made about _Telmatornis_ in the parrot thread are given additional 
weight - the pre-split Mirandornithes would thus be just like primitive 
charadriiforms, but with the odd Mirandornithes shared apomorphy. Whether this 
can be gleaned from so fragmentary a fossil I'd have to check, since at least 
in the Podicipediformes there is no (undisputed) basal fossil record at all, 
and the earliest Phoenicopteriformes are highly apomorphic already. But both 
differ so vastly that one cannot tell which of the presumed unique apomorphies 
in either crown group are really plesiomorphic, retained from the basal 
Mirandornithes (where it was an apomorphy) in one lineage but nmot the other.


Regards,

Eike


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