Re: emu history

["Tim Williams" <twilliams_alpha@hotmail.com>]

David Marjanovic wrote:

> I was surprised to see that he had mtDNA from five genera of moas, and that 
> they cluster out quite reasonably, so the kiwi placement seems to be rather 
> sound.

Alas, just because one part of a phylogenetic tree "makes sense", this does 
not necessarily mean that the entire tree is good.  Some DNA sequences are 
better at capturing deeper divergences than others, depending on whether the 
constituent gene(s) evolve 'fast' or 'slow', and on the method of analysis 
used to crunch the data.  Unfortunately, you cannot get the best of both 
worlds by combining 'fast' and 'slow' genes, and hoping the right tree comes 
out in the wash.  (This is usually the intention behind sequencing WHOLE 
mitochondrial genomes, rather than relying on a small number of genes.  This 
doesn't usually solve the problem.)

Thus, the recovery of a monophyletic moa clade in a given molecular study (a 
result that is strongly supported by morphological data) does not 
automatically mean that deeper branching events can be treated with equal 
confidence (even if they have strong bootstrap support).  To give an 
analogous example, mitochondrial-DNA analyses of mammals typically recover 
most of the modern mammal 'orders' as monophyletic, but jump around at the 
base of the tree, making interordinal relationships difficult to interpret.  
For ratites, it may be that published mitochondrial-based analyses are 
giving us the shallow clades (moas; emu+cassowaries), but are misleading 
with regard to basal ratite relationships.  In this respect, they may not be 
too different from anatomy-based phylogenies.



Tim

_________________________________________________________________
Find high-speed ?net deals ? comparison-shop your local providers here. 
https://broadband.msn.com