Re: Tuataras are the roadrunners of molecular evolution.

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

I've now read the paper. I expected a tetrapod phylogeny or something, where 
*Sphenodon* has a longer branch than *Alligator*... nope. It just takes 
subfossil tuatara DNA (8800 to 500 years old), compares it to modern tuatara 
DNA, and calculates a rate of evolution assuming all samples represent a 
single population through time. Also, the DNA in question is the 
mitochondrial control region, not a gene.

> I was wondering if the rate estimate might be inflated by assuming that 
> the
> early Holocene bones were ancestral to modern samples, whereas common 
> sense
> would suggest they might well be distinct species, or in any case that 
> their
> mitochondrial lineages had split much earlier.

I agree. The subfossils come from all over North and South Island.

> though there is mention there of sequences
> clustering by geographic region regardless of date.
> Is the analysis OK, do you think?

As explained in the supplementary information, that analysis has two steps. 
First, a neighbor-joining analysis that was used to get a topology. This 
topology was then fed into a maximum-likelihood analysis to get branch 
lengths. The reasons for this bizarre approach aren't explained anywhere. 
It's bizarre because:
1) neighbor-joining is phenetics, not phylogenetics;
2) neighbor-joining alone produces both a topology and branch lengths;
3) maximum-likelihood alone, too, produces both a topology and branch 
lengths.
The logical approach would have been to use maximum-likelihood alone. Unless 
I've missed something -- but given the complete lack of explanations, I 
can't figure out what that could be.

> On the other hand, what would hyper-variable mitochondrial control regions
> have to do with rate of morphological (or indeed physiological) evolution?

As far as I can tell, nothing whatsoever...