Re: Fixing dinosaurian carnivour question

["T. Mike Keesey" <tkeese1@gl.umbc.edu>]

On Sun, 6 Jun 1999, Stanley Friesen wrote:

> A *quick* summary:
> 
> 1. Derive a cladogram by any of the usual methods.
> 
> 2. Annotate each internode with a measure of the change it represents (its
> length)
>     - the simplest, but often least useful, measure is simply the number of
> synapomorphies on it.
>     - a slightly more sophisticated variant is to weight the synapomorphies
> inversely by their frequency on the cladogram (thus a unique synapomorphy
> has the highest weight, and a highly convergent character has very low 
> weight).
>     - more sophisticated measures can be devised, as desired.
> 
> 3. At any given taxonomic rank, "cut" the cladogram at those internodes
> having a weighted length greater than a selected threshold - perhaps using
> some statistical measure of significance or information content to
> determine the threshold value.

Interesting! Of course, it does look a bit hard to apply on a large
scale...

Although I find virtually no utility in ranking taxa, this practice
actually could come in handy for me -- namely in figuring out the optimal
way to divvy up the taxon pages on my site. In anticipation of the site's
revival (September at the latest), I have been preparing a new version
with more taxon pages, e.g. where I once had Ornithothoraces & Neornithes,
I now have Enantiornithes, Ornithurae, Hesperornithiformes, and
Neornithes. I've been (mildly) agonizing over where it would be best to
draw such arbitrary cuts, but perhaps this method could be of use.

Now if only there were a recent cladistic analysis (with character matrix)
of all Ornithodira ...

--T. Mike Keesey                                    <tkeese1@gl.umbc.edu>
WORLDS                                  <http://www.gl.umbc.edu/~tkeese1>
THE DINOSAURICON                               <http://dinosaur.umbc.edu>