Cladistics can be fun (joke)

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

After I finished my bird matrix (40 ingroup taxa, an allzero outgroup, 51  
characters), I accidentally started the calculation before I found out how  
to make some multistate characters ordered but keep others unordered. This 
guarantees that the outcome is nonsense, because some of my characters 
must obviously be ordered (e. g. the number of sacral vertebrae). 
        I got 120629 most parsimonious trees (well, after 1.3 billion 
branch-swapping replications I stopped it). I checked out 26 of these and 
was surprised. They are pretty funny, for example *Lectavis* is always a 
euornithine and *Yungavolucris* is always the sistergroup of *Sinornis*, 
and pretty interesting, for example *Iberomesornis* and *Noguerornis* are 
sistergroups in 99 % of the trees even though I didn't include the one 
character that optimizes as a synapomorphy for them in the analyses by 
Chiappe et al., and *Liaoningornis* is always a euornithine. Resolution of 
Enantiornithes is surprisingly good, for example Avisauridae the way I 
mentioned it earlier today comes out in 53 % of all trees: 
 
--53--*Concornis* 
   `--92--*Eoalulavis* 
       |--*Neuquenornis* 
       `--70--*Soroavisaurus* 
           `--*Avisaurus* 
 
. Don't ask me how *E.* got in there. 
        If anyone is interested in these 26 (or fewer), as well as the 
strict consensus bush and the 50 % and/or 60 % majority-rule consensus 
trees, I can send them as .tre files. :-) If you don't have PAUP*, 
TreeView or suchlike, you can open them as text files and will see the 
tree written as (A((B,C)(D(E,F)))). 
        You'll notice that I didn't add a character to take the 
plesiomorphic long tail of *Yandangornis* into account. :-} 

-- 
+++ GMX - Mail, Messaging & more  http://www.gmx.net +++
Bitte lächeln! Fotogalerie online mit GMX ohne eigene Homepage!