Re: Adding taxa or characters?

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

> I seem to remember a discussion on-list a while back about whether
> it's better to add more taxa or more characters to a phylogenetic
> analysis, and that there's a paper that reaches a conclusion on this.
> Unfortunately, I don't remember the reference.  Anyone know it?

There are several. One is by Alan Graybeal in Systematic Biology in 1998. I 
think there's at least one in the June 2003 issue of JVP... if not, search 
the DML archives. -- In any case, if you keep adding characters beyond some 
ratio of characters per taxon, they give you less and less additional 
information, while there's no such limit for adding taxa (as long as there 
are enough characters). There are probably no morphological analyses that 
have reached the character limit yet, however (maybe Livezey & Zusi's opus 
magnum, no idea), so adding characters is very unlikely to be a waste of 
time.

What there is anyway in the June 2003 JVP is a heap of papers on whether to 
add taxa or characters in spite of the missing data they produce. Short 
answer: Yes, yes, yes! :-)