[Taxacom] Phylogenetic classification?

[Richard Zander]

Taxacom has a large number of users who simply read the sometimes
entertaining messages and seldom participate actively in discussions, a
much smaller group of fanatics with special agendas, and an even smaller
number of blessedly centered people with clear ideas of how science
works. What Taxacom needs is a lot more fanatics.

For example, gratis:

What's all this about "trees"? A phylogenetic tree can be completely
replaced by a nested set of parentheses. Darwin's Tree of Life (page 95
of the paperback reprint I have) has actual taxa giving rise to
derivative taxa along the central caulis and branches. The present Tree
of Life project much in the news should actually be named the Nested
Parentheses of Life. 

One might admit that phenetics and cladistics both share a method that
produces nested clusters that may be modeled as trees, with a node
representing a pair of parentheses, but one might say cladistics uses a
model of evolution in analysis. But it really isn't a model of descent
with modification, is it? It is more a more limited model of
cladogenesis. But wait a minute, is it really even that? 

Parsimony only works if the data are statistically I.I.D. (independent
and uniquely distributed), but fixation of traits during speciation is
hardly random. Suppose we have ((AB)C)... and ((AC)B)... as possible
terminal configurations with AB supported by three traits and AC by 2
different traits. IF these traits were independent, then there is more
support for ((AB)C) ... BUT if the three traits were linked selectively,
while the other two traits were, say, one neutral and the other
selection-based, then ((AC)B ... would be the correct configuration.
This is always the problem, where details are NOT even commonly
parsimonious although the cladogram at low resolution seems quite
satisfactory.

Are molecular traits independent and uniquely distributed? Given that
whole genes and associated non-coding regulator sites are commonly used
in molecular taxonomy, linkage and therefore convergence is to be
expected (in addition to simple codon bias). Check discussion on page
1267 of Grant's paper:
http://www.mobot.org/plantscience/resbot/Repr/Add/Grant-Incong2003AJB.pd
f 

or if this URL is too long

http://tinyurl.com/lrzy55 



*****************************
Richard H. Zander 
Voice: 314-577-0276
Missouri Botanical Garden
PO Box 299
St. Louis, MO 63166-0299 USA
richard.zander at mobot.org
Web sites: http://www.mobot.org/plantscience/resbot/
and http://www.mobot.org/plantscience/bfna/bfnamenu.htm
Non-post deliveries to:
Missouri Botanical Garden, 4344 Shaw Blvd., St. Louis, MO 63110
*****************************