Paraphytlly is real.
rzander at SCIENCEBUFF.ORG
Wed Jan 16 09:11:14 CST 2002
A lot of this is arguing about how many angels can dance on a cladogram.
Both inference and probabilistic demonstration of phylogeny are used in phylogenetic estimation. With molecular data sets, a non-parametric test like bootstrapping will give a probability that a particular arrangement of the three lineages terminating any internal branch is due to chance; and a bootstrap value of 95% or greater is pretty good evidence of a phylogenetic signal in a gene tree. You need at least three agreeing gene trees (an no contradiction), however, to get a 95% chance or greater that the gene trees reflect species evolution (given a pre-selected 95% confidence level).
Morphological data sets, however, because of the paucity of characters, seldom actually demonstrate probabilistic inequalities of support for the three different arrangements of lineages terminating an internal branch, but instead good support must be based on inference; even with no support for contrary arrangements of the three lineages terminating an internal branch, you simply can have 100% confidence in the inference of a phylogenetic relationship based on evolution (fossils, microevolution) demonstrated elsewhere.
So we have confidence in inference of well-supported morphological relationships, and confidence in high bootstrap values for molecular results. In the first we have no actual demonstrated signals, and in the second few demonstrated signals at a decent confidence level.
Given that demonstrated reliability is low for the published results of the past 30 years of phylogenetic analysis, arguing about how to treat paraphylly in nomenclature is a bit premature, I think. Although there are uncontested groups, most relationships that now warrant discussion need more research first. We need more genuine phyly before we can profitably discuss paraphyly.
Richard H. Zander
Emeritus Curator of Botany
FNA Editorial Center at Buffalo
The Buffalo Museum of Science
1020 Humboldt Pkwy
Buffalo, NY 14211 USA
email: rhzander at sciencebuff.org NEW EMAIL ADDRESS! (Note h infix)
voice: 716-896-5200 x 351 FAX: 716-897-6723
BFNA home: http://www.buffalomuseuofscience/BFNA/BFNAmenu.htm
FNA home: http://hua.huh.harvard.edu/FNA/
----- Original Message -----
From: Ken Kinman <kinman at HOTMAIL.COM>
To: <TAXACOM at USOBI.ORG>
Sent: Tuesday, January 15, 2002 4:04 PM
Subject: Re: Paraphyly is real (+ strict cladist defined)
> We can only approximate the
> reality through inference, so conventions will always be necessary.
More information about the Taxacom