[Taxacom] Patterns of homoplasy
mesibov at southcom.com.au
Tue Aug 19 22:04:45 CDT 2008
I'm asked what I mean by 'evaluation of trees'. It's not a new process,
it's been happening in systematics for a long time.
In my own area of arthropodology, relationships of certain higher taxa
as 'treed' from certain nuclear protein-coding genes were rejected by
virtually every specialist when the work was published. Rejection (a
kind of evaluation) didn't happen because there was something faulty
about the tree-building method used, or because morphology-based trees
were culturally entrenched. Rejection happened because the specialists
said: 'If this tree is correct, then we've completely misunderstood
numerous morphological homologies. That's *unlikely*.'
The same genes were then shown to yield an anomalous tree (I'm being
very polite here) in another higher taxon, in which separate trees from
mtDNA, rRNA and morphology were more or less congruent. Here the
evaluation was: 'Three independent data sources converge on an answer
different from the fourth. The fourth is *unlikely* to be correct.'
Taxacomers will have other examples. My point is that no matter how
soundly or shakily based your tree-building method, the output tree(s)
should be evaluated in the light of evidence *other than* the particular
morphological character states or sequence alignments that were the
inputs. I suspect that character-tracking could help in this.
Dr Robert Mesibov
Honorary Research Associate
Queen Victoria Museum and Art Gallery and
School of Zoology, University of Tasmania
Home contact: PO Box 101, Penguin, Tasmania, Australia 7316
(03) 64371195; 61 3 64371195
More information about the Taxacom