[Taxacom] Patterns of homoplasy

Bob Mesibov mesibov at southcom.com.au
Mon Aug 18 19:11:56 CDT 2008


There's a memorable courtroom scene in Oliver Stone's film 'JFK' that I
often think of when looking at tree-form phylogenetic hypotheses.

New Orleans DA Jim Garrison is tracing the course of the bullet that
killed Kennedy, *if* the findings of the official inquiry are accepted.
The bullet's track, with its supposed richochets and passages through
people, is wildly improbable. It's far more likely that the
interpretation of evidence by the official inquiry is flawed.

The connection with phylogenetic hypotheses goes like this:

We use character evidence (sequences, morphological character states) to
generate a set of phylogenetic hypotheses, and tentatively propose
(usually) a single hypothetical tree based on that set. We know the
assumptions involved and we're aware of the uncertainties involved in
the 'phylogenising' method of choice. (Although as Richard Zander has
pointed out, many phylogeneticists don't know how to determine those
uncertainties, or aren't interested in them.)

But inspection of the 'best' tree(s) always reveals character conflict,
or possible (minor?) disagreements with the evolutionary model we used
to link raw sequences in a branching network. You might well then say to
yourself, 'Hmm. So this structure - if I accept the hypothesis - appears
here, then disappears in the lineage, then reappears later, and
meanwhile appears independently over here. Wonder if other structures do
that?'

What's more interesting to me is that I don't see this bullet-tracking
in the systematic literature. It's as though homoplasy (in the case of
morphology) is something of an embarrassment. We have our tree, we don't
want to think about what that means for character evolution. I'm aware
that there are papers and books dealing with patterns of homoplasy, but
these seem to focus on similarity as a biological phenomenon, not on the
ins, outs, ricochets and surprising passages of individual homologue
bullets.

I'm wondering if study of homoplasy and sequence anomalies could reveal
widely occurring patterns of character and sequence evolution. Having
discovered these, we might then have a new tool to assess a tree's
plausibility.

The first problem would be digging out those bullet-tracks from the
systematics literature. Not many systematists do a Jim Garrison and say,
'If we accept this best-supported tree, this means...'
-- 
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
http://www.qvmag.tas.gov.au/mesibov.html





More information about the Taxacom mailing list