chloroplast and other genes (was Lucy in Newsweek)

Richard.Zander at MOBOT.ORG Richard.Zander at MOBOT.ORG
Wed Apr 7 09:04:57 CDT 2004

I figure there is a definite limit to statistically distinguishing branch
arrangements. Sure, with more information, you can get better phylogenetic
resolution, for instance, in flipping a coin, the chance by chance alone of
6 trials out of 10 trials is much greater than 60 out of 100 or 600 out of
1000, though the proportions stay the same. Problems exist, however:

A large percentage of genes are not quite suitable for particular problems
because they are saturated or they evolve at too slow or too fast a rate, so
there is a limit on genes available to solve a problem.

Each gene can have a different history so you can't just add all
information. Even though a branch arrangement for a particular gene may be
highly probable, the species history may be different. You have to identify
the gene histories that match the species histories first (at each
internode), THEN add the data before analyzing it.

Sequence alignment may be ambiguous, especially if parsimony or gap costs
figure in the alignment. Any change in an arrangement of interest caused by
slightly unparsimonious arrangements or gap costs of a different arbitrary
size must lower the final probability of being correct.

In the "Tree of Life" there are doubtless many branch arrangements
determined by "strict cladism" where the justification is that the
arrangement is "best." This is nonsense but stays with us.

Then there is sample error, and other problems often cited in papers but
ignored under the assumption that they are not problems.

Of course, many molecular branch arrangements agree with morphological
arrangements, but why can't there be problematic taxa that simply can't be
phylogenetically resolved? This suggestion is a possibly extreme counter to
the extremism of the last 30 years in which phylogenetic arrangements are
ALWAYS presented as resolved in publications except those arrangements that
collapse into multifurcations because the support pro and con is EXACTLY the
same. If statistics is of any use, then arrangements should be collapsed
into multifurcations when support pro and con is indistinguishable from
chance variation.

Richard H. Zander
Bryology Group
Missouri Botanical Garden
PO Box 299
St. Louis, MO 63166-0299
richard.zander at <mailto:richard.zander at>
Voice: 314-577-5180
Fax: 314-577-9595
Bryophyte Volumes of Flora of North America:
Res Botanica:
Shipping address for UPS, etc.:
Missouri Botanical Garden
4344 Shaw Blvd.
St. Louis, MO 63110

More information about the Taxacom mailing list