kennethkinman at webtv.net
Sun Nov 7 22:21:21 CST 2010
I haven't read the paper, but locating nodes
"centroidally" seems like it could often be too simplistic and very
misleading. Especially given the much longer history of fossil
exploration in certain parts of the world (such as Europe and North
America), but far less in places like Africa or parts of Asia.
For instance, before hummingbird fossils were discovered in
Europe, a "centroidal" location for the known living and fossil
hummingbirds would be somewhere in the middle of the Americas. So with
the more recent discoveries of older hummingbirds in Europe, does that
place their ancestor "centroidally" in North America?
This ignores the possibility that hummingbirds actually
originated in parts of the Old World that haven't been sampled nearly as
much as Europe. They could have started out somewhere in Asia and
spread eastward through Europe and then the Americas, or they could have
started in North America and spread southward into South America and
eastward into Europe. I don't see how "centroidally" locating a node
would help us much in such cases. When sampling is poor, and especially
when preservation is problematic (hummingbird bird bones are certainly
not robust, so preservation and sampiing are huge problems), things
getting extremely uncertain and any number of different scenarios could
have yielded the presently known distributions.
It sort of reminds me of "main massings" in panbiogeography. It
might sometimes work, but when things are not so simple and
straightforward, a "centroidal" approach might not be appropriate
(whether we are taking fossils into account or not).
Bob Mesibov wrote:
An article in the latest issue of Systematic
Biology attempts to address the placelessness of phylogenies by
generating space-time diagrams with explicit procedures:
Kidd, D.M. 2010. Geophylogenies and the Map of Life. Systematic Biology
If you're interested in this subject, I recommend that you read this
paper critically. Spoiler alert: after a marvellous introduction, the
author chooses to locate nodes at points 'centroidally' intermediate
between locations for the two terminals. This needs work.
More information about the Taxacom