[Taxacom] Incongruent phylogeny and geology

[David Campbell]

Geological evidence is often relatively coarse-scale, and neontologists'
ideas of geology often much vaguer still.  In the real world, connections
and gaps between continents or ocean basins do not abruptly jump from
impassible to easily traveled and vice versa; different species have
different dispersal ability and are stopped or not by different barriers.
Thus, an event such as the closing of Panama does not have a magic fixed
date at which every land and freshwater species crossed and every marine
species stopped crossing.  Land mammals (probably decent swimmers) began to
cross by about 10 million years ago, but the trickle became a flood about
2.5 million.  The southernmost part of the isthmus geologically is
geographically northwestern Colombia, and probably was in place about 10
million years ago, but shallow marine connections persisted as recently as
2.5 million.

These issues are quite commonly a problem in molecular clock calculations.
Calibrations based on tectonic events rarely involve careful examination of
the fossil record, dispersal abilities, and detailed geological analysis
that would be necessary to actually correlate a particular point on the
tree with the event in question.  Likewise, when dates are calculated, many
different things are happening geologically at any given time, and honest
error bars for molecular clocks are quite wide, so the range of possibly
correlating events is in fact very wide.