biogeographic homology

[John Grehan]

Peter Hovenkamp's response concerning the distributin of a hypothetical
taxon represented in the Netherlands, Japan, and Columbia illustrates the
contrasts in methodology and application of homology concepts in biogeography.

The first issue concerns level of differentiation. I am inclinded to agree that
if a taxon is apparently undifferentiated between these localities, a recent
establishment of the range is a likely possibility. However, there have been
cases where such distributions have later been shown to actually comprise
distinct species so some caution is required.

Hovenkamp responds to the qeustion of how to determine what geography
is involved by correlation of the taxonomic ordering of the component
taxa with (1) each other and (2) earth history.

The question that remains is what earth history? How is the choice to be made
from biogeographic analysis. In most if not all cases of phylogenetic approaches
this invovles a preference by the athour that a particular earth history is
true (earth history is, after all a story or theory about the past which may,
or may not, be true to an entire or partial extent). If the postulated sequence
matches, a shared history is considered a likely possibility.

The trouble with such approaches is that there is no biogeographic input or
orientation of the distribution. The distribution is seen as uninformative about
the probable geographic sector involved. Instead, preference is drawn from
external disciplines such as geology. It is my view that such approaches render
biogeography non-scientific since the biogeography itself is not assigned any
empirical contnet (i.e. the spatial structure of the distribution is itself
uninformative).

A classic example of the differences in biogeographic technique is that of
Nothofagus. Traditional approaches interpreted the origins by matching the
 group's cladistic sequence with the breakup of Gondwana. Now that some
cladistic reconstructions don't match, some biogeographers are trying to
match the cladogram with the postulated breakup of Pacifica. In either case
the biological relatinships are being matched with hypothesized historical
sequences with no biogeographic indication as to which geographic sector
is inolved. Lately, if I understand it correctly, one cladistic study renders
the cladogram of Nothofagus completely uninformative anyway.

Panbiogeographic analysis places the homology of the group with a
Pacific baseline according to the criteria of minimum spanning distance
between localities that have their main massing along parts of the the Pacific
basin. This spatial correlation leads to the hypothesis that the Pacific basin
is more likely to be involved with the historical origin of the distribution
than any other major geological/geomorphic feature of the earth.

I've probably gone on longer than most would have the patience for, but to
address the last point about the hypothetics distribution not corresponding
to any actual distribution - there may be no taxa with these specific localities
along, but there are taxa which are present in Europe, Japan, and western
South Amreica. The hypothetical example would nest within such a pattern,
and therefore belong to the same standard (or generalized) track.

John Grehan