Panbiogeography of the Americas

[John Grehan]

Stuart Poss wrote

>This sounds to me as if you concede that track analysis requires more prior
>information than can simply be inferred by examining the "track" alone when
>deciding which of these hypotheses (or the several other logical
possibilities
>that you did not specifically mention) provides us with the correct
explanation
>for the observed distribution.

I think you misunderstand what tracks are about. They are about spatial
relationships, so I don't think I am conceding anything not conceded already.

  While we can draw a track, it is not clear that
>we can be sure what it means.

A track describes the spatial geometry of a distribution. Yes, its
"meaning" is potentially
multiple. However, when one identifies a track homology, and offers
correlation with tectonic patterns, it is possible to assert or predict a
possible historical relationships
between the origin of the biological and geological patterns.


>The problem with curving coastlines is more evident in the tracks drawn by
Rosen
>in his 1976 paper in Systematic Zoology on the origins of the caribbean fish
>fauna.  It is unclear to me why all the logically possible "tracks" showing
>floridean affinities were not drawn.  I would suspect it is because he
thought
>some routes were more likely than others to have been taken and not
because of an
>intrinsic requirement specifying track geometry of the panbiogeographic
method.

I have never analysed Rosen's track method so could not defend it one way
or another.

I must admit I still have not located the Page
>reference that was alluded to in an earlier post.  Does any one have the full
>citation to this work?

I don't have the full citation to hand, but one of his papers was in a 1987
Systematic
Zoology issue if I recall correctly.


>If I understand your argument up to this point, so far the only part of the
>panbiogeographic algorithm I have been able to program my computer to
perform is
>to draw lines among the shortest disjunct distances among southern  beeches.
>I still do not see the algorithm as providing me a single explanation for
their
>distribution that is more informative than that provided by Hooker many years
>before Croizat.

The track by itself explains nothing. However, the identification of a
Pacific track homology and its correlation with Pacific tectonics becomes
more informative. Whether Croizat was any more informative than Hooker is
not an issue to me at least.

 Perhaps there is something about the biology and phylogeny of
>these plants that is more interesting than their panbiogeographic tracks.

Perhaps

John Grehan