dispersal fantasy

[pierre deleporte]

A 22:27 27/06/2002 +0200, Peter Hovenkamp wrote :

>To me, the basic problem here seems to be how we go about confronting
>geological theory with biogeographical evidence.

Agreed, and hence what is considered relatively strong or doubtful evidence
in geology or historical biogeography. Apparent conflict of course appeals
to improved explanation.

>John seems to want to collect a whole lot of biogeographical evidence, use
>this evidence to derive a biogeographical statement with some degree of
>generality, and then confront geology with that statement.

Seems to. I still don't grasp the general rule for making a geological
prediction from a "baseline" (series of redundant tracks), without
implementing some "biogeographical evolutionary model". Would some model be
implemented, then panbiogeography would be comparable to other approaches
in historical biogeography .

>Pierre seems to accept a "piecemeal" approach, which confronts each piece
>of biogeographical evidence with the geological theory separately, and
>returns a "narrative" which deals with the similarities and differences.
>Is that more or less a correct interpretation?

I't's a possible debate, but not my own proposition. I perfectly accept
that a general pattern emerging from largely sampled biotas (and not only
considering separate lineages independently) can be taken into account. If
such a pattern emerges from "noise", then some common historical fate of
the biotas is likely. My question in is: which pattern means what, and why?
Clearly, I suggest that not any arbitrary designed pattern goes.
Thus I question biogeographers of "areas" for the meaning of there areas
and the way they interpret the biological of geological connections between
areas, I question biogeographers of distribution limits (Peter ?!) the same
way, as so goes for proponents of tracks and baselines. What does this
mean, and how do you draw explanations or predictions from that ?

>I suggest that, in the long term (when we're all dead...) it will come to
>the same thing: when a lot of conflicting evidence emerges, something will
>have to break, and alternative theories will have to be considered.

All right for alternative theories, confrontations and new suggestions. You
mean theories, let's talk of theories.
A classic vicariance theory is that speciation was principally allopatric,
that areas of endemicity may be defined, and that redundant pattern of
areas connected by the phylogenetic relatedness of the taxa they harbour is
indicative of general splitting events (barriers) having affected the
biotas in the past, or alternatively of oriented dispersal across such
barriers.
A variant vicariance theory (Hovenkamp's approach, in short) considers the
same way that allopatric speciation is the general rule, but that the
problem at stake is the identification of barriers, and that redundant
distribution limits between sister taxa is the right pattern to be searched
for (rather that cladograms connecting pre-defined areas of endemicity), by
the way eluding the problem of very large distributions (as irrelevant)...
All this is debatable.
Now, panbiogeography looks after redundant tracks = baselines. But what is
the panbiogeographic theory?

>  That
>may be a geological theory (new evidence for land bridges across the
>Mozambique channel), or it may be about the biogeographical evidence
>(someone finding chameleons can swim after all..., or revealing fatal flaws
>in the cladograms).
>Ultimately, it's about strength of evidence. I would need pretty strong
>evidence to convince me of the effectiveness of homeopathic medicine - and
>some recent transoceanic dispersal hypotheses come close to that in their
>implausibility in the light of other evidence...

Yes. You mean evidence. My point is that patterns of any kind are not
relevant evidence for anything, without an interpretative theory. A purely
"structural" biogeography disconnected from interpretative theories would
not make more biological sense than a purely "structural" systematics.

Pierre.