Biogeographic memory

[P.Hovenkamp]

At 11:00 AM 8/24/01 -0400, you wrote:
>>Granted. But still not something that Turner et al can be accused of...
>
>No, and I did not make any such accusation.

Must have been a misreading on my part then. Sorry. Some of the subtleties
of the English language will be forever beyond me...

>>I may have been hasty, basing my comments only on Croizat 1962 (Space,
>>Time, Form: The Biological Synthesis), and possibly (excusably?) even
>>overlooking some statements in those 822 pages.
>>The point remains that Craw, Grehan and Heads are quite adamant that
>>biogeographic analysis is not to be informed by phylogenetic analysis
>>based on other characters.
>
>Perhaps if some specific quotes were given here this could be discussed.

The request for quotes suggests that John Grehan does not recognize this
position as being one that is taken up in the book. That is curious,
considering the content of chapter 4, which starts with:
"An understanding of geographical evidence is *basic* to interpreting the
phylogeny, systematics and taxonomy of species and higher taxa", and in the
same vein on the following page:
"The panbiogeographic project (...) advocates that biogeographic analysis
is a necessary *prerequisite*  for a coherent presentation of phylogeny and
constructioin of a natural classification (...)" (p. 89). (*emphasis* mine).
The chapter then continues with defending the position that heritable
characters and geographical occurrence can be treated with equal standing
("Geographical distributions as a Systematic Character") and concludes with
"(...)a natural classification can only be established after, not before, a
biogeographic analysis." (p. 113).

Moreover, and pehaps more telling, Craw, Grehan and Heads give a great many
examples and illustrations, but only a very small minority of these refer
to explicit hypotheses about phylogeny. Most of this minority is taken up
by examples in Ch. 4, and deal with the prediction of phylogenetic
relations from biogeographic patterns. Phylogenetic data as input for
biogeographic analysis is obviously not what the book is about.

>>This raises the question how different two distribution patterns must be
>>to be distinguishable by such different means of dispersal. Given the
>>choice between a butterfly and a freshwater group, would one be safe in
>>venturing that a pattern ranging from New Guinea to Samoa is more likely
>>to be the butterfly and a pattern restricted to Borneo more likely the
>>fish? I do not ask for certainty - simply for a rational reason to make a
>>choice.
>
>One might make such assertions but that would be on the basis of other
>preconceptions about the origin and evolution of particular areas. The
>point about this example is that Mayr said that poor dispersing primary
>freshwater fish have totally different distribution patterns from
>butterflies. I emphasize the word 'totally'. The examples given were
>presented as a fish and butterfly group with similar distribution ranges
>with some local differences. There is no objective criterion to predict
>which range is necessarily the fish or butterfly. One may claim that the
>range in the Pacific is "more likely" to be a butterfly based on some
>preconceived position, but that's all it would be.

Would not that be enough to dismiss the claim that "their distiributions
are not distinguishable according to their different means of dispersal"?
Perhaps this illustrates the rather arbitrary way in which Panbiogeography
uses "similarity". Thus, the two distributions discussed here are treated
as "similar", despite clear dissimilarities - as in other cases tracks
treated as similar "Indian Ocean Baselines" can be as dissimilar as fig.
2-2 (connecting India to Sumatra), 3-3 b (connecting Cape to S. Australia)
or 3-6  c (connecting Madagascar and India).

>The same goes for the
>more restricted distribution of one group than the other in the Middle
>East. Means of dispersal does not provide a predicable relationship with
>distribution range or pattern.
Here, "predictable" seems to imply "predictable with 100 % certainty".
(...)

>>Quite a lot of systematic analyses pertaining to the region have been
>>performed since Croizat (1968). It might be a good idea if someone would
>>integrate these in a panbiogeographical framework. Preferably in less than
>>400 pages...
>
>Why less than 400 pages? Does this mean that such issues do not warrant
>such pagination while other issues do?

IMHO, hardly any issues warrant 400 pages. For one thing, it makes it so
difficult to trace quotations that are referred to as (Croizat, 1968).

>Hovenkamp will be pleased to know, however, that steps in that
>'integration' are current. A couple of examples are Heads on Abrontanella,
>and Heads in press (J. Biogeography) on birds of paradise. Craw of course
>has published numerous papers integrating phylogeny in general with
>panbiogeography and of course the panbiogeography book does quite a lot of
>this.

Quite a lot? Again, an analysis of the illustrated examples shows
otherwise. Hardly any of the examples deal with the complications of Wallacea.

I've just obtained a copy of Croizat (1968, all 400 pages) from our
library. To summarize this is beyond me at the moment, but a quick glance
is enough to notice that this does deal with Wallacea in rather a big way.
Nevertheless, I have not been able to find a summary statement more
specific than "The biogeographic status of Borneo is both very important
and quite complex" or "The same is true of Southern Formosa/ Northern
Luzon" (p. 389).

In addition, Turner et al. use 29 groups for which detailed data have been
published (consisting of phylogenies and distribution maps), most of which
published before 1995.

None, repeat, none, of all this work has made its way into the
panbiogeography book. Croizat (1968) is cited only in the context of
pacific or indian ocean baselines. It makes me wonder what is wrong with
Wallacean biogeography. In an earlier post, John Grehan wrote that:
"Croizat's analyses suggests the biogeography is not complicated in this
respect" . Perhaps it is yet too complicated to summarize?


P. Hovenkamp
Nationaal Herbarium Nederland - Leiden
PO Box 9514
2300 RA  Leiden
The Netherlands
hovenkamp at nhn.leidenuniv.nl
http://nhncml.leidenuniv.nl/rhb