More on Biogeographic memory

P.Hovenkamp Hovenkamp at NHN.LEIDENUNIV.NL
Tue Oct 9 11:50:21 CDT 2001

At 10:50 AM 10/8/01 -0400, John R. Grehan wrote:

>>If this discussion has made clear that there is indeed a distinction, even
>>if it is not absolute, it has served some purpose.
>Progress in panbiogeography certainly did not stop with Croizat so it
>stands to
>reason that there will be differences.

Readers interested in Croizat's methodology will need to read Croizat then
- there seems to be no "friendly" summary then that does not go
significantly beyond Croizat's approach.

>>I would have thought that an author of a book would have some idea about
>>what is the major conclusion - but then, I have no first-hand experience.
>Did I indicate otherwise? As I said, in my mind the reader is entitled to
>decide what "the" major conclusion is if they want to. I do not see it as
>my role to impose "the" major conclusion on anyone. If Hovenkamp feels that
>the major conclusion of panbiogeography of the book is that major
>biogeographical regions are the major ocean basins, he is entitled to make
>that reading. In the book we did not make a formal declaration of there
>being a single major conclusion as such (i.e. our major conclusion for the
>book is ....). Each chapter ends with conclusions. Perhaps the first
>chapter speaks for panbiogeography in general (abbreviated slightly here)
>(1) Panbiogeography is a constructive approach to biogeography.
>(2) It attempts to resolve the dispersal-vicariance opposition.
>(3) Panbiogeography emphasizes the analysis of locality distribution data.
>(4) The panbiogeographic method does not assign taxa to already defined
>biogeographic elements, regions and areas of endemism.
>(5) Panbiogeography can lead to novel biogeographic hypotheses.
>(6) Panbiogeography reaffirms the importance of the geographical context to
>understanding the history of life.
>(7) In biogeography there is only location, location, location.
>If I were to be asked what I thought of as "the" major conclusion of
>panbiogeography, it might well be summed up by (7). Such a conclusion may
>be viewed as profound or trite. So each to their own.

And I would call these starting points - hardly conclusions. If they are
conclusions - they are often curiously vague: "attempts to", "emphasizes",
"can lead".
But I admit it might have been clearer if earlier I had asked for the major
*result*  of Panbiogeographic analysis.

>>Is it really possible that it has escaped your attention that a a lot of
>>work in historical biogeography is concerned simply with finding
>>commonality in patterns for different groups?
>No it has not.

Then why the remark that "vicariance biogeography seems to simply be a
version of Darwinian biogeography" you made earlier?

>  I would note that the commonality in patterns for the most
>part has been concerned with either geographic overlap, or biological

Agreed. In vicariance biogeography (and in panbiogeography as well) we can
observe commonality in geographic overlap and biological relationships. And
very little commonality without these two, I might add.

>However, there is nothing new in finding commonality in
>patterns for different groups. I have to disagree with myself on statement
>about Darwinian biogeography not being able to predict biogeographic
>history without some other external criterion. Of course Darwinian
>biogeography does make historical predictions about centers of origin and
>dispersal for example, without using external criteria - although it does
>adopt external information on geological hypotheses, for example, to frame
>the predictions. I am not aware of Darwinian biogeography ever making novel
>predictions about geological structure however.

Here, it may be useful to introduce the distinction between "Earth history"
and "Taxon history" biogeography which I made earlier (Hovenkamp, 1997). In
brief: Earth history is about inferring common patterns, which can be used
to make predictions about geological structure, Taxon history is about
developing a coherent explanation for the current distribution pattern of a
particular taxon (often carried out for a group of taxa forming a
monophyletic group).
Grehan's "Darwinian biogeography" seems to be virtually synonymous with
"Taxon history", and thus does not aim at making novel predictions about
geology. The criticism is invalid. It may be John Grehan's view that
explaining the distributions of individual taxa is not a useful enterprise
- but that would leave us with a large number of unexplainable phenomena.
It would also run counter to the general trend in the CGH
Panbiogeography-book, where most examples (and many of Croizat's examples
also) are concerned not with finding common patterns (generalized tracks)
but with explaining single distribution patterns.

>>McLennan. But Turner et al. are very clear: they "hoped to detect common
>>patterns (...)" (p. 221).
>Given that Croizat had already demonstrated comprehensively that there are
>common (shared) spatial patterns in biogeography it would not be going
>against the wind to have such a hope.
>  But then it
>>forces these spatial relationships into patterns dictated by geology or
>>tectonics (by assigning "baselines" in the way that it does).
>"Forcing" is not a term I would have applied.

But which I applied deliberately - I see no better term to describe the way
in which different biological patterns are assigned the same baseline
(examples cited in earlier posts).

>  The technique is simply one
>of spatial correlation between spatial biological and geological patterns.
>It is this spatial correlation that provides the biogeographic criterion
>for hypothesizing a historical relationship by correlating a pattern
>(biological) with a pattern (geological) in contrast to Darwinian
>biogeography where the correlation is between a pattern of biological
>relationship and a hypothesized geological history (represented as a
>narrative or cladogram).

This makes the distinction one of degree: CGH Panbiogeography simple makes
use of less detailed geological data.

>>So there is
>>*no* independence.
>There is independence of method with respect to constructing tracks, nodes,
>and main massings.

Baselines significantly left out here. Q.E.D.

>Of course a methodological connection with geology is
>then proposed as otherwise there would be no criteria for hypothesizing a
>particular historical association.
>>And I do not understand the remark that the track method is explicitly
>>constructed using phylogeny. "A track is a line drawn on a map that
>>connects the different localities pr distribution areas of a particular
>>taxon or group of taxa. The simplest way to construct such a graph is to
>>form a minimal spanning tree" (p. 20). Not a word about phylogeny there.
>As the sentence states - the simplest way is to form a track is to use a
>minimal spanning tree. However, to draw a track one must have some kind of
>systematic (biological) hypothesis. I agree with Hovenkamp in that it would
>be interesting in a future panbiogeographic book to include more examples
>of track construction with details of the phylogeny also presented. This
>approach is present in some of the panbiogeographic literature so anyone
>more interested in this aspect can access those examples for now.

Can you give some specific references? (this also on behalf of some of the
readers of this list who have graciously expressed an interest in this

>>>>One would have to check the literature from the '70's to find out in how
>>>>far vicariance biogeography actually derived from or was inspired by
>>>>Croizat. My statement is based on rereading Hull (1988).
>>>There is an important distinction I believe between derivation and
>>>inspiration. I would
>>>agree that Croizat may have inspired vicariance biogeography, but
>>>derivation is not substantiated.
>>I agree with the semantics - that is why I suggested that a literature
>>study would be necessary to decide whether "derived from" or "inspired by"
>>is the best term to apply.
>Given that the difference is in concept of biogeographic homology
>(Croizat's being spatial, vicariance cladistics being biological) it is
>not, in my mind, simply a matter of semantics.

Sorry - I  was were dealing with the question whether "derived from" or
"inspired by" applies best - so the different concepts Grehan mentions are
irrelevant here.

>>But the point I was trying to make still stands: that Croizat's main
>>purpose, to establish biogeography as an independent (independent from
>>geology) science, is better served by vicariance biogeography than by CGH
>Any my view of this is that vicariance (cladistic) biogeography lacks a
>spatial criterion for biogeographic relationship.

I do not see the relevance of this remark. Any "spatial criterion" lacking
or not in vicariance biogeography is irrelevant. Croizat's main purpose was
independence of biogeography, and CGH Panbiogeography lacks an independent
spatial criterion:
1. Baselines are not independent of geology (admitted by Grehan, see above)
2. Spatial homology in CGH Panbiogeography is "designated" by the
"baseline" (CGH, p. 148)
Thus Spatial homology is not independent from geology. Q.E.D.

>>In earlier posts I already noted that throughout the entire book,
>>demonstrations of this ability of CGH Panbiogeography are virtually absent.
>I agree. The book focused on spatial relationships and homologies for given

Making John Grehan's claim that "Panbiogeography can incorporate as much
detail that resolved (biological) phylogenies provide in track
construction" a rather unsubstantiated one.

>Comparisons between detailed biological relationships and spatial
>relationships for specific could prove to be useful for a future book. Of
>course one may also take the converse, and see vicariance (cladistics)
>biogeographers incorporating spatial homologies into their analyses.

And thus relinquishing the hard-fought independence from geology?

P. Hovenkamp
Nationaal Herbarium Nederland - Leiden
PO Box 9514
2300 RA  Leiden
The Netherlands
hovenkamp at

More information about the Taxacom mailing list