Biogeographic units and classification

[John Grehan]

My comments here are written in response to specific points. It is posted
to the list on the assumption that the issue of classification in biogeography
is at least of general interest to some of the list members. If any list member
feels that the dialogue here is not of general interest I please let me know and
I may reconsider.

I appreciate the comments by the following person and others on the subject.


De Jong wrote

>Possibly I misinterpret John Grehan's discussion, but does not a baseline
>connect two areas of endemism that are connected by phylogenetic
>relationships of their inhabitants?

The baseline does not connect areas. Localities or distribution areas are
connected by a track. The distribution areas are themselves just an arbitrary
grouping of localities, and are used as a shorthand for the distribution
where scale constrains the detail (i.e. disjunctions are so small they would
not show on, for example, a global map. The baseline is assinged to a track
according the the geological/geomorphological feature crossed by the track.

> Is
>not a baseline a hypothesized biogeographical province or whatever in
>another geological period? Why would such an area in, say, the Cretaceous be
>more natural than nowadays?

My understanding of baselines is that they do not represent former
biogeographical provinces. They do represent biogeographic homologies, and
in that context they represent biogeographic units comprising
relationships rather than geographically set areas (thus a Pacific track is
defined by the track crossing the Pacific, not by a particular line enclosing
the Pacific).


 If we agree that biogeographic units are areas
>of endemism (i.e., are characterized by endemic species), then they are
>natural in the sense that they are characterized by natural (not man-made)
>phenomena, notwithstanding the fact we may disagree about the limits of the
>area.

If we can disagree about the limits of the area it seems to me that this is
a prime indication of their being artifacts. The view that areas of endemism
are artificial may represent one of the substantial differences in
classification theory between panbiogeography and all other biogeographic
methods.

As far as I am able to determine, there is nothing in a distribution that
provides
empirical evidence of that distribution comprising an "area" representing a
biogeographic unit.

>Apparently, the problem arises from attempts to find a hierarchical order in
>a system that did not per se develop hierarchically.

Agreed, and in panbiogeography you will find that in place of a
hierarchical order
there is a network.

Even in cases where we
>can discern a hierarchical development of physical areas (sch as the
>break-up of Gondwana, and not counting the fact that the pieces partly
>became inbedded in or connected to areas wtih another origin), it remains to
>be seen if distribution areas agree with that, and even if they do, why
>neglect all the distribution areas that don't?

Agreed

>I fully agree with Peter Hovenkamp's comment that biogeography is a
>discipline of its own, and that we should apply methods suited to
>biogeographic problems. If in panbiogeography the principle of synapomorphy
>has been established as a matter of methodological principle, it may well be
>time to reconsider this principle.

I agree that biogeography is a discipline of its own, but most biogeographic
account make biogeography subservient to another discipline such as
biological systematics (where biological relationships are represented as
biogeographic relationships), or historical geology where biogeographic
reconstructions (whether cladistic or not) are made according to their
fit with the geological history (e.g. all organisms dispersed to Galapagos
over water because geologists say there is no direct geological connection
between Galapagos and the American continent).

It is possible that the use of spatial synaomorphies in panbiogeography is
innappropriate in some way, but to my knowledge such a view has not
been substantiated.

John Grehan