Ken Kinman comments on tracks

[John Grehan]

Ken Kinman wrote:

>John,
>     I agree with Tom that this putative clade (and its putative
>synapomorphies) should be reexamined.  If Oreobolus distichus or goeppingeri
>were found on any intervening islands (as in the fly genus Cymatopus), I
>wouldn't find the distribution so strange.

There are certainly many groups found in the Americas and eastern Asia that
are not found within the Pacific itself (e.g. the Blennospermatinae
example), but the intriguing aspect of the Oreobolus distribution is the
absence of the Hawaiian clade from the trans-Pacific distichus or
goeppingeri group.

>     Just because one cladist's analysis came up with this clade doesn't
>make it holophyletic.

Agreed. This is true of any phylogeny. If one is to analyze the
biogeography of many taxa one is in the position of examining the
biogeographic implications of phylogenetic relationships as they are
presented. The Oreobolus case provides an example where one may consider
geographic relationships as being informative about the biological
relationships. According to Platnick and Nelson, incongruence between
geographic proximity and biological proximity requires one to give up
geographic proximity at an informative character whereas one might say the
contrary may be the case.

>   I'm not saying that
>this is the problem with Oreobolus, but your biogeographic analysis
>indicates to me that this so-called clade could be based on homoplasies (and
>would make a mess of your panbiogeographic analysis).

It would not change the fact that the genus has a Pacific baseline. It is
probable that the nodes may not differ substantially. A different set of
geographic coordinates may be proposed.

>     No matter how that particular issue might turn out, I find the absence
>of Antarctica on such maps VERY disturbing.  If Oreobolus was starting to
>break up into clades during the Oligocene, Antarctica would probably be the
>most important part of the puzzle.

There is nothing I am ware of in the biogeography of Oreobolus that
suggests Antarctica would 'probably' be the most important part of the
puzzle (and I am not aware that there is a 'puzzle' as such). It would be
consistent with the Pacific orientation of Oreobolus to expect that the
genus may have once occupied Antarctica, but this is no more significant
than its occurrence anywhere else that it is known from. I would not be
surprised at any discovery of fossils between China and Europe either.

>  I would strongly suggest you plot the
>distributions of Oreobolus on an Oligocene world map (or at least back to
>the Miocene if "circularity of reasoning" has you worried).

This would not help from my point of view. It is simply using a historical
narrative from another discipline to 'explain' a biogeographic pattern.
This is what I was critiquing in Raxworthy's approach.

>      I suspect that the main tract of Oreobolus is trans-Antarctic and that
>it spread northward on both sides of the Pacific.

This is traditional Darwinian biogeography invoking dispersal away from an
imaginary center of origin using landbridge narratives in a plate tectonics
context.

>*IF*
>distichus-goeppingeri really is a clade, it is probably either an old
>slow-evolving vicariant clade (with lots of extinctions having taken place),
>or it is a recent clade with some bizarre dispersal event having taken
>place.

The standard geographic character of Oreobolus may suggest that it is not
the result of some bizarre dispersal event.

>      But I certainly wouldn't automatically assume that it really is a
>clade just because one cladist's analysis came out that way (especially if
>based on "selected" characters, as in some of Sereno's dinosaur phylogenies
>which probably make tyrannosauroids too derived among other problems).

As a biogeographer I am not in the position of second guessing every
phylogeny - whether proposed by a cladist or any other kind of systematist.
Any or all phylogenies may be wrong. What interests me is the kinds of
biogeographic pattern represented by taxa as they are proposed by
systematists. Where phylogenies are highly controversial one may be in a
position to directly contribute from a biogeographic perspective. Even
where there is no controversy, the Oreobolus example shows that geographic
information may provide indications of alternative arrangements.

>Otherwise you could be missing out on some "red flags" your panbiogeographic
>analysis might be sending you.  If it really is paraphyletic (or worse yet
>polyphyletic), ANY biogeographic analysis which assumes holophyly would be
>severely flawed.

This applies to any research discipline that works with a given phylogeny.
Any or all phylogenies may be wrong. Biogeographic analysis does not
'assume' holophyly (or any other phyly) in a vacuum. If a systematist
presents an analysis supporting a particular phylogenetic arrangement, the
biogeographer is then in the position of examining the spatial implications.

John Grehan