[Taxacom] Hedges /Kumar (eds) The Timetree of Life

[John Grehan]

-----Original Message-----
From: taxacom-bounces at mailman.nhm.ku.edu [mailto:taxacom-bounces at mailman.nhm.ku.edu] On Behalf Of Jason Mate


Dear John,

> All data can be "problematic" in its interpretation. 

Sure

> There are uncertainties in the timing, the method, etc. I fail to see why 
> you specify fossils. It could also be argued, correctly, that tectonic 
> data can be problematic.

Sure, but fossils are made more problematic when they, or derivative estimates (molecular) are misrepresented as not being minimal.

> As regards to extant taxa morphology resulting in the wrong assignment of 
> fossil taxa, who ever said it was error free? This is a source of error in 
> timing, like any other and open to reinterpretation by researchers armed 
> with better data or methods. 

I think here was the irony of molecular theorists on one hand condemning morphology and on the other using it as the ultimate authority for molecular divergence.

> All these points notwithstanding, the right fossil at the right location 
> can destroy an argument, even if this argument was the most parsimonious 
> explanation for the previously available data. It doesn´t matter who or 
> how a hypothesis was developed, if it doesn´t fit the facts (fossils, 
> geology) then it is wrong. 

In principle, but in practice I can certainly see how this would apply to the misapplication of fossils as minimal when all it takes is an older fossil as refutation for a notion that should not have been made in the first place.

>> "Phylogeny gives biological relationships. It cannot give spatial 
>> relationships. 
 
> The phylogeny provides the structure on to which organise the 
> distributional data. 

Yes, but only the BIOLOGICAL structure. It does not provide the spatial structure. One must have a biogeographic criterion for that - which is what panbiogeography provides.

> Also saying that one starts from the lowest level (i.e. genus) hardly 
> matters if your genus is something more speciose than hominids 
> (Onthophagus for example?).

Perhaps so, but there are plenty of examples with more speciose groups.

"True, one does not correlate tracks with geological events since the latter are historical theories. One must first correlate tracks with tectonics and from that one has a basis for suggesting the track is linked to the history of those tectonics. A simple example is where a taxon is found either side of a transform fault, but disjunct from each other at distances that match past fault movement. This correlation would suggest the ancestor originated at a time when the fault positions the disjuncts together. When this is found for many taxa of various ecologies the conclusion of a historical relationship becomes even more apparent. One could deny it of course. .... The point is that they can lead to the prediction of geological features or events that were not already apparent to geologists."

> Maybe you didn´t mean to say it this way, but what I am reading here 
> suggests that if you have lots and lots (how many?) 

Look at the NZ Apline fault example. Enough to justify 'lots' although I have never counted. Perhaps you could do that.

of tracks joining two 
> areas then you can somehow justify that a geological event occurred. 

When tracks overlap or cross a tectonic feature one may hypothesize that the track and the tectonic feature have a shared history


If 
> so, is vicariance disproved if we look and not find any features or events? Or better, if the actual phylogenies of the organisms, under closer scrutiny, do not match the tracks? 


> To make it clearer, lets have a taxon (tribe P) with 2 genera, one in 
> South America (Genus Sam, 3 species) and one in Africa (Genus Afri, also 3 
> species). If I ignored the phylogeny to the extent that you seem to 
> suggest then I would have two groups linked by a minimum spanning tree 
> which fits other such patterns ((Sam1,Sam2,Sam3)(Afri1,Afri2,Afri3)) thus 
> suggesting that the tribe P shows a typical vicariant distribution. 
> However if I were to look at the phylogeny of the group and find the 
> relationships to be ((Sam1,Sam2)Afri3)(Sam3(Afri2,Afri1)) would your story 
> be so neat? I´d guess you would need to time the divergences to know which 
> (or if any) of the 2 splits is the result of vicariance.

Nice analysis. Yes, the patterns of track connections between South America and Africa can be complex, but the decision to even identify them as 'trans-Atlantic' is a purely panbiogeographic technique based on spatial proximity to a tectonic formation. So you were applyng a panbiogeographic principle and you did just fine with that. I will give you an 'A' for that effort.

> Finally, in regards to my silly comments. They are an extension of your 
> system, i.e., panbiogeography can lead to silly situations. If you look at 
> the distribution of organisms armed with just the basic knowledge of their 
> relations (i.e. monophyletic tribes, genera, etc) then you can, at best 
> say that either the lineage broke apart (vicariance) or they somehow got 
> to the other side. If you want to know which you need to know how the taxa 
> relate to each other precisely  and time their divergences as best as 
> possible. 

The more phylogenetic detail available the more detailed the analysis. Nothing silly about that. But don't forget that it is important to consider the spatial details of individual groups in relation to other groups as well.

John Grehan