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

[Jason Mate]

Dear John,

when you mentioned the secondary nature of time in cladistic biogeography I was somehow confused but I think I understand what you mean. You seem to take issue with the creation of cladograms which are afterwards calibrated, usually with fossils. As fossils are the only truly objective datum you have however, it is the only way to time, objectively, a divergence. Afterwards you can check to see if it correlates with known geological events, suggesting vicariance, or not in which case you would lean towards dispersal. 

As for the root timing, it can be a serious issue as Sergio has pointed out, but only if you cannot directly calibrate the basal node to fossil evidence. This is because the probability range from a fossil and the node it calibrates is effectively 0 up the tree(or close depending on the time uncertainty of the stratum) whereas backwards you have the compounded errors of the fossil itself and all the internodal branch lengths. The further back you move away from the fossil points in the tree the worse the range becomes. In a graphic way fossils pull the nodes above them towards themselves.
However this is a very different issue to the maximal age of fossils which, I reiterate, can be tackled by reference to other fossils, if they exist, and the phylogenetic relationships of your target group to the outside fossils.

In regards to the use of tectonic events, I see them more as explanatory events than as supporting evidence. A divergence and a geological event can coincide and you can make a case for it but little else. This is the way that panbiogeography (sensu Croizat or Craw)works: patterns of distribution, superimposed on others without phylogenetic information (and therefore not with no check on the ´homology´ of the distributions). You can draw attention to a pattern but not test it. And if you want an example I offer the ongoing discussion on Madagascar´s mammalian fauna.


Good night

Jason