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

[Jason Mate]

Dear John,

in response to your last email:

"I would say that molecular or DNA sequence tree of life would be better as the relationships are presumably based only on DNA sequence similarities. The temporal dimension is given to the tree, but only secondarily, and if the authors explicitly stated that these were minimal estimates I would have no problem with that. If the divergence estimates use fossils for calibration then all dating is based on fossils directly or indirectly."

I would like to see how you would apply a temporal dimension a priori.

"As for 'probability' of a much older origin than the youngest fossil, one may believe anything at all as to what is an 'educated guesstimate" , but that's all it is and its not based on anything empirical. They are not weaving a story based on the data at hand at all, but an interpretation of that data which usually amounts to accepting origins at or about that time and precluding the possibility of a much older origin.

It may be in any case that the oldest fossil does date to the origin of a group, but unless there is any other evidence to direct to the matter, one may be limited as to what may be said. But when other evidence is available it changes the picture entirely. That is where tectonic correlation comes in (this is a panbiogeographic technique). If a tectonic correlation points to an older origin, then one now has an empirical basis to predict that the oldest fossil is an underestimate. But the tectonic correlation is also itself only a minimal estimate (for example that the ancestor existed by that time). So panbiogeography can indeed sometimes take out the guesstimating and provide a more empirical approach to divergence estimation."

You bring panbiogeography to the mix at your own risk. The risk being that you will start subordinating the facts to the pattern. Now if I had two sister clades, as in the rodent article, one in South America and one in Africa, it could be vicariance or it could be dispersal. The key is timing. We are all in agreement that fossils provide a minimal age but there is (or should) also a reasonable maximal age (best estimated by the fossil record of related groups). Of course with no bottom it becomes impossible (-ish) to refute vicariance. And that is my point with Michael´s article.


"My reading of biogeographic papers is that most practitioners do indeed assume that the lack of evidence for former land is evidence for a lack of former land."

In my experience there are not many 4m tall humans. In fact I have never met a single one and neither has anybody else I know. Based on this I make the inference that there are no 4m tall humans. This is a logical fallacy, as has been pointed out by Michael. However, and in spite of Popperian flogging, science operates not on certainties but on probabilities based on the available data. So although I cannot prove the non-existance of 4m tall humans, the fact that there is not one shred of evidence til now makes it highly improbable that one exists. However if one were to walk throught the door right now then all my pretty hypothesis would be wrong. I can live with this uncertainty. 

"As for 'fitting the evidence' this could mean anything at all. Not fitting the 'evidence' is another possibility. This happened with the Galapagos where Croizat not only predicted the existence of former island arcs (the modern term for his insular clusters)before geological evidence was available, he also predicted the associated tectonic center, again before the evidence (through seafloor mapping) was available, all through reference to organisms that everyone else supposed dispersal from the mainland."

A broken clock tells the right time twice a day. In 10,000 pages I am sure that there are many right guesses.

 
"The irony of all of this is that the various papers by Heads demonstrates the utility of molecular phylogenies as hypotheses of relationship for many, many groups. While I have my doubts about molecular phylogenies and regard them as provisional with respect to morphological corroboration and problematic when there is substantial morphological incongruence, they do provide many examples that are biogeographically coherent, that often demonstrate classic vicariism, corroborate the standard tracks of the world (which no one has yet falsified), and providing tectonic correlations. One could not ask for more."

You are moving way beyond the topic here. But in any case tracks are not a method of analysis. You are describing a pattern and then trying to impose it on other patterns. Humans have this tendency to learn a pattern and then look for it. Like finding animals in the clouds or in a piece of wood, just because it looks like something it doesn´t mean it is the same thing.

Good night

Jason