[Taxacom] RE dispersal

[Jason Mate]

Dear Michael,

It is self-evident that distribution can, to a certain degree, reflect phylogeny. When species diverge, the split presuposes an initial connection between populations which implies a degree of geographical coincidence (although I concede that when extinction works as the speciation wedge this may not be apparent). In this picture dispersal acts as noise, eventually scrambling "the" signal or adding paradoxes that cannot be explained but vicariance. It may be argued that noise is useless for our purposes but its existence is undeniable. Furthermore, can it be categorically asserted that dispersal is truly random? I can presume, as a purely mental experiment, that particular dispersal routes from point A to point B, can have occured repeatedly amongst several groups, simply due to a facilitating feature that is no longer present (wind or sea currents for example). 
This pattern, due to its repetition, can mislead us into thinking that it is the result of vicariance.  Hence, support for biogeographical hypotheses requires outside evidence, be it phylogenies, fossils, molecular calibrations (as imperfect as they may be), etc. Phylogenies are built to "classify" organisms attributes, and distribution but one of them.  The biogeographical pattern itself is being tested by the phylogeny but if the distribution reflects the phylogeny then even better, for we have congruence, which is the ultimate test.
An interesting parallel is host-parasite coevolution. Although phylogenies show remarkable congruence (there are exceptions) there are always cases of, often dramatic, host-switching. Given enough time and iterations the improbable (chance dispersal/switching) will happen. It is unlikely that it will happen repeatedly in the same taxon or clade as the probabilities are low, but given time and numbers of organisms, it is a mathematical certainty. Of course the probabilities are so low that we can only hope to "see" it in the incongruences between phylogenies and and distributions. Sitting in the beach waiting for a raft of colonisers would be a waste of time... or may be not ;)
Finally, as novel and groundbreaking as vicariance may seem, it has been flogged for a long time. It is de facto assumption for organisms distribution, particularly in the Gonwanan landmasses. Editors will favour the unusual (man biting dog principle if you will) so articles supporting dispersal may well be over represented in journals. In a way noise is more interesting!
Lets end with a closing remark you made:
"Why/how does the molecular variation show such unexpected, deep geographic structure? What does this tell us about evolution?"
The wording tells me more about people than nature. If you build a phylogeny of, lets say, Acacia trees and it reflects its distribution, what are you going to write: "And here we have yet another Gondwanan pattern..." or instead will you take some creative license to make it appear more contentious? Writing factual article is not newsworthy and even in the sacred field of science self-promotion is common. It definitely helps with next years?grants.
Best regards
Jason