[Taxacom] Dispersal clarifications

[Jason Mate]

Hi Michael, my thoughts on your comments below:
MH:About the terms you mention: 'dispersalist theory' and its advocates accept chance dispersal as a mode of speciation. 
JFM:So is dispersal of a species from a mainland source to an island acceptable dispersal?
MH:An organism 'disperses across a barrier', often by mysterious means. For example, an organism that moves a few centimeters in its lifetime is theorized to have jumped - just once - 10 000 km across the Pacific. Vicariance theory accepts that allopatric speciation is caused by vicariance and that range expansion (physical movement) only explains overlap.
JFM:Are all purported cases of dispersal similarly extreme? Do you consider the dispersal of flying organisms equally mysterious? Don´t you think that the probability of dispersal correlates with the organisms ability to disperse?
MH:Chance: this term can be used in the sense of sweeping something under the carpet - 'oh, event x must have just been an accident'. Alternatively, you can go out and collect data and actually work out the 'probabilities'. These are also referred to as 'chance' - as in 'the chance of event x happening', but 'chance' here is a completely different concept. For example, a road accident may have happened on a corner where it turns out, following a statistical analysis, 'accidents' regularly happen. It wasn't simply an accident, there was a general, underlying explanation. This change in approach was the psychological revolution of Fermat and Pascal that gave birth to probability theory. 
JFM:This is a qualitative remark. A 1 in a million event, although rare, although ´chance´, still has a probability, and the sum of all the probabilities describes and event. Thus, although accidents can cluster and these are what people may find interesting, a well lit corner in a speed-controlled zone with good visibility can also have the odd accident. The underlying cause of this accident may be something completely random, with a low repeatability (maybe old drivers with astigmatism at dusk fail to see pedestrians). But this isolated case is still part of the totality of accidents for the area under study. So I´d say that you are confusing probability with repeatability. By studying the biota of an area you work out the frequency (probability) of each mechanism. Maybe dispersal is not of interest to you since you are only interested in vicariance, but to taxonomists or ecologists dispersal can be equally interesting.
MH:Dispersal theory as seen in biogeography and systematics journals deals with phylogeny and relies on 'chance' in the earlier sense. In contrast, dispersal studies in ecological journals deal with observations (not inferences) of simple movement and are full of probability analyses - they use chance in the modern sense.
JFM:Chance in the earlier sense? You mean the magical sense? Each throw of the dice is independent from the previous ones. So even if I could work out the probability of dispersal under your system (I can´t because your philosophical frame does not allow it) it would be based on studying the origin of a certain number of clades and 

MH:'Ecological dispersal' is just the simple movement of organisms observed every day. A weed disperses into a garden, but it doesn't differentiate into a new taxon. An organism can be where it is because it moved there or because it evolved there.
JFM:The simple movement observed every day? So if you don´t observe it every day it does not occur, i.e. migration? What about vagrant specimens? Dispersal is not an everyday, week or even year event. It is rare, precisely because it is a high failure, overwhelmingly lethal, no second prize strategy.
 MH:Neatly allopatric (dovetailing) geographic structure repeated in different groups can't be explained by (mere) 'chance'. It could be the result of a general process such as vicariance.
JFM:You are the only one here saying that vicariance is not an accepted mechanism that accounts for organisms distributions. At the same time, you would expect that sources of colonists would be limited to certain areas that have some sort of advantage (i.e. currents, prevailing winds, shortest distance).
MH:Vicariance might be falsified by finding that the distributions of taxa in a region don't share similar phylogenetic/biogeographic breaks. But this could be due to original allopatry followed by range expansion. Unlike the text-book examples which everyone learns, most broad theories don't fall over instantaneously because someone finds a single fact (see the detailed critiques of 'naive falsificationism' by Lakatos, Feyerabend etc.). Usually there is a slow accumulation of 'anomalous' data and eventually there is a broad shift of opinion. If all the molecular phylogenies showed no geographic structure and instead taxa tended to sort on colour or size or hairiness, rather than locality, no-one would be looking at  vicariance and it would just fade away. 
JFM:Once again, vicariance is an accepted mechanism.
MH:But the clades nearly all show impressive allopatry - this is why so many authors are  putting distribution maps of their clades in the graphical abstracts at Mol. Phylogen. Evol. and lots of people are discussing vicariance. If you Google scholar 'vicariance', here are the numbers of articles mentioning the word for 5-year periods, starting with 1960-1965 when Croizat wrote his big books on the topic: 27, 40, 66, 218, 770, 1040, 1380, 2280, 4140, 6600 (2005-2010). 
JFM:So vicariance is the only mechanism allowable unless everything there shows otherwise? Why would you assume that the entire biota of an area got there by one mechanism only? What is the logical basis for this?