Tsunamis and the Colonization of Islands

The story I have always heard about how islands are colonized by terrestrial animals like lizards, turtles, and rodents has never made sense; even if a storm washes one or two trees out to sea, how likely is it that a tree will have a surviving male and female of the same species, or that two trees, on with a male and one with a female, will wash ashore on the same island?  It certainly seems like it would be outrageously rare, especially when coupled with the fact that if the female arrives first, the male must follow before she dies, or vice versa.  Of course, I had in mind a thunderstorm, or even a storm front, but I don't think even a hurricane or typhoon would make much difference. 

A tsunami would be a different story, though.  As we have seen all too clearly, a big one would affect a much broader region of coast, simultaneously washing out to sea countless trees, animals, whatever.  An animal would not need to be in the tree to start with; it might well find itself swept off shore and only later find a "raft" floating nearby.  On the other hand, if a given species of lizard or rat favored a particular type of tree, animals of that species might find themselves arriving on the same island at nearly the same time, just as the action of wind and wave has coordinated the arrival of oyster buoys and refrigerators from Japan.

I'm not sure how this could be tested.  Maybe a correlation could be found between coasts that are unusually subject to large tsunamis -- places like Japan, Indonesia, and Peru -- with the origins of colonizing small animals.