NEW DIRECTIONS IN BIOGEOGRAPHY

The deduction of biogeographic process from the study of its history i s complicated by the fact that history is a singular thing. This singu larity allows no estimation of the degree of determinism surrounding t he realization of biogeographic processes, and consequently we know al most nothing about the generality of such deductions. Here we discuss a new approach, called 'experimental biogeography', that exploits comp uters to model faunal build-up repeatedly against a fixed vicariant ba ckground over ecological and evolutionary time scales. This new approa ch enables a biogeographer to be omniscient - to know both vicariant h istory and actual phylogeny. Moreover, history can be replayed repeate dly to accumulate a sample of multiple phylogenies and to estimate pro bability density functions for biogeographic variables. Roles of stoch astic, historical and ecological processes in adaptive radiations can also be assessed. Experimental biogeography allows examination of the reliability of various methods of recovering historical patterns.