STOCHASTIC-MODELS OF PHYLOGENY AND EVOLUTION OF DIVERSITY - DISCUSSION

Raup et al. (1973) presented an exercise in which family trees are simulated in such a manner that branching and extinction occurs randomly through time once an arbitrarily defined equilibrium number of taxa have evolved. Boucot points out that if they are correct, the fossil record should reveal an overall uniform rate of cladogenesis. Cladogenesis actually occurs in bursts which correlate highly with marked changes in the physical environment that open up previously vacant niches or give rise to reproductive isolation. Boucot states that the Raup et al. (1973) computer generated simulation accounts for an insufficient portion of the fossil record. Raup et al. reply that the model was used as a null hypothesis against which to view the real world. The simulations show what the record might look like if cladogenesis and related phenomena behaved as random variables. A uniform rate of cladogenesis is predicted only in the sense that if a rate is stochastically uniform there is likely to be considerable variation about some mean value but the variation is not correlated with time. The fact that specific causes can sometimes be identified for specific events in the geologic past in no way invalidates the model.