Abundance-Distributions in Artificial Life and Stochastic Models: "Age and Area" revisited

Chris Adami, C. Titus Brown, and Michael R. Haggerty

In Proc. of 3rd Europ. Conf. on Artificial Life, June 4-6, 1995, Granada, Spain, Lecture Notes in Computer Science, Springer Verlag (1995), p.503

Using an artificial system of self-replicating strings, we show a correlation between the age of a genotype and its abundance that reflects a punctuated rather than gradual picture of evolution, as suggested long ago by Willis. In support of this correlation, we measure genotype abundance distributions and find universal coefficients. Finally, we propose a simple stochastic model which describes the dynamics of equilibrium periods and which correctly predicts most of the observed distributions.


Click here to download the (gzipped) postscript file