Studying Long-term Evolution with Artificial Life

This paper describes an artificial ecology system populated by self-replicating machines, which is able to capture many of the qualitative and quantitative characteristics observed in real ecosystems, from the fractal nature of taxonomic and phylogenetic trees to the pattern of originations and extinctions in long-term evolution. The self-replicating machines are composed by elementary particles. The composition and behaviour of a self-replicating machine is determined by its genetic code, and new species arises by means of random mutations. The resulting dynamics is characterised by the co-evolution of species competing for resources. The interplay between the species and environment plays a crucial role for self-regulation: extinctions and originations are mechanisms used by the ecosystem to maintain the dynamical equilibrium that is favourable to life itself.