CPM Seminar

Statistical Physics and Biological Coevolution

Per Arne Rikvold

Florida State University

Biological evolution presents many problems involving systems that contain a large number of interacting entities (e.g., individuals or species) that interact through a complex network of interactions. Consequently, this research area has attracted increasing attention from statistical physicsists.

In this talk I will introduce a class of simple models of biological coevolution, in which the basic interacting entities are individual organisms. On short time scales up to hundreds or thousands of generations, these models describe ecological dynamics. However, mutations that occur during reproduction introduce new species that can lead to the extinction of existing ones. The result is an evolutionary dynamics on time scales of thousands to millions of generations.

The models I present are simple enough that many of their ecological features can be obtained exactly, yet they produce complex, intermittent dynamics that lead to approximate power laws for such quantitities as the lifetimes of species and communities, and to 1/f noise in their power spectra. Some of the models also create simple food webs that display features consistent with observations in Nature.