Dynamics of Evolutionary Processes

ABSTRACT: Starting with a brief historical introduction, we review recent results in the theory of evolutionary dynamics, emphasizing new mathematical and simulation methods that promise to provide experimental access to evolutionary phenomena. One impetus for these developments comes from a fresh look at evolutionary population dynamics from the computer science perspectives of stochastic search and adaptive computation. These applications of evolutionary processes are complemented by new theoretical approaches to Wright's concept of adaptive landscapes—analyses that borrow heavily from the theories of phase transitions, critical phenomena, and self-organization in statistical physics and of bifurcations, pattern formation, and chaos in dynamical systems theory. We also discuss extensions of basic neo-Darwinian dynamics that include,for example, morphological and functional aspects of phenotypes, as well as cooperative interactions between individuals and between species. We argue that understanding the behavioral and structural richness these extensions engender requires new levels of mathematical and theoretical inventiveness that are appropriate for the massive amounts of data soon to be produced by automated experimental evolutionary systems.