Fitness Landscape

Fitness Landscape

In evolutionary biology, fitness landscapes or adaptive landscapes are used to visualize the relationship between genotypes (or phenotypes) and reproductive success. It is assumed that every genotype has a well-defined replication rate (often referred to as fitness). This fitness is the "height" of the landscape. Genotypes which are very similar are said to be "close" to each other, while those that are very different are "far" from each other.

The two concepts of height and distance are sufficient to form the concept of a "landscape". The set of all possible genotypes, their degree of similarity, and their related fitness values is then called a fitness landscape. The idea of a fitness landscape helps explain flawed forms in evolution, including exploits and glitches in animals like their reactions to supernormal stimuli.

In evolutionary optimization problems, fitness landscapes are evaluations of a fitness function for all candidate solutions (see below). The idea of studying evolution by visualizing the distribution of fitness values as a kind of landscape was first introduced by Sewall Wright in 1932.