display | more...

In evolutionary biology, refers to a topological concept of fitness as a function of an organism's phenotype or genotype. The origin of this concept is usually attributed to Sewall Wright, although it is a ubiquitous metaphor across disciplines. If he could not be properly called the father of the adaptive landscape, he must certainly be one of its popularisers in evolutionary biology. G.G. Simpson was also involved in promoting the landscape concept, particularly for phenotypic evolution over geological time. In its most intuitive form, the landscape consists of a two-dimensional plane that represents individual phenotype or genotype as a continuous space. Above this plane, one may draw some surface with "peaks" and "valleys". The perpendicular height of the surface above the plane represents the fitness attributable to a particular type.

One appealing aspect of this concept is that the dynamics of a population on this landscape can be readily converted into mathematical formulae. The abstract concept of adaptation and divergence can thus be formalised. However, it is easier to apply the model to phenotypic than genotypic evolution, as the former translates readily to a continuous space. Genotype space has been described as an aggregate of points, each corresponding to some combination of alleles. It has also been described as a collection of allele frequencies, though a frequency is an attribute of a population and consequently violates the nature of the landscape model. For a detailed and historical discussion of this, see Sewall Wright and Evolutionary Biology by WB Provine, ©1986 University of Chicago Press.

The recent literature has experienced renewed controversy over the importance of the adaptive landscape as an accurate model of evolution (Whitlock and Phillips 2000). S. Gavrilets pointed out that although we are restricted to visualising the landscape in three dimensions, no such limitation exists for biology. Thus, the notion of peaks and valleys separating fit types is flawed. Furthermore, fitness is a function of the environment as well as phenotype/genotype. Fluctuations in the environment may be large enough to render inaccurate a static landscape metaphor.

Despite these controversies, the adaptive landscape undeniably remains one of the most influential and inspirational concepts in evolutionary biology.

One of the primary problems with the adaptive landscape concept is that only three variables can be graphed - two alleles and fitness. This ignores linkage disequilibrium, which will be another factor in determining the way in which the population will evolve. Another problem is that it assumes that mating is random, which is often not the case.

Overall, the adaptive landscape is useful for telling us how fit a population is at a given time point corresponding to a given allelic combination. It's pretty useless when it comes to telling us what the population is going to do next. It's a shame, really, because it's not possible to work out what a population is going to do without using iterative models and continuing until equilibrium is reached. Being able to start from a point on a landscape and simply draw a line to the top of the nearest peak would make life significantly easier.

(Of course, there's the other problem that given a constant selection pressure populations will not necessarily reach the point of maximum fitness given the alleles available to them. But that's another issue.)

Log in or register to write something here or to contact authors.