Barrier transitions driven by fluctuations, with applications to ecology and evolution (Q1324992): Difference between revisions
From MaRDI portal
Added link to MaRDI item. |
Set profile property. |
||
Property / MaRDI profile type | |||
Property / MaRDI profile type: MaRDI publication profile / rank | |||
Normal rank |
Revision as of 03:56, 5 March 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Barrier transitions driven by fluctuations, with applications to ecology and evolution |
scientific article |
Statements
Barrier transitions driven by fluctuations, with applications to ecology and evolution (English)
0 references
7 July 1994
0 references
\textit{T. Price} et al. [Peak shifts produced by correlated response to selection. Paper submitted for publication (1992)] recently proposed a model for the evolution of correlated characters in a changing environment. According to this model, we envision two correlated traits, \(X\) and \(Y\), which are near one of two peaks in the fitness surface of individual fitness. After an environmental change, \(X\) is subject to directional selection. As a correlated response, \(Y\) may be driven from one fitness peak to the other. When genetic drift is taken into account, the stochastic fluctuations may produce a peak shift when it is not predicted in the deterministic model, or may prevent it when it would occur deterministically. (Crossing this valley of fitness is equivalent to crossing the deterministic separatrix in the ecological setting.) We wish to compute the probability that fluctuations cause a shift from one outcome to another. This involves (i) specifying an underlying model, (ii) deriving the equations for the probability of outcome, given initial conditions, and (iii) solving those equations. This program is carried out in the paper. Models based on stochastic differential equations, stochastic difference equations, and a Markov chain are introduced. In addition a ``practice'' problem is solved. The solution of this problem, which involves only one state variable, indicates the form of the general solution. The method of general solution is summarized and explained. Two simple examples of the method are illustrated, and the method is extended to models in which there is no underlying deterministic system. A thorough study of the results illustrates why the diffusion approximation is appropriate in this case, and how macroscopic population dynamics arise from an underlying Markov chain model.
0 references
evolution of correlated characters
0 references
changing environment
0 references
fitness surface
0 references
directional selection
0 references
genetic drift
0 references
stochastic fluctuations
0 references
stochastic difference equations
0 references
diffusion approximation
0 references