Optimal timing in biological processes

American Naturalist
By:  and 



A general approach for obtaining solutions to a class of biological optimization problems is provided. The general problem is one of determining the appropriate time to take some action, when the action can be taken only once during some finite time frame. The approach can also be extended to cover a number of other problems involving animal choice (e.g., mate selection, habitat selection). Returns (assumed to index fitness) are treated as random variables with time-specific distributions, and can be either observable or unobservable at the time action is taken. In the case of unobservable returns, the organism is assumed to base decisions on some ancillary variable that is associated with returns. Optimal policies are derived for both situations and their properties are discussed. Various extensions are also considered, including objective functions based on functions of returns other than the mean, nonmonotonic relationships between the observable variable and returns; possible death of the organism before action is taken; and discounting of future returns. A general feature of the optimal solutions for many of these problems is that an organism should be very selective (i.e., should act only when returns or expected returns are relatively high) at the beginning of the time frame and should become less and less selective as time progresses. An example of the application of optimal timing to a problem involving the timing of bird migration is discussed, and a number of other examples for which the approach is applicable are described.
Publication type Article
Publication Subtype Journal Article
Title Optimal timing in biological processes
Series title American Naturalist
Volume 123
Issue 1
Year Published 1984
Language English
Contributing office(s) Patuxent Wildlife Research Center
Description 1-19
Larger Work Type Article
Larger Work Subtype Journal Article
Larger Work Title American Naturalist
First page 1
Last page 19
Google Analytic Metrics Metrics page
Additional publication details