Stability of stage-structured population models

Fred Brauer,Ma Zhien
DOI: https://doi.org/10.1016/0022-247X(87)90041-2
IF: 1.417
1987-01-01
Journal of Mathematical Analysis and Applications
Abstract:In order to model populations which go through distinct stages, such as laboratory insect populations it is necessary to take the age structure of the population into account. The oscillations which have frequently been observed in the sizes of such populations cannot be explained by a model which ignores the effects of age structure. While it is feasible to use a fully age-structured model of McKendrickkvon Foerster type [3] or an analogous size-structured model [4], it is generally impossible to derive qualitative information about the behavior of solutions of such models. Accordingly, models have been proposed which subdivide the lifetime of an individual into two stages (larvae and adults), with different dynamics in each stage. Such models may be formulated as systems of differentialdifference equations, and several different types of systems of differentialdifference equations to model different types of larval competition have been proposed in [S]. Our purpose is to analyze the stability properties of the various models proposed in [S], using the standard method of determining equilibria, linearizing about an equilibrium, and locating the roots of the corresponding characteristic equation. By an equilibrium of a system of differential-difference equations we mean simply a constant solution; thus equilibria are found by setting the right side of each equation equal to zero (assuming that each equation of the system has been solved for the derivatives of the unknown functions). In cases where every constant is a solution of one of the equations in a system, it is necessary to replace that equation by an integrated form which incorporates the initial data in order to solve for equilibria. The linearization of a system about an equilibrium is obtained by using the deviations of the unknown functions from equilibrium as new variables, expanding by Taylor’s theorem, and retaining only the linear terms. The
What problem does this paper attempt to address?