Existence and stability of equilibria in age-structured population dynamics (Q759679): Difference between revisions

The dynamics of an age-structured closed population in a spatially homogeneous environment are described by the McKendrick equations \[ \partial \rho /\partial t+\partial \rho /\partial a+D(\rho)(a)\rho =0,\quad t>0,\quad 0<a<A\leq +\infty, \] \[ \rho (t,0)=\int^{A}_{0}F(\rho)(a)\rho (t,a)da,\quad t>0, \] where \(\rho =\rho (t,a)\geq 0\) is the density of individuals of age a in the population at time t, A is the maximum lifespan, and D and F are the (per unit density per unit time) death and fertility rates. Usually, D and F are functions of a and of the total population size \(P(t)=\int^{A}_{0}\rho (t,a)da.\) Here, D and F are functions of a and are allowed to be general functionals of \(\rho\), as long as they are smooth enough near \(\rho =0\). An initial condition \(\rho\) (0,a) \((0<a<A)\) must be provided. Equilibrium solutions \(\rho =\rho (a)\geq 0\) of the McKendrick equations and their stability are studied. There is always a trivial equilibrium solution \(\rho\) (a)\(\equiv 0\) (population extinction) and there may be other equilibrium positive solutions. Bifurcation behavior may be observed depending upon the value of an adequate parameter. Previous papers by the author show the existence of a primary bifurcation from the trivial solution as the intrinsic net reproductive rate \[ n=\int^{A}_{0}F(0)(a)\exp (-\int^{A}_{0}D(0)(s)ds)da \] passes through \(n=1\) and the existence of a secondary Hopf-type bifurcation to time periodic solutions for a large value of n. These global results are complemented here by a local linear analysis of the primary bifurcation. It is shown that the trivial solution looses stability as n increases through 1 while the branch solution is stable iff the bifurcation is supercritical. The lower order terms in the Lyapunov-Schmidt expansions are obtained and through them the direction of bifurcation can be related to the properties of D and F as functionals of \(\rho\). Some examples illustrate these general results.