Abstract:In this paper we propose some Harris-like criteria in order to study the long time behavior of general positive and periodic semiflows. These criteria allow us to obtain new existence results of principal eigenelements, and their exponential attractiveness. We present applications to two biological models in a space-time varying environment: a non local selection-mutation equation and a growth-fragmentation equation. The particularity of this article is to study some inhomogeneous problems that are periodic in time, as it appears for instance when the environment changes, due for instance to the seasonal cycle or circadian rhythms.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is related to the study of the long - time behavior of non - conservative periodic semi - flows, especially for non - local linear partial differential equations (PDEs). Specifically, the author proposes some methods similar to the Harris criterion to study the long - time behavior of general positive periodic semi - flows. These methods enable the author to obtain the existence result of the principal eigen - element and its exponential attractivity. The paper pays special attention to the applications of two biological models in the spatiotemporal - varying environment: the non - local selection - mutation equation and the growth - fission equation. The special features of these problems are that they are time - periodic, for example, when the environment changes due to seasonal cycles or circadian rhythms.
### Background and Problem Description of the Paper
Many biological phenomena depend on periodic time scales, such as the problem of adaptive dynamics and the relationship between circadian rhythms and cancer cells. Therefore, it is crucial to understand the evolution equations with time - periodic coefficients and their long - time behavior. For ordinary differential equations (ODEs), Floquet theory has well - established the long - time behavior of linear ODEs. However, for linear partial differential equations (PDEs), especially non - local equations, the relevant research is still very limited.
### Main Objectives
The main objective of this paper is to study the long - time behavior of non - local linear PDEs \(\partial_t u(t)=A_t u(t)\), where \(A_t\) is a \(T\)-periodic operator, that is, \(A_{t + T}=A_t\) holds for all \(t\geq0\). More specifically, the author hopes to find some conditions such that the solution \(u(t)\) exhibits asynchronous Malthusian behavior as \(t\rightarrow\infty\), that is:
\[u(t)\sim e^{\lambda_F(t - s)}\gamma_t\langle h_s, u_s\rangle\]
where \(\lambda_F\) is the so - called Floquet eigenvalue, and \((\gamma_t)_{t\geq0}\) and \((h_t)_{t\geq0}\) are the direct and dual periodic eigen - elements respectively.
### Research Methods
To achieve this goal, the author draws on techniques in probability theory, especially the Harris method. The Harris method is a contraction method that can be used to deal with problems in non - homogeneous environments and can quantify the convergence rate of periodic Malthusian behavior. The author proposes two sets of assumptions, which are applicable to the selection - mutation model and the growth - fission model respectively, and obtains the corresponding convergence results through these sets of assumptions.
### Application Examples
1. **Non - local Selection - Mutation Equation**:
\[
\begin{cases}
\partial_t u(t,x)+\partial_x u(t,x)=\int_R u(t,y)Q(y,dx)dy + a(t,x)u(t,x),&(t,x)\in(s,\infty)\times\mathbb{R}\\
u(s,x)=u_s(x).
\end{cases}
\]
This equation describes the dynamics of a population structured by the phenotypic trait \(x\in\mathbb{R}\). \(a(t,x)\) represents the birth and death balance of the trait \(x\) at time \(t\), and \(Q(y,dx)\) describes the mutation process from the trait \(y\) to the trait \(x\).
2. **Growth - Fission Equation**:
\[
\partial_t u(t,x)+\partial_x(\tau(t,x)u(t,x))+\beta(t,x)u(t,x)=\int_x^\infty b(t,y,x)u(t,y)dy.
\]
This equation describes the process in which each individual in a population experiences growth and fission. Each individual grows at a speed of \(\tau(t,x)\) and divides at a rate of \(\beta(t,x)\), and the fission kernel \(b(t,x,y)\) describes the change of traits during the fission process.