Onset and termination of oscillation of disease spread through contaminated environment

  • Received: 29 August 2016 Accepted: 30 December 2016 Published: 01 October 2017
  • MSC : Primary: 34K18, 35B35, 35B32; Secondary: 35K57, 92D40

  • We consider a reaction diffusion equation with a delayed nonlocal nonlinearity and subject to Dirichlet boundary condition. The model equation is motivated by infection dynamics of disease spread (avian influenza, for example) through environment contamination, and the nonlinearity takes into account of distribution of limited resources for rapid and slow interventions to clean contaminated environment. We determine conditions under which an equilibrium with positive value in the interior of the domain (disease equilibrium) emerges and determine conditions under which Hope bifurcation occurs. For a fixed pair of rapid and slow response delay, we show that nonlinear oscillations can be avoided by distributing resources for both fast or slow interventions.

    Citation: Xue Zhang, Shuni Song, Jianhong Wu. Onset and termination of oscillation of disease spread through contaminated environment[J]. Mathematical Biosciences and Engineering, 2017, 14(5&6): 1515-1533. doi: 10.3934/mbe.2017079

    Related Papers:

    [1] Mingzhu Qu, Chunrui Zhang, Xingjian Wang . Analysis of dynamic properties on forest restoration-population pressure model. Mathematical Biosciences and Engineering, 2020, 17(4): 3567-3581. doi: 10.3934/mbe.2020201
    [2] Bruno Buonomo, Marianna Cerasuolo . The effect of time delay in plant--pathogen interactions with host demography. Mathematical Biosciences and Engineering, 2015, 12(3): 473-490. doi: 10.3934/mbe.2015.12.473
    [3] Bruno Buonomo, Alberto d’Onofrio, Deborah Lacitignola . Rational exemption to vaccination for non-fatal SIS diseases: Globally stable and oscillatory endemicity. Mathematical Biosciences and Engineering, 2010, 7(3): 561-578. doi: 10.3934/mbe.2010.7.561
    [4] Guangxun Sun, Binxiang Dai . Stability and bifurcation of a delayed diffusive predator-prey system with food-limited and nonlinear harvesting. Mathematical Biosciences and Engineering, 2020, 17(4): 3520-3552. doi: 10.3934/mbe.2020199
    [5] Honghua Bin, Daifeng Duan, Junjie Wei . Bifurcation analysis of a reaction-diffusion-advection predator-prey system with delay. Mathematical Biosciences and Engineering, 2023, 20(7): 12194-12210. doi: 10.3934/mbe.2023543
    [6] Shunyi Li . Hopf bifurcation, stability switches and chaos in a prey-predator system with three stage structure and two time delays. Mathematical Biosciences and Engineering, 2019, 16(6): 6934-6961. doi: 10.3934/mbe.2019348
    [7] Zuolin Shen, Junjie Wei . Hopf bifurcation analysis in a diffusive predator-prey system with delay and surplus killing effect. Mathematical Biosciences and Engineering, 2018, 15(3): 693-715. doi: 10.3934/mbe.2018031
    [8] Peng Feng, Menaka Navaratna . Modelling periodic oscillations during somitogenesis. Mathematical Biosciences and Engineering, 2007, 4(4): 661-673. doi: 10.3934/mbe.2007.4.661
    [9] Hongying Shu, Wanxiao Xu, Zenghui Hao . Global dynamics of an immunosuppressive infection model with stage structure. Mathematical Biosciences and Engineering, 2020, 17(3): 2082-2102. doi: 10.3934/mbe.2020111
    [10] Huan Kong, Guohong Zhang, Kaifa Wang . Stability and Hopf bifurcation in a virus model with self-proliferation and delayed activation of immune cells. Mathematical Biosciences and Engineering, 2020, 17(5): 4384-4405. doi: 10.3934/mbe.2020242
  • We consider a reaction diffusion equation with a delayed nonlocal nonlinearity and subject to Dirichlet boundary condition. The model equation is motivated by infection dynamics of disease spread (avian influenza, for example) through environment contamination, and the nonlinearity takes into account of distribution of limited resources for rapid and slow interventions to clean contaminated environment. We determine conditions under which an equilibrium with positive value in the interior of the domain (disease equilibrium) emerges and determine conditions under which Hope bifurcation occurs. For a fixed pair of rapid and slow response delay, we show that nonlinear oscillations can be avoided by distributing resources for both fast or slow interventions.


    1. Introduction

    We consider the spread of a disease carried by a biological species and transmitted through contaminated environment. We assume the diseased individuals move randomly in a spatial domain Ω (a smooth open bounded set in a finite dimensional space) following the standard diffusion, and subject to the Dirichlet condition on the boundary Ω as the boundary is not suitable for the diseased individuals to survive (due to disease prevention and control, or due to the natural environmental constraints). We model the situation where the growth of the infection in the biological population is proportional to the number of diseased individuals as the amount of pathogen loads released to the environment is proportional to this number of diseased cases. We further consider the case where a certain amount of resources is available to clean the environment, a portion of the sources can be used to respond to the contamination relatively faster (with a delay given by τ1) and the rest can be used for slower response characterized by another average delay τ2>τ1. This yields the following model

    ut=dΔu+ru[1a1ΩP1(x,y)u(y,tτ1)dya2ΩP2(x,y)u(y,tτ2)dy], (1)

    where u(t,x) is the population density of infected individuals at time t and location x, (t,x)(0,)×Ω, d is the diffusion rate, r is the reproduction ratio of the diseased populations. The total environment available for the pathogen contamination is normalized to 1. In the first nonlocal delayed integration, u(y,tτ1) is the pathogen loads released by the infected individuals at time tτ1 and spatial location y and P1(x,y) is the probability of the pathogen moved from the spatial location y to current location x. A certain biosafety intervention measure is implemented, in proportion to the pathogen loads ΩP1(x,y)u(y,tτ1)dy, but with a time lag τ1. Similar interpretations apply to the second integration, but with a longer delay τ2. The constants a1 and a2 satisfy a1+a2=1, where a1[0,1] represents the allocation of resources to be allocated to implement the intervention measure for either rapid or slow response to protect the environment from being used to be contaminated to spread the disease back to the biological species under consideration. The kernel function are relevant to the mobility of the virus and this can be derived in a similar fashion as in [14].

    Note that we assume the time for the biosafety intervention is much slower than the virus spread in the environment, and hence the delay in the spread process is ignored. Δ stands for the Laplacian operator, with following Dirichlet boundary condition

    u(x,t)=0,xΩandt(0,+)

    which implies that the exterior environment is hostile and the species cannot move across the boundary of environment, and initial condition satisfies

    u(x,s)=η(x,s)0,xΩandt[τ,0],

    where ΩRn(n1) is a bounded domain with smooth boundary Ω, τ=max(τ1,τ2), ηC:=C([τ,0],Y) and Y=L2(Ω).

    This study is motivated by the spread of avian influenza, an infectious disease of birds that is caused by influenza virus type A strains. The involvement of different bird species and their interactions with environments together lead to complex transmission pathways which include birds to birds, birds to mammals, birds to human, birds to insects, human to human, and environment to birds/mammals/human and vice-versa [9]. How to model the interplay of different transmission pathways and its impact on the spread of avian influenza imposes significant challenge [1][11][16]. In the study of Wang et al.[17], a system of reaction diffusion equations on unbounded domains was proposed to establish the existence and nonexistence of traveling wave solutions of a reaction-convection epidemic model for the spatial spread of avian influenza involving a wide range of bird species and environmental contamination. In the earlier studies of Gourley et al.[6], the role of migrating birds were examined using partial differential equations and their reduction to delay differential systems. Here we focus on the spread of avian influenza among the wild birds, where the virus is shredded into the environment, through which the virus further spreads and infects other wild birds coming to contact with contaminated environment. The parameter r represents the intrinsic susceptibility and transmissibility of the environment, which can be reduced through biosafety intervention so the nonlinearity in the kinetic equation for the infected individuals resembles the classical delayed non-local logistic equations. Note that a large portion of the environment for the virus spread and contamination involves water, the kernel functions P1 and P2 for the virus spread in the environment can involve both diffusion and convercation.

    Our goal in this paper is to 1). Determine whether there is a critical value of r above which the disease will persist in the population in the form of a nonnegative non-trivial equilibrium (note this is necessarily spatially varying due to the Dirichlet condition); 2). Identify critical value of the rapid response delay where the nontrivial equilibrium remains locally stable when all resources are committed for the rapid biosafety intervention; 3). Identify critical value of the slow response delay where the nontrivial equilibrium loses its locally stability even when all resources are committed for the slow biosafety intervention; 4). Identify the critical resource allocation parameter α when a Hopf bifurcation takes place from the nontrivial equilibrium, in this case we examine the patterns of bifurcated periodic solutions to examine impact of parameters on the peak and frequency of the spatiotemporally varying stable patterns.

    We use r to denote the principal eigenvalue of the following one-dimensional eigenvalue problem

    {dΔu(x)=ru(x),xΩ,u(x)=0,xΩ,

    and ϕ is the corresponding eigenfunction of r with ϕ(x)>0 for xΩ. The following notations are needed. Let Lp(Ω) (p1) be the space consisting of measurable functions on Ω that are p-integrable, and Hk(Ω) (k0) be the space consisting of functions whose k-th order weak derivatives belong to L2(Ω). Denote the spaces X=H2(Ω)H10(Ω) and Y=L2(Ω), where H10(Ω)={uH1(Ω)|u(x)=0 for all xΩ}. For any real-valued vector space Z, we also denote the complexification of Z to be ZC:=ZiZ={x1+ix2|x1,x2Z}. For a linear operator L:Z1Z2, we denote the domain of L by D(L), the null space by N(L) and the range of L by R(L). For the complex-valued Hilbert space YC, the standard inner product is <u,v>=Ωˉu(x)v(x)dx. In what follows, we assume the comparability condition between the kernel functions Pi(x,y), i=1,2 and the eigenfunction ϕ. Namely, we assume ΩΩ(P1(x,y)+P2(x,y))ϕ(y)ϕ(x)dydx0. This is because one of the kernel functions can be zero.


    2. Existence of steady state solution

    The positive steady state solutions of (1) satisfy the following equation:

    {dΔu+ru[12i=1aiΩPi(x,y)u(y)dy]=0,xΩ,u(x)=0,xΩ. (2)

    Let N(dΔ+r) and R(dΔ+r) be the null space and the range of the operator dΔ+r, then

    N(dΔ+r)=span{ϕ},R(dΔ+r)={yL2(Ω)|<ϕ,y>=0}.

    Then we have the following decompositions:

    X=N(dΔ+r)ˆX,Y=N(dΔ+r)R(dΔ+r),

    where

    ˆX={yX|<ϕ,y>=0}.

    Then we have the following result on positive steady state solution of model (1).

    Theorem 2.1. There exist r>r and a continuously differential mapping r(ξr,αr) from [r,r] to ˆX×R+ such that the model (1) has a positive steady state solution as follows

    ur(x)=αr(rr)[ϕ(x)+(rr)ξr(x)],r[r,r].

    Moreover,

    αr=Ωϕ2(x)dxr(2i=1aiΩΩPi(x,y)ϕ2(x)ϕ(y)dxdy)

    and ξrX1 is the unique solution of the following equation

    (dΔ+r)ξ+ϕ[1rαr(2i=1aiΩPi(x,y)ϕ(y)dy)]=0.

    The proof is standard. Namely, we let f:ˆX×R×RY be defined by

    f(ξr,αr,r)=(dΔ+r)ξr+ϕ(x)+(rr)ξrrαr(ϕ(x)+(rr)ξr)(2i=1aiΩPi(x,y)(ϕ(y)+(rr)ξ(y))dy),

    then f(ξr,αr,r)=0. The partial derivative of f at (ξr,αr,r) is given by

    D(ξr,αr)f(ξr,αr,r)(η,ϵ)=(dΔ+r)ηrϕ(x)ϵ(2i=1aiΩPi(x,y)ϕ(y)dy).

    Under the comparability condition, we have ϕ(x)Ω(P1(x,y)+P2(x,y))ϕ(y)dyR(dΔ+r). Then D(ξr,αr)f(ξr,αr,r) is bijective from ˆX×R to Y. From the implicit function theorem, there exist r>r and a unique continuously differential mapping r(ξr,αr) from [r,r] to ˆX×R+ such that

    f(ξr,αr,r)=0,r[r,r].

    Therefore, ur(x)=αr(rr)[ϕ(x)+(rr)ξr(x)] solves the boundary value problem (2).

    In what follows, we always assume that r(r,r] and rr1.


    3. Eigenvalue analysis

    It is easy to see that the linearized equation of the model (1) at the steady state solution ur can be written as

    {v(x,t)t=dΔv(x,t)+rv(x,t)[1(2i=1aiΩPi(x,y)ur(y)dy)]rur(x)(2i=1aiΩPi(x,y)v(y,tτi)dy),xΩ,t>0,v(x,t)=0,xΩ, t>0,v(x,t)=η(x,t),(x,t)Ω×[τ,0], (3)

    where ηC.

    Define a operator Ar:D(Ar)Y with domain D(Ar)=X by

    Ar=dΔ+r[1(2i=1aiΩPi(x,y)ur(y)dy)].

    From [13], Ar is an infinitesimal generator of a strong continuous semigroup and Ar is also self-adjoint. Then the study of the stability of ur is transferred to the analysis of the following eigenvalue problem

    Λ(r,λ,τ1,τ2)ψ=Arψrur(2i=1aieλτiΩPi(x,y)ψ(y)dy)λψ=0, (4)

    where ψXC{0}, i.e., the study of the following spectral set

    σ(Aτ1τ2,r)={λC:Λ(r,λ,τ1,τ2)ψ=0,forψXC{0}},

    where Aτ1τ2,r is the infinitesimal generator of the semigroup induced by the solutions of equation (3) with

    Aτ1τ2,rψ=˙ψ,

    and

    D(Aτ1τ2,r)={ψCCC1C:ψ(0)XC,˙ψ(0)=Arψ(0)rur(2i=1aiΩPi(x,y)ψ(y,τi)dy)},

    where C1C=C1([τ,0],YC).

    Then Aτ1τ2,r has a purely imaginary roots λ=iω(ω0) for τ1, τ20 if and only if

    Arψrur(2i=1aieiωτiΩPi(x,y)ψ(y)dy)iωψ=0

    is solvable for some ω>0 and ψXC{0}.

    Next, we discuss the effects of two nonlocal delays on the stability at the positive steady state solution ur in four different cases.

    Case 1. τ1=0 and τ2>0.

    In this case, the equation (4) can be reduced into

    Λ(r,λ,0,τ2)ψ=Arψrur[a1ΩP1(x,y)ψ(y)dy+a2eλτ2ΩP2(x,y)ψ(y)dy]λψ=0. (5)

    If there exit iωτ2(ωτ2>0) and ψτ2XC{0} satisfying (5), then

    Arψτ2rurΩ(a1P1(x,y)+a2eiωτ2τ2P2(x,y))ψτ2(y)dyiωτ2ψτ2,ψτ2=0. (6)

    Separating the real and imaginary parts, we obtain

    ωτ2ψτ2,ψτ2=ImrurΩ(a1P1(x,y)+a2eiωτ2τ2P2(x,y))ψτ2(y)dy,ψτ22i=1|rurΩaiPi(x,y)ψτ2(y)dy,ψτ2|.

    Thus,

    ωτ2rr(a1+a2)rαr(ϕ+(rr)ξr)maxˉΩ×ˉΩPi(x,y)|Ω|. (7)

    It implies that ωτ2rr is uniformly bounded for r(r,r].

    Ignoring a scalar factor, we know that ψτ2 can be expressed as

    ψτ2=βτ2ϕ+(rr)zτ2,ϕ,zτ2=0,βτ20,ψτ22YC=β2τ2ϕ2YC+(rr)2zτ22YC=ϕ2YC. (8)

    Substituting (8) and ωτ2=(rr)kτ2 into (5), we have the following equivalent equation to (5)

    g1(zτ2,βτ2,kτ2,r)=(dΔ+r)zτ2+[1ikτ22i=1ΩairαrPi(x,y)(ϕ(y)+(rr)ξr(y))dy](βτ2ϕ+(rr)zτ2)rαr(ϕ(x)+(rr)ξr(x))Ω(a1P1(x,y)+a2eiωτ2τ2P2(x,y))(βτ2ϕ+(rr)zτ2)dy=0,g2(zτ2,βτ2,kτ2,r)=(β2τ21)ϕ2YC+(rr)2zτ22YC=0.

    Define Gτ2:ˆXC×R3YC×R by Gτ2=(g1,g2). It is clear that

    Gτ2(zτ2,r,βτ2,r,kτ2,r,r)=0.

    Denote

    ˜ai=aiΩΩPi(x,y)ϕ2(x)ϕ(y)dxdy,fori=1,2.

    Separating the real and imaginary parts of g1(zτ2,r,βτ2,r,kτ2,r,r)=0, we have

    {(dΔ+r)z1τ2,r+[1(2i=1airαrΩPi(x,y)ϕ(y)dy)]ϕrαrϕ(x)Ω(a1P1(x,y)+a2P2(x,y)cos(ωτ2,rτ2))ϕ(y)dy=0,(dΔ+r)z2τ2,rkτ2,rϕ+a2rαrϕ(x)ΩP2(x,y)ϕ(y)dysin(ωτ2,rτ2)=0, (9)

    where zτ2,r=z1τ2,r+iz2τ2,r.

    From Theorem (2.1), it can be seen that Eq. (9) is solvable if and only if

    zτ2,r=(1ikτ2,r)ξr,kτ2,r=˜a22˜a21˜a1+˜a2,(˜a2>˜a1)βτ2,r=1,ωτ2,rτ2=arccos(˜a1˜a2)+2nπ,n=0,1,2,.

    Case 2. τ1>0 and τ2=0.

    Assume that ˜a1˜a2. Using a similar analysis as in the Case 1, we can obtain the following result:

    If there exit iωτ1(ωτ1>0) and ψτ1XC{0} satisfying

    Arψτ1rurΩ(a1eiωτ1τ1P1(x,y)+a2P2(x,y))ψτ1(y)dyiωτ1ψτ1=0,

    then ωτ1rr is uniformly bounded for r(r,r]. The equivalent equation to (5) is Gτ1=(g1(zτ1,βτ1,kτ1,r),g2(zτ1,βτ1,kτ1,r)), where

    g1(zτ1,βτ1,kτ1,r)=(dΔ+r)zτ1+[1ikτ12i=1ΩairαrPi(x,y)(ϕ(y)+(rr)ξr(y))dy](βτ1ϕ+(rr)zτ1)rαr(ϕ(x)+(rr)ξr(x))Ω(a1eiωτ1τ1P1(x,y)+a2P2(x,y))(βτ1ϕ+(rr)zτ1)dy=0,g2(zτ1,βτ1,kτ1,r)=(β2τ11)ϕ2YC+(rr)2zτ12YC=0.

    Moreover, it is easy to see that Gτ1(zτ1,r,βτ1,r,kτ1,r,r)=0, where

    zτ1,r=(1ikτ1,r)ξr,kτ1,r=˜a21˜a22˜a1+˜a2,(˜a1>˜a2)βτ1,r=1,ωτ1,rτ1=arccos(˜a2˜a1)+2nπ,n=0,1,2,.

    Case 3. τ1(0,τ10) and τ2>0.

    In this case, we consider τ2 as a parameter and τ1 being in the stable interval (0,τ10). Assume that, for some τ2>0, iωτ1τ2 (ωτ1τ2>0) and ψτ1τ2XC{0} are a solution of the equation (4). If we substitute this solution into the inner product Δ(r,iωτ1τ2,τ1,τ2)ψτ1τ2,ψτ1τ2 and separate the imaginary part, then we obtain the following equation:

    ωτ1τ2ψτ1τ2,ψτ1τ2=ImrurΩ(a1eiωτ1τ2τ1P1(x,y)+a2eiωτ1τ2τ2P2(x,y))ψτ1τ2(y)dy,ψτ1τ2.

    Therefore, ωτ1τ2rr has the same boundary as shown in (7), i.e., ωτ1τ2rr is uniformly bounded for r(r,r]. Then we can rewrite ψτ1τ2 as

    ψτ1τ2=βτ1τ2ϕ+(rr)zτ1τ2,ϕ,zτ1τ2=0,βτ1τ20,ψτ1τ22YC=β2τ1τ2ϕ2YC+(rr)2zτ1τ22YC=ϕ2YC. (10)

    Based on (10) and ωτ1τ2=(rr)kτ1τ2, we obtain the equivalent equation to (5) as follows:

    g1(zτ1τ2,βτ1τ2,kτ1τ2,r)=(dΔ+r)zτ1τ2+[1ikτ1τ22i=1ΩairαrPi(x,y)(ϕ(y)+(rr)ξr(y))dy](βτ1τ2ϕ+(rr)zτ1τ2)rαr(ϕ(x)+(rr)ξr(x))2i=1Ωaieiωτ1τ2τiPi(x,y)(βτ1τ2ϕ+(rr)zτ1τ2)dy=0,g2(zτ1τ2,βτ1τ2,kτ1τ2,r)=(β2τ1τ21)ϕ2YC+(rr)2zτ1τ22YC=0.

    Define Gτ1τ2:ˆXC×R3YC×R by Gτ1τ2=(g1,g2). By separating the real and imaginary parts of g1(zτ1τ2,r,βτ1τ2,r,kτ1τ2,r,r)=0 similar to Case 1, it is easy to see that Gτ1τ2=0 at r=r when the following equations are satisfied

    zτ1τ2,r=(1ikτ1τ2,r)ξr,kτ1τ2,r=˜a22˜a21˜a1+˜a2,(˜a2>˜a1)βτ1τ2,r=1,ωτ1τ2,rτ2=arccos(˜a1˜a2)+2nπ,n=0,1,2,.

    Case 4. τ1>0 and τ2(0,τ20).

    Assume that there exit iωτ2τ1(ωτ2τ1>0) and ψτ2τ1XC{0} such that

    Arψτ2τ1rur(2i=1aieiωτ2τ1τiΩPi(x,y)ψτ2τ1(y)dy)iωτ2τ1ψτ2τ1=0.

    By the similar analysis as in the Case 3, the following result can be obtained:

    ωτ2τ1rr is uniformly bounded for r(r,r]. The equivalent equation to (5) is Gτ2τ1=(g1(zτ2τ1,βτ2τ1,kτ2τ1,r),g2(zτ2τ1,βτ2τ1,kτ2τ1,r))=0, which has the same form as Gτ1τ2. Moreover, Gτ2τ1=0 at r=r if the following equations are satisfied

    zτ2τ1,r=(1ikτ2τ1,r)ξr,kτ2τ1,r=˜a21˜a22˜a1+˜a2,(˜a1>˜a2)βτ2τ1,r=1,ωτ2τ1,rτ1=arccos(˜a2˜a1)+2nπ,n=0,1,2,.

    Since stability analysis is similar for the above four cases, we will only discuss Case 3. For other cases, we omit them in this paper.

    Theorem 3.1. There exists a continuously differentiable mapping

    r(zτ1τ2,r,βτ1τ2,r,kτ1τ2,r)

    from [r,r] to XC×R3 such that Gτ1τ2(zτ1τ2,r,βτ1τ2,r,kτ1τ2,r,r)=0. Furthermore, the solution of Gτ1τ2=0 is unique for r(r,r].

    Proof. Define θτ1τ2=ωτ1τ2τ2. Let T=(T1,T2):XC×R3YC×R be defined by the Fréchet derivative of G at r=r as follows

    T1(z,β,k,θ)=(dΔ+r)z+[1rαr(a1eiωτ1τ2,rτ1ΩP1(x,y)ϕ(y)dy+(a2ia22a21)ΩP2(x,y)ϕ(y)dy)a22a21a1+a2i]ϕβiϕkrαr(a1ia22a21)ϕ(x)θΩP2(x,y)ϕ(y)dy,T2(z,β,k,θ)=2ϕ2YCβ.

    Then T is one-to-one from XC×R3 to YC×R. Hence, from the implicit function theorem, the proof of the existence is completed. Since the proof of the uniqueness is similar to Theorem 2.4 in [15], we omit it here.

    Now, from the analysis above, we can obtain the following conclusion:

    Remark 1. For r(r,r], the eigenvalue problem

    Δ(r,iωτ1τ2,τ1,τ2)ψτ1τ2=0,ωτ1τ2>0,τ2>0,ψXC{0}

    has a solution ψτ1τ2=βτ1τ2ϕ+(rr)zτ1τ2 if and only if

    ωτ1τ2=(rr)kτ1τ2,

    where zτ1τ2,βτ1τ2,kτ1τ2 are defined in Case 3.


    4. Hopf bifurcation

    Theorem 4.1. When τ1=τ2=0, all eigenvalues of Aτ1τ2,r have negative real parts for any r(r,r], i.e., ur is locally asymptotically stable for τ1=τ2=0.

    The proof is essentially same as Proposition 2.9 in [3], hence is omitted.

    Next, we introduce the adjoint operator of Aτ1τ2,r and Δ(r,iωτ1τ2,τ1,τ2), denoted by Aτ1τ2,r and Δ(r,iωτ1τ2,τ1,τ2), respectively. Δ(r,iωτ1τ2,τ1,τ2) is defined as follows

    Δ(r,iωτ1τ2,τ1,τ2)ψ=Arψ+iωτ1τ2ψr2i=1Ωaieiωτ1τ2τiPi(x,y)ur(y)ψ(y)dy.

    Similar to the analysis of (4), we conclude that the following adjoint equation

    Arψr(2i=1aieiωτ1τ2τiΩPi(x,y)ur(y)ψ(y)dy)+iωτ1τ2ψ=0 (11)

    is solvable and the solution is denoted by ωτ1τ2>0 and ψXC{0}. It is well-known that the spectrum set satisfies

    σ(Δ(r,iωτ1τ2,τ1,τ2))=σ(Δ(r,iωτ1τ2,τ1,τ2)).

    Define the following function

    Sn(r):=Ω¯ψ(x)ψ(x)dxra1τ1ΩΩP1(x,y)ur(x)¯ψ(x)ψ(y)dxdyeiωτ1ra2τ2nΩΩP2(x,y)ur(x)¯ψ(x)ψ(y)dxdyeiωτ2n.

    It is easy to see that

    Sn(r)[(a1+a2)ΩΩP2(x,y)ϕ2(x)ϕ(y)dxdy2i=1aiΩΩPi(x,y)ϕ2(x)ϕ(y)dxdy(arccos(a1a2cos(ωτ1τ2,rτ1))+2nπ)a1sin(ωτ1τ2,rτ1)a22a21cos2(ωτ1τ2,rτ1)a1a2(ia22a21cos2(ωτ1τ2,rτ1)+a1cos(ωτ1τ2,rτ1))+1]Ωϕ2(x)dx,asrr,

    which leads to Sn(r)0 for any r(r,r].

    Theorem 4.2. For r(r,r], iωτ1τ2 is a simple eigenvalue of Aτ1τ2n,r, n=0,1,2.

    Proof. Notice that N[Aτ1τ2n,riωτ1τ2]=Span{eiωτ1τ2ψτ1τ2}. If ξD(Aτ1τ2n,r)D([Aτ1τ2n,r]2), then we can obtain

    [Aτ1τ2n,riωτ1τ2]2ξ=0,

    which leads to

    [Aτ1τ2n,riωτ1τ2]ξN[Aτ1τ2n,riωτ1τ2]=Span{eiωτ1τ2ψτ1τ2}.

    Thus, there exists a constant l such that

    [Aτ1τ2n,riωτ1τ2]ξ=leiωτ1τ2ψτ1τ2,

    i.e.,

    ˙ξ(θ)=iωτ1τ2ξ(θ)+leiωτ1τ2θψτ1τ2θ[τ2n,0]˙ξ(0)=Arξ(0)a1rurΩP1(x,y)ξ(τ1)(y)dya2rurΩP2(x,y)ξ(τ2n)(y)dy. (12)

    The first equation of (12) leads to

    ξ(θ)=ξ(0)eiωτ1τ2θ+lθeiωτ1τ2θψτ1τ2˙ξ(0)=iωτ1τ2ξ(0)+lψτ1τ2.

    Thus, we have

    Δ(r,iωτ1τ2,τ1,τ2n)ξ(0)=l[ψτ1τ2rur(a1τ1eiωτ1ΩP1(x,y)ψ(y)dy+a2τ2neiωτ2nΩP2(x,y)ψ(y)dy)].

    Moreover,

    0=Δ(r,iωτ1τ2,τ1,τ2n)ψτ1τ2,ξ(0)=ψτ1τ2,Δ(r,iωτ1τ2,τ1,τ2n)ξ(0)=l[Ω¯ψ(x)ψ(x)dxrΩΩ¯ψ(x)ur(x)(a1τ1eiωτ1τ2τ1P1(x,y)+a2τ2neiωτ1τ2τ2P2(x,y))ψ(y)dxdy]=lSn(r).

    Due to Sn(r)0, the coefficient l=0 and this leads to ξN[Aτ1τ2n,riωτ1τ2]. Hence, we have

    ξN[Aτ1τ2n,riωτ1τ2]j=N[Aτ1τ2n,riωτ1τ2],j=1,2,3,n=0,1,2,

    and this shows that λ=iωτ1τ2 is a simple eigenvalue of Aτ1τ2n,r for n=0,1,2. This completes the proof.

    From the implicit function theorem, we can obtain that there is a neighborhood On×Dn×HnR×C×XC of (τ2n,iω,ψ) and a continuously differential function (λ,ψ):OnDn×Hn such that for each τ2On, the only eigenvalue of Aτ1τ2,r in Dn is μ(r) and

    λ(τ2n)=iωτ1τ2,r,ψ(τ2n)=ψτ1τ2,r,Δ(r,μ,τ1,τ2)ψ=(Arμ(τ2))ψrur2i=1Ωaieμ(τ2)τiPi(x,y)ψ(τ2)(y)dy=0. (13)

    Then, the following result describes the transversality condition of Hopf bifurcation:

    Theorem 4.3. For any r(r,r], Redλ(τ2n)τ2>0, n=0,1,2.

    Proof. Differentiating (13) with respect to τ2 at τ2=τ2n, we obtain that

    dλ(τ2)dτ2=a2riωΩΩ¯ψ(x)ur(x)P2(x,y)ψ(τ2)(y)dxdyeiθΩ¯ψ(x)ψ(x)dxrΩΩ(2i=1aiτieiωτiPi(x,y))¯ψ(x)ur(x)ψ(y)dxdy=1|Sn(r)|2{a2riωeiθΩψ(x)ˉψ(x)dxΩΩP2(x,y)ur(x)¯ψ(x)ψ(y)dxdya1a2r2iωτ1ei(ωτ1θ)ΩΩP1(x,y)ur(x)ψ(x)ˉψ(y)dxdyΩΩP2(x,y)ur(x)¯ψ(x)ψ(τ2)(y)dxdya22r2τ2iωΩΩP2(x,y)ur(x)ψ(x)ˉψ(y)dxdyΩΩP2(x,y)ur(x)¯ψ(x)ψ(τ2)(y)dxdy}.

    Therefore,

    limrrRe(dλ(τ2)dτ2)=1|Sn(r)|2a22a21cos2(ωτ1τ2τ1)arccos(a1a2cos(ωτ1τ2τ1))rαrΩϕ2(x)dxΩΩP2(x,y)ϕ2(x)ϕ(y)dxdy>0.

    Then we conclude

    Theorem 4.4. For r(r,r], the positive steady state solution ur of model (1) is locally asymptotically stable for τ2[0,τ20) and there undergoes Hopf bifurcation at τ2=τ20.


    5. Stability of bifurcated periodic solutions

    This section contains lengthy and technical discussions about the direction of Hopf bifurcation, stability and period of the periodic solution bifurcating from the positive steady solution ur. Following the ideas of Wu [19], we derive the explicit formulae for determining the properties of Hopf bifurcation at the critical value τ20 for fixed τ1(0,τ10) by employing the normal form method and center manifold theorem. Without loss of generality, this section assumes that τ1<τ20. Let U(t)=u(,t)ur and τ2=τ20+ν. Then ν=0 is the Hopf bifurcation value of model (1). Re-scaling the time by ttτ2 to normalize the delay, model (1) is transformed into the following form

    dU(t)dt=τ20dΔU(t)+τ20L0(Ut)+F(Ut,ν), (14)

    and L0:CC, F:C×RC are given respectively by

    L0(ψ)=r[1(2i=1aiΩPi(x,y)ur(y)dy)]ψ(0)rur[a1ΩP1(x,y)ψ(τ1τ20)dy+a2ΩP2(x,y)ψ(1)dy],F(ψ,ν)=νdΔψ(0)+νL0(ψ)r(ν+τ20)(a1ΩP1(x,y)ψ(τ1τ20)dy+a2ΩP2(x,y)ψ(1)dy)ψ(0),

    where ψC([1,0],Y).

    There exists a function η(θ,x,ψ(θ)) of bounded variation for θ[1,0] such that

    L0ψ=01dη(θ,x,ψ(θ)),

    where

    η(θ,x,ψ(θ))=r[1(2i=1aiΩPi(x,y)ur(y)dy)]δ(θ)ψ(θ)a1rurΩP1(x,y)δ(θ+τ1τ20)ψ(θ)(y)dya2rurΩP2(x,y)δ(θ+1)ψ(θ)(y)dy.

    For ψC([1,0],Y), define

    Aτ2ψ={dψ(θ)dθθ[1,0),τ2dΔψ(0)+τ201dη(θ,x,ψ(θ))θ=0,

    and

    R(ψ,ν)={0θ[1,0)F(ψ,ν)θ=0

    Then system (14) is equivalent to

    dUtdt=Aτ2Ut+R(Ut,ν), (15)

    where Ut(θ)=U(t+θ) for θ[1,0].

    For ψC([0,1],Y), define

    Aτ2˜ψ(s)={d˜ψ(s)dss(0,1],τ2dΔ˜ψ(0)+τ201dη(s,x,˜ψ(s))s=0,

    and the formal duality

    ˜ψ,ψ≫=˜ψ(0),ψ(0)01θξ=0˜ψ(ξθ),dη(θ,y,ψ(ξ))dξ.

    From the previous definition, we have

    Aτ2˜ψ,ψ=Aτ2˜ψ(0),ψ(0)a1rτ200τ1τ2Aτ2˜ψ(s+τ1τ2),ur(x)ΩP1(x,y)ψ(s)(y)dydsa2rτ2001Aτ2˜ψ(s+1),ur(x)ΩP2(x,y)ψ(s)(y)dyds=τ2dΔ˜ψ(0)+τ2r(12i=1ΩaiPi(x,y)ur(y)dy)˜ψ(0)ra1τ2ur(x)ΩP1(x,y)ψ(τ1τ2)(y)dyra2τ2ur(x)ΩP2(x,y)ψ(1)(y)dy,ψ(0)a1rτ200τ1τ2˙˜ψ(s+τ1τ2),ur(x)ΩP1(x,y)ψ(s)(y)dydsa2rτ201˙˜ψ(s+1),ur(x)ΩP2(x,y)ψ(s)(y)dyds=˜ψ(0),τ2dΔψ(0)+τ2r(12i=1ΩaiPi(x,y)ur(y)dy)ψ(0)ra1τ2˜ψ(τ1τ2),ur(x)ΩP1(x,y)ψ(0)(y)dyra2τ2˜ψ(1),ur(x)ΩP2(x,y)ψ(0)(y)dy+a1rτ200τ1τ2˙˜ψ(s+τ1τ2),ur(x)ΩP1(x,y)ψ(s)(y)dyds+a2rτ201˙˜ψ(s+1),ur(x)ΩP2(x,y)ψ(s)(y)dyds=˜ψ(0),Aτ2ψ(0)a1rτ20τ1τ2˜ψ(s+τ1τ2),ur(x)ΩP1(x,y)˙ψ(s)(y)dydsa2rτ2001˜ψ(s+1),ur(x)ΩP2(x,y)˙ψ(s)(y)dyds=˜ψ,Aτ2ψ.

    Since ±iωτ1τ2τ20 are eigenvalues of Aτ2, they are also eigenvalues of Aτ2. Based on the previous eigenvalue analysis, ψτ1τ2eiωτ1τ2τ20θ and ˉψτ1τ2eiωτ1τ2τ20θ are the eigenfunctions of Aτ2 corresponding to iωτ1τ2τ20 and iωτ1τ2τ20, respectively. Let Φ=(q(θ),ˉq(θ))=(ψτ1τ2eiωτ2θ,ˉψτ1τ2eiωτ2θ), θ[1,0], then P=Span{Φ} is the generalized eigenspace of Aτ2 with respect to eigenvalues set {iωτ1τ2τ20,iωτ1τ2τ20}. Similarly P=Span{q(s),¯q(s)} is generalized eigenspace of the adjoint operator Aτ2, where q(s)=ψτ1τ2eiωτ2s is the eigenfunction with respect to iωτ2. Then the phase space CC can be decomposed as CC=PQ, where

    Q={ψCC:≪ψ,ϕ≫=0,for allψP}.

    Denote

    Ψ=(1ˉSn(r)q(s)1Sn(r)¯q(s)),

    then Ψ,Φ≫=I2×2.

    Let Ut be the solution of (15) when ν=0. Define

    z(t)=≪1ˉSn(r)q(s),Ut,W(t,θ)=UtΦ(θ)(z(t),ˉz(t))T. (16)

    Then we obtain the following center manifold

    W(t,θ)=W(z(t),ˉz(t),θ)=W20(θ)z22+W11(θ)zˉz+W02(θ)ˉz22+ (17)

    with the range in Q. From the definition of (16), we have

    ˙z(t)=ddt1ˉSn(r)q(s),Ut=≪1ˉSn(r)q(s),Aτ2Ut+1ˉSn(r)q(s),R(Ut,0)=≪1ˉSn(r)Aτ2q(s),Ut+1Sn(r)q(0),F(Ut,0)=iωτ20z(t)+1Sn(r)q(0),F(W(z(t),ˉz(t),0)+2Re{z(t)q(θ)},0).

    We rewrite this equation as

    ˙z(t)=iωτ20z(t)+g(z,ˉz),

    where

    g(z,ˉz)=1Sn(r)q(0),F(W(z(t),ˉz(t),0)+2Re{z(t)q(θ)},0)=g20z22+g11zˉz+g02ˉz22+g21z2ˉz2+. (18)

    Computing the coefficients of (18), we have

    g20=2τ20rSnΩΩ(a1eiωτ1P1(x,y)+a2eiωτ20P2(x,y))ˉψτ1τ2(x)ψτ1τ2(x)ψτ1τ2(y)dxdy,
    g11=τ20rSn[ΩΩ(a1eiωτ1P1(x,y)+a2eiωτ20P2(x,y))ˉψτ1τ2(x)ˉψτ1τ2(x)ψτ1τ2(y)dxdy+ΩΩ(a1eiωτ1P1(x,y)+a2eiωτ20P2(x,y))ˉψτ1τ2(x)ψτ1τ2(x)ˉψτ1τ2(y)dxdy],
    g02=2τ20rSnΩΩ(a1eiωτ1P1(x,y)+a2eiωτ20P2(x,y))ˉψτ1τ2(x)ˉψτ1τ2(x)ˉψτ1τ2(y)dxdy,
    g21=2τ20rSnΩΩ(a1eiωτ1P1(x,y)+a2eiωτ20P2(x,y))ˉψτ1τ2(x)ψτ1τ2(y)W11(0)(x)dxdyτ20rSnΩΩ(a1eiωτ1P1(x,y)+a2eiωτ20P2(x,y))ˉψτ1τ2(x)ˉψτ1τ2(y)W20(0)(x)dxdyτ20rSnΩΩˉψτ1τ2(x)ˉψτ1τ2(x)(a1P1(x,y)W20(τ1τ2)(y)+a2P2(x,y)W20(1)(y))dxdy2τ20rSnΩΩˉψτ1τ2(x)ψτ1τ2(x)(a1P1(x,y)W11(τ1τ2)(y)+a2P2(x,y)W11(1)(y))dxdy.

    From the expression of g21, we need to compute W20(θ) and W11(θ). From (15) and (16), we have

    ˙W={Aτ2WΦ(θ)Ψ(0),F(W(z,ˉz)+Φ(z,ˉz)T,0),1θ<0,Aτ2WΦ(θ)Ψ(0),F(W(z,ˉz)+Φ(z,ˉz)T,0)+F(W(z,ˉz)+Φ(z,ˉz)T,0)θ=0,=Aτ2W+H(z,ˉz,θ), (19)

    where

    H(z,ˉz,θ)=H20(θ)z22+H11(θ)zˉz+H02(θ)ˉz22+. (20)

    Due to the chain rule

    ˙W=Wz˙z+Wˉz˙ˉz,

    we have

    (2iωτ1τ2τ20+Aτ2)W20(θ)=H20(θ),Aτ2W11(θ)=H11(θ). (21)

    From (19), we know that for θ[1,0)

    H(z,ˉz,θ)=Φ(θ)Ψ(0),F(W(z,ˉz,θ)+Φ(z,ˉz)T,0)=gq(θ)ˉgˉq(θ).

    Comparing the coefficients with (20), we obtain

    H20(θ)=g20q(θ)ˉg02ˉq(θ),H11(θ)=g11q(θ)ˉg11ˉq(θ). (22)

    From (21) and (22) and the definition of Aτ2, it follows that

    ˙W20(θ)=2iωτ1τ2τ20W20(θ)+g20q(θ)+ˉg02ˉq(θ). (23)

    Hence,

    W20(θ)=ig20ωτ1τ2τ20q(θ)+iˉg023ωτ1τ2τ20ˉq(θ)+M1e2iωτ1τ2τ20θ. (24)

    Similarly, we can obtain

    W11(θ)=ig11ωτ1τ2τ20q(θ)+iˉg11ωτ1τ2τ20ˉq(θ)+M2. (25)

    In the following we shall find out M1 and M2. From (19) and (20), we have

    H20(0)=g20q(0)ˉg02ˉq(0)2rτ20ψτ1τ2(x)Ω(a1eiωτ1τ2τ1P1(x,y)+a2eiωτ1τ2τ20P2(x,y))ψτ1τ2(y)dy,H11(0)=g11q(0)ˉg11ˉq(0)rτ20[ψτ1τ2(x)(a1ΩP1(x,y)ˉψτ1τ2(y)eiωτ1τ2τ1dy+a2ΩP2(x,y)ˉψτ1τ2(y)eiωτ1τ2τ20dy)+ˉψτ1τ2(x)(a1ΩP1(x,y)ψτ1τ2(y)eiωτ1τ2τ1dy+a2ΩP2(x,y)ψτ1τ2(y)eiωτ1τ2τ20dy)].

    Thus, we can compute M1 and M2 satisfying

    M1=2rΔ1(r,2iωτ1τ2,τ1,τ2)ψτ1τ2(x)Ω(a1eiωτ1τ2τ1P1(x,y)+a2eiωτ1τ2τ20P2(x,y))ψτ1τ2(y)dy,M2=rΔ1(r,0,τ1,τ2)[ψτ1τ2(x)Ω(a1eiωτ1τ2τ1P1(x,y)+a2eiωτ1τ2τ20P2(x,y))ˉψτ1τ2(y)dy+ˉψτ1τ2(x)Ω(a1P1(x,y)eiωτ1τ2τ1+a2P2(x,y)eiωτ1τ2τ20)ψτ1τ2(y)dy]

    Now, we can determine W20(θ) and W11(θ) from (24) and (25). Furthermore, g21 can be expressed. Hence, we can compute the following values

    c1(0)=i2ωτ1τ2τ20(g20g112|g11|2|g02|23)+g212,μ2=Re{c1(0)}Re{λ(τ20)},β2=2Re{c1(0)},T2=Im{c1(0)}+μ2Im{λ(τ20)}ωτ1τ2τ20.

    From the conclusion of [19], [8], we have the following results.

    Theorem 5.1. μ2 determines the direction of the Hopf bifurcation: if μ2>0 (μ2<0), the Hopf bifurcation is supercritical (subcritical); β2 determines the stability of the bifurcating periodic solution: the bifurcating periodic solution is stable (unstable) if β2<0 (β2>0) and T2 determines the period of the bifurcating periodic solution: the period increases (decrease) if T2>0 (T2<0).


    6. Numerical simulation

    In this section, we present some numerical simulations to demonstrate our analytical results.

    We choose the parameters d=1, r=2.5 and the initial condition u(x,t)=0.9sin2x for all τ<t0. And a1=0.6, a2=0.4 satisfying a1+a2=1. Without loss the generality, let a1=α and then a2=1α. Kernel function takes the form Pi(x,y)=14παie(xy)24αi, i=1,2. For the simplicity, α1=α2=1. For τ2=0, we can get τ10=3.4294. Now, let the delay τ1=1[0,τ10) be fixed. Based on theoretical analysis above, we can figure out ωτ1τ2=1.4933, λ(τ20)=0.18100.0225i, c1(0)=0.09410.1298i, μ2=0.52, β2=0.1882, T2=0.0838 and delay critical value τ20=1.1299. Then we know that the positive steady state solution ur is locally asymptotically stable when τ2<τ20. This is numerically illustrated in left panel of Fig. 1. According to Theorem 5.1, model (1) undergoes a supercritical Hopf bifurcation at the positive state solution ur(x) and bifurcating periodic solution exists for τ2 slightly larger than τ20 and the bifurcated periodic solution is stable, as depicted in right panel of Fig. 1. Moreover, Fig. 2 plots a critical curve τ2=f(α) w.r.t. two parameters τ2 and α for the fixed delay τ1=1. We shall explore the significance of the change of monotonicity of this curve in the next section.

    Figure 1. Solutions of model (1) approach to a positive steady state with τ2=0.6 and a periodically oscillatory orbit with τ2=1.2, respectively.
    Figure 2. The critical value of time delay τ2 with respect to varying α(0,0.8).

    7. Discussion

    Here we interpreted the classical logistic model with two non-local delayed terms in the framework of avian influenza spread between wild birds and the environment---the environment is contaminated by infected birds and the contaminated environment then pass on the pathogen to other susceptible birds. Due to the random movement of the infected birds and pathogens, the disease spreads in the geographical domain and pathogen loads in any given spatial location are not just the consequence of local contamination. Here we consider the case where resources are available for cleaning the environment. These resources can be used to launch either rapid or slow environment cleaning interventions, but the resources are limited so optimal allocations will be needed. Our study shows that disease outbreak in the form of a nontrivial equilibrium is possible assuming the intrinsic reproduction number is sufficiently large, and nonlinear oscillations around this nontrivial equilibrium can take place. Our analysis and simulations show that to prevent this oscillation, the resources should be distributed for both rapid and slow responses, focusing on either rapid or slow response will require the slow response to be also very rapid. For example, in Figure 3, if we normalized the delay so that the rapid response takes place with τ1=1, then the critical value for nonlinear oscillation (τ20) to take place can be large, and close to 5 when α is close to 0.5.

    In recent years, reaction-diffusion equations with time delay have been investigated extensively. Su et al.[15] studied a diffusive logistic equation with mixed delayed and instantaneous density dependence, with some interesting results on global continuation of Hopf bifurcation branches. Hu and Yuan[10] proposed a coupled system of reaction-diffusion system with distributed delay and studied stability of the positive steady state solution and the occurrence of Hopf bifurcation. The Hopf bifurcation was also considered in Ma [12] for a coupled reaction-diffusion systems involving three interacting species. The earlier work introducing nonlocal terms into the diffusive Fisher equation included the paper of Britton[2]. Guo[7] investigated the existence, stability and multiplicity of spatially nonhomogeneous steady state solutions and periodic solutions for reaction-diffusion models with nonlocal delay effect by using the Lyapunov-Schmidt reduction. Deng and Wu[5] established a comparison principle and constructed monotone sequences to show the global stability for a nonlocal reaction-diffusion population model. Zuo and Song[22] studied the effect of three weight functions on the dynamics of a general reaction-diffusion equation with nonlocal delay and showed that the average delay for the case of strong kernel may induce the stability switches. Chen and Yu [4] considered a nonlocal delayed reaction-diffusion equation with general form of nonlocal delay. More discussions about the biological backgrounds of non-local reaction diffusion equations with delay and further results on the existence of nontrivial equilibria and Hopf bifurcations can be found in [21][20][18] and references therein. Here we link a logistic model with two non-local delay terms to the understanding of optimal strategies to prevent nonlinear oscillations in disease spread involving environment contamination and resources allocation, and we believe this line of research in modeling and analysis may generate interest for further expanding the models to reflect more biological realities and disease spread such as temporal heterogeneity and multiple routes of transmission.


    Acknowledgments

    This work was supported by the Fundamental Research Funds for the Central University of China (N140504005) and the Natural Sciences and Engineering Research Council of Canada (105588-2011-RGPIN).


    [1] [ L. Bourouiba, S. Gourley, R. Liu, J. Takekawa and J. Wu, Avian Influenza Spread and Transmission Dynamics. In Analyzing and Modeling Spatial and Temporal Dynamics of Infectious Diseases, John Wiley and Sons Inc., 2014.
    [2] [ N. Britton, Reaction-diffusion Equations and Their Applications to Biology, Academic Press, London, 1986.
    [3] [ S. Chen,J. Shi, Stability and Hopf bifurcation in a diffusive logistic population model with nonlocal delay effect, J. Differ. Equations, 253 (2012): 3440-3470.
    [4] [ S. Chen,J. Yu, Stability and bifurcations in a nonlocal delayed reaction-diffusion population model, J. Differ. Equaitons, 260 (2016): 218-240.
    [5] [ K. Deng,Y. Wu, Global stability for a nonlocal reaction-diffusion population model, Nonlinear Anal. Real World Appl., 25 (2015): 127-136.
    [6] [ S. Gourley,R. Lui,J. Wu, Spatiotemporal distributions of migratory birds: Patchy models with delay, SIAM Journal on Applied Dynamical Systems, 9 (2010): 589-610.
    [7] [ S. Guo, Stability and bifurcation in a reaction-diffusion model with nonlocal delay effect, J. Differ. Equations, 259 (2015): 1409-1448.
    [8] [ B. D. Hassard, N. D. Kazarinoff and Y. H. Wan, Theory and Applications of Hopf Bifurcation, Cambridge University Press, Cambridge, 1981.
    [9] [ Y. Hsieh, J. Wu, J. Fang, Y. Yang and J. Lou, Quantification of bird-to-bird and bird-to-human infections during 2013 novel H7N9 avian influenza outbreak in China, PLoS One, 9 (2014), e111834.
    [10] [ R. Hu,Y. Yuan, Spatially nonhomogeneous equilibrium in a reaction-diffusion system with distributed delay, J. Differ. Equations, 250 (2011): 2779-2806.
    [11] [ R. Liu,V. Duvvuri,J. Wu, Spread patternformation of H5N1-avian influenza and its implications for control strategies, Math. Model. Nat. Phenom., 3 (2008): 161-179.
    [12] [ Z. P. Ma, Stability and Hopf bifurcation for a three-component reaction-diffeusion population model with delay effect, Appl. Math. Model., 37 (2013): 5984-6007.
    [13] [ A. Pazy, Semigroups of Linear Operators and Applications to Partial Differential Equations, Springer-Verlag, New York, 1983.
    [14] [ J. So,J. Wu,X. Zou, A reaction-diffusion model for a single species with age structure. I travelling wavefronts on unbounded domains, Proceedings of the Royal Society: London A, 457 (2001): 1841-1853.
    [15] [ Y. Su,J. Wei,J. Shi, Hopf bifurcation in a diffusive logistic equation with mixed delayed and instantaneous density dependence, J. Dyn. Differ. Equ., 24 (2012): 897-925.
    [16] [ X. Wang,J. Wu, Periodic systems of delay differential equations and avian influenza dynamics, J. Math. Sci., 201 (2014): 693-704.
    [17] [ Z. Wang,J. Wu,R. Liu, Traveling waves of the spread of avian influenza, Proc. Amer. Math. Soc., 140 (2012): 3931-3946.
    [18] [ Z. C. Wang,W. T. Li,J. Wu, Entire solutions in delayed lattice differential equations with monostable nonlinearity, SIAM J. Math. Anal., 40 (2009): 2392-2420.
    [19] [ J. Wu, Theory and Applications of Partial Functional Differential Equations, Spinger-Verlag, New York, 1996.
    [20] [ T. Yi,X. Zou, Global dynamics of a delay differential equaiton with spatial non-locality in an unbounded domain, J. Differ. Equations, 251 (2011): 2598-2611.
    [21] [ G. Zhao,S. Ruan, Existence, uniqueness and asymptotic stability of time periodic traveling waves for a periodic Lotka-Volterra competition system with diffusion, J. Math. Pures Appl., 95 (2011): 627-671.
    [22] [ W. Zuo,Y. Song, Stability and bifurcation analysis of a reaction-diffusion equaiton with spatio-temporal delay, J. Math. Anal. Appl., 430 (2015): 243-261.
  • This article has been cited by:

    1. Houfu Liu, Yuanyuan Cong, Ying Su, DYNAMICS OF A TWO-PATCH NICHOLSON'S BLOWFLIES MODEL WITH RANDOM DISPERSAL, 2022, 12, 2156-907X, 692, 10.11948/20210268
  • Reader Comments
  • © 2017 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(3125) PDF downloads(522) Cited by(1)

Article outline

Figures and Tables

Figures(2)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog