MPE Mathematical Problems in Engineering 1563-5147 1024-123X Hindawi Publishing Corporation 495275 10.1155/2014/495275 495275 Research Article Permanence and Extinction of a Stochastic Delay Logistic Model with Jumps http://orcid.org/0000-0002-0639-0746 Lu Chun 1, 2 Ding Xiaohua 1 Ibeas Asier 1 Department of Mathematics Harbin Institute of Technology (Weihai) Weihai 264209 China hit.edu.cn 2 School of Science Qingdao Technological University Qingdao 266520 China qtech.edu.cn 2014 1622014 2014 03 10 2013 10 12 2013 25 12 2013 16 2 2014 2014 Copyright © 2014 Chun Lu and Xiaohua Ding. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This paper is concerned with a stochastic delay logistic model with jumps. Sufficient and necessary conditions for extinction are obtained as well as stochastic permanence. Numerical simulations are introduced to support the theoretical analysis results. The results show that the jump process can affect the properties of the population model significantly, which conforms to biological significance.

1. Introduction

Recently, Freedman and Wu  considered the following delay logistic model: (1)dx(t)dt=x(t)[r(t)-a(t)x(t)+b(t)x(t-τ(t))], where r(t) is the net birth rate, a(t) denotes the self-inhibition rate, b(t) represents the reproduction rate, and τ(t) is the time-varying delay. There is an extensive literature concerned with the properties of system (1) and we here mention  among many others.

As we know, stochastic population models have recently been investigated by many authors (see, e.g., ). Particularly, May  has revealed that due to environmental noises, the growth rate should be stochastic. Suppose that the growth rate r(t) is perturbed by white noise (see, e.g., [9, 10]) r(t)r(t)+σ(t)xθ(t)ω˙(t) where ω˙(t) is the white noise, namely, w(t) is a Brownian motion defined on a complete probability space (Ω,,𝒫) with a filtration {t}tR¯+ satisfying the usual conditions; σ2(t) represents the intensity of the white noise. As a result, (1) becomes the following model: (2)dx(t)=x(t)[r(t)-a(t)x(t)+b(t)x(t-τ(t))]dt+σ(t)x1+θ(t)dw(t).

On the other hand, the population may suffer from sudden environmental shocks, for example, massive diseases like avian influenza and SARS, earthquakes, hurricanes, epidemics, and so forth. Bao et al. [11, 12] and Liu and Wang [13, 14] incorporate a jump process into the underlying population system, which can describe these phenomena well and provide a more practical model. Particularly, the books by Applebaum  and Situ  are also good references in this area. Furthermore, some scholars have researched the theory about stochastic differential delay equations with jumps recently (see, e.g., ). However, as far as our knowledge is concerned, no articles on introducing a jump process into stochastic delay population has been introduced. Motivated by these, we will consider the stochastic delay logistic model with jumps: (3)dx(t)=x(t)[r(t)-a(t)x(t)+b(t)x(t-τ)]dt+σ(t)x1+θ(t)dw(t)+x(t-)𝕐γ(u)N~(dt,du), with the initial data ξ(t)C0b([-τ,0];R+), where C0b([-τ,0];R+) denotes the family of all bounded, 0-measurable, C([-τ,0];R+)-valued random variables and R+=(0,+). Here, x(t-)=limstx(s), N(dt,du) is a real-valued Poisson counting measure with characteristic measure λ on a measurable subset 𝕐 of R¯+ with λ(𝕐)<+, N~(dt,du)=N(dt,du)-λ(du)dt, γ(u) is bounded function, and γ(u)>-1, u𝕐. Furthermore, we assume that w(t) is independent of N and τ(t)τ is nonnegative constant.

Based on the fact that model (3) describes a population dynamics, it is very important to investigate the permanence and extinction. The main aims of this work are to investigate how jump process affect the permanence and extinction of model (3). Our results demonstrate that the jump process can change the permanence and extinction, which accords with biological significance. In addition, we establish the sufficient and necessary conditions for stochastic permanence and extinction of model (3).

For model (3) we always assume the following.

θ>0.75, r(t), a(t), b(t), and σ(t) are continuous bounded functions on R¯+ with inftR¯+a(t)>0, inftR¯+b(t)>0, and inftR¯+σ(t)>0, where R¯+=[0,).

For each m>0 there exists Lm such that 𝕐|H(x,u)-H(y,u)|2λ(du)Lm|x-y|2 where H(x,u)=γ(u)x(t-) with |x||y|m.

There exists a positive constant c such that |ln(1+γ(u))|c for γ(u)>-1.

For the simplicity, we define the following notations: (4)x(t)=1t0tx(s)ds,x*=liminft+x(t),x*=limsupt+x(t),g(t)=r(t)-𝕐[γ(u)-ln(1+γ(u))]λ(du).

The rest of the paper is arranged as follows. In Section 2, we show that model (3) has a unique positive global solution. Afterward, sufficient and necessary conditions for extinction and stochastic permanence are established in Section 3. Section 4 mainly concentrates on introducing some figures to illustrate the main results. Finally, we close the paper with conclusions and remarks in Section 5.

2. Global Positive Solution

The classical existence and uniqueness result for solutions of a stochastic differential delay equation with jumps requires the coefficient functions to satisfy a local Lipschitz condition and a linear growth condition (see, e.g., ). Clearly, the coefficients of (3) satisfy the local Lipschitz condition, while they fail the linear growth condition. In this section, using the Lyapunov analysis method (mentioned in ), we will show that the jump processes can suppress the explosion and the solution of model (3) is positive and global. For later applications, let us cite the following lemma.

Lemma 1.

The following inequalities hold: (5)ln(x+1)x,x>-1,(6)xq1+q(x-1),x0,0q1.

Theorem 2.

Let assumptions (A1)–(A3) hold. For any given initial value ξ(t)C0b([-τ,0];R+), (3) has a unique positive solution x(t)R+ for any t0 almost surely.

Proof.

Since the coefficients of the equation are locally Lipschitz continuous, for any given initial value ξ(t)C0b([-τ,0];R+), there is a unique local solution x on t[-τ,τe), where τe is the explosion time. To show that this solution is global, we need to show that τe=+ a.s. Let k0>0 be sufficiently large for (1/k0)<min-τt0ξ(t)max-τt0ξ(t)<k0. For each time integer kk0, define the stopping time: (7)τk=inf{t[0,τe):x(t)1korx(t)k}, where throughout this paper we set inf=+ (as usually denotes the empty set). Clearly, τk is increasing as k+. Set τ+=limk+τk; hence τ+τe a.s. If we can show that τ+=+ a.s., then τe=+ a.s. and x(t)R+ a.s. for all t0. In other words, to complete the proof all we need to show is that τ+=+ a.s. To show this statement, let us define a C2-function V:R+R+ by V(x)=x-1-0.5lnx. Let kk0 and T>0 be arbitrary. For 0tτkT, applying the Itô’s formula, we obtain (8)d[t-τtx2(s)ds+V(x(t))]=(x2(t)-x2(t-τ))dt+0.5(x-0.5(t)-x-1(t))×[𝕐x(t)(r(t)-a(t)x(t)+b(t)x(t-τ))dt+σ(t)x1+θ(t)dw(t)𝕐]+0.5[-0.25x-1.5(t)+0.5x-2(t)]σ2(t)x2+2θ(t)dt+𝕐[(1+γ(u))0.5-1-0.5γ(u)]λ(du)x0.5(t)dt+0.5𝕐[γ(u)-ln(1+γ(u))]λ(du)dt+𝕐[(1+γ(u))0.5-1]N~(dt,du)x0.5(t)-0.5𝕐[ln(1+γ(u))]N~(dt,du)[x2(t)-x2(t-τ)]dt+[𝕐0.5r(t)(x0.5(t)-1)-0.5a(t)x(t)(x0.5(t)-1)+x2(t-τ)+0.0625b2(t)(x0.5(t)-1)2𝕐]dt+0.5[-0.25x-1.5(t)+0.5x-2(t)]σ2(t)x2+2θ(t)dt+𝕐[(1+γ(u))0.5-1-0.5γ(u)]λ(du)x0.5dt+0.5𝕐[γ(u)-ln(1+γ(u))]λ(du)dt+0.5xθ(t)(x0.5(t)-1)σ(t)dω(t)+𝕐[(1+γ(u))0.5-1]N~(dt,du)x0.5(t)-0.5𝕐ln(1+γ(u))N~(dt,du)=(x0.5+2θ-0.5r(t)+0.0625b2(t)-0.125b2(t)x0.5(t)+0.5r(t)x0.5(t)+0.0625b2(t)x(t)+0.5a(t)x(t)-0.5a(t)x1.5(t)+x2(t)+0.25σ2(t)x2θ(t)-0.125σ2(t)x0.5+2θ(t))dt+𝕐[(1+γ(u))0.5-1-0.5γ(u)]λ(du)x0.5dt+0.5𝕐[γ(u)-ln(1+γ(u))]λ(du)dt+0.5xθ(t)(x0.5(t)-1)σ(t)dω(t)+𝕐[(1+γ(u))0.5-1]N~(dt,du)x0.5(t)-0.5𝕐ln(1+γ(u))N~(dt,du)=F(x(t))dt+0.5xθ(t)(x0.5(t)-1)σ(t)dω(t)+𝕐[(1+γ(u))0.5-1]N~(dt,du)x0.5(t)-0.5𝕐[ln(1+γ(u))]N~(dt,du), where (9)F(x)=-0.5r(t)+0.0625b2(t)-0.125b2(t)x0.5+0.5r(t)x0.5+0.0625b2(t)x+0.5a(t)x-0.5a(t)x1.5+x2+0.25σ2(t)x2θ-0.125σ2(t)x0.5+2θ+𝕐[(1+γ(u))0.5-1-0.5γ(u)]λ(du)x0.5+0.5𝕐[γ(u)-ln(1+γ(u))]λ(du). From the inequality xq1+q(x-1) for x0,   0q1, and assumptions (A1) and (A3), it is easy to see that F(x) is bounded, say by K, in R+. We therefore obtain that (10)d[t-τtx2(s)ds+V(x(t))]Kdt+0.5xθ(t)(x0.5(t)-1)σ(t)dω(t)+𝕐[(1+γ(u))0.5-1]N~(dt,du)x0.5-0.5𝕐[ln(1+γ(u))]N~(dt,du). Integrating both sides from 0 to t, and then taking expectations, yields (11)E[t-τtx2(s)ds+V(x(t))]-τ0ξ2(s)ds+V(x(0))+Kt. Letting t=τkT, we obtain that EV(x(τkT))-τ0ξ2(s)ds+V(x(0))+KT. Note that for every ω{τkT}, x(τk,ω) equals either k or 1/k, and hence V(x(τk,ω)) is not less than either k-1-0.5log(k) or 1/k-1-0.5log(1/k)=1/k-1+0.5log(k). Consequently, (12)V(x(τk,ω))[k-1-0.5log(k)][1k-1+0.5log(k)]. It then follows from (11) that (13)-τ0ξ2(s)ds+V(x(0))+KTE[1{τkT}V(x(τk,ω))]P{τkT}([1k-1+0.5log(k)][k-1-0.5log(k)][1k-1+0.5log(k)]), where 1{τkT} is the indicator function of {τk}. Letting k+ gives limk+P{τkT}=0. Since T>0 is arbitrary, we have P{τ+<+}=0, and so P{τ+=+}=1 as required.

3. Permanence and Extinction for Model (<xref ref-type="disp-formula" rid="EEq1.3">3</xref>) Theorem 3.

Let assumptions (A1)–(A3) hold. If g*<0 and inftR¯+[a(t)-b(t+τ)]0; then the population x(t) represented by (3) goes to extinction a.s.

Proof.

Now applying Itô’s formula to (3) leads to (14)dt-τtb(s+τ)x(s)ds+dlnx(t)=(b(t+τ)x(t)-b(t)x(t-τ))dt+[r(t)-a(t)x(t)+b(t)x(t-τ)-σ2(t)x2θ(t)2+𝕐[ln(1+γ(u))-γ(u)]λ(du)σ2(t)x2θ(t)2]dt+σ(t)xθ(t)dω(t)+𝕐ln(1+γ(u))N~(dt,du). Then we have (15)t-τtb(s+τ)x(s)ds--τ0b(s+τ)x(s)ds+lnx(t)-lnx(0)=0t[r(s)-(a(s)-b(s+τ))x(s)-σ2(s)x2θ(s)2]ds-t𝕐[γ(u)-ln(1+γ(u))]λ(du)+M1(t)+0t𝕐ln(1+γ(u))N~(ds,du), where M1(t)=0tσ(s)xθ(s)dω(s). The quadratic variation of M1(t) is M1(t),M1(t)=0tσ2(s)x2θ(s)ds. By virtue of the exponential martingale inequality, for any positive constants T0, α, and β, we have (16)𝒫{sup0tT0[M1(t)-α2M1(t),M1(t)]>β}e-αβ. Choose T0=k, α=1, and β=2lnk. Then it follows that (17)𝒫{sup0tk[M1(t)-12M1(t),M1(t)]>2lnk}1k2. Making use of the Borel-Cantelli lemma yields that, for almost all ωΩ, there is a random integer k0=k0(ω) such that, for kk0, (18)sup0tk[M1(t)-12M1(t),M1(t)]2lnk. That is to say, M1(t)2lnk+(1/2)M1(t),M1(t)=2lnk+(1/2)0tσ2(s)x2θ(s)ds, for all 0tk, kk0 a.s. Substituting this inequality into (15), we can obtain that (19)lnx(t)-lnx(0)-τ0b(s+τ)x(s)ds+0t[r(s)-(a(s)-b(s+τ))x(s)]ds-t𝕐[γ(u)-ln(1+γ(u))]λ(du)+M1(t)+2lnk+0t𝕐ln(1+γ(u))N~(ds,du), for all 0tk, kk0 a.s. In other words, we have shown that for k-1tk, kk0 a.s., (20)t-1{lnx(t)-lnx(0)}t-1-τ0b(s+τ)x(s)ds+t-10t[r(s)-(a(s)-b(s+τ))x(s)]ds-𝕐[γ(u)-ln(1+γ(u))]λ(du)+2(k-1)-1lnk+t-1M1(t)+t-10t𝕐ln(1+γ(u))N~(ds,du). Define, for t0, M2(t)=0t𝕐ln(1+γ(u))N~(ds,du). Under assumption (A3), M2(t)=0t𝕐ln(1+γ(u))2λ(du)dsc2tλ(𝕐), by the strong law of large numbers for local martingales (see, e.g., ); we then obtain (21)limt+1t0t𝕐ln(1+γ(u))N~(ds,du)=0a.s. Taking superior limit on both sides of (20) and then making use of (21) yield limsupt+t-1lnx(t)g*. That is to say, if g*<0, one can see that limt+x(t)=0 a.s.

Definition 4 (see, e.g., Bao et al. [<xref ref-type="bibr" rid="B11">11</xref>]).

Population size x(t) is said to be stochastic permanence if, for arbitrary ε>0, there are constants β>0 and M>0 such that liminft+𝒫{x(t)β}1-ε and liminft+𝒫{x(t)M}1-ε.

Theorem 5.

Let assumptions (A1)–(A3) hold. If g*>0 and θ1, then the population x(t) modeled by (3) will be stochastic permanence.

Proof.

First, we prove that for arbitrary ε>0, there is constant M>0 such that liminft+𝒫{x(t)M}1-ε.

Let 0.5<p<1; we compute (22)dxp(t)=pxp-1(t)dx(t)+12p(p-1)xp-2(t)(dx(t))2+𝕐[(1+γ(u))p-1-pγ(u)]λ(du)xp(t)dt+𝕐[(1+γ(u))p-1]N~(dt,du)xp(t)=pxp-1(t)×[x1+θx(t)(r(t)-a(t)x(t)+b(t)x(t-τ))dt+σ(t)x1+θ(t)dω(t)]+12p(p-1)σ2(t)xp+2θ(t)dt+𝕐[(1+γ(u))p-1-pγ(u)]λ(du)xp(t)dt+𝕐[(1+γ(u))p-1]N~(dt,du)xp(t)[r(t)pxp(t)+p2b2(t)x2p(t)4+x2(t-τ)]dt+pσ(t)xp+θ(t)dω(t)-12p(1-p)σ2(t)xp+2θ(t)dt+𝕐[(1+γ(u))p-1-pγ(u)]λ(du)xp(t)dt+𝕐[(1+γ(u))p-1]N~(dt,du)xp(t)=F(x(t))dt-[ε1xp(t)+eε1τx2(t)-x2(t-τ)]dt+pσ(t)xp+2θ(t)dω(t)+𝕐[(1+γ(u))p-1]N~(dt,du)xp(t), where ε1 is a positive constant and (23)F(x)=eε1τx2+(ε1+r(t)p)xp+p2b2(t)x2p4-12p(1-p)σ2(t)xp+2θ+𝕐[(1+γ(u))p-1-pγ(u)]λ(du)xp. Making use of (6) in Lemma 1, we obtain 𝕐[(1+γ(u))p-1-pγ(u)]λ(du)xp0. In view of the inequality above, θ>0.75 and 0.5<p<1, we have that F(x) is bounded in R¯+; namely, M3=supxR¯+F(x)<+. Therefore (24)dxp(t)[M3-ε1xp(t)-eε1τx2(t)+x2(t-τ)]dt+pσ(t)xp+2θ(t)dω(t)+𝕐[(1+γ(u))p-1]N~(dt,du)xp(t). Once again by the Itô’s formula we have (25)d[eε1txp(t)]=eε1t[ε1xp(t)dt+dxp(t)]eε1t[M3-eε1τx2(t)+x2(t-τ)]+eε1tpσ(t)xp+2θ(t)dω(t)+eε1t(𝕐[(1+γ(u))p-1]N~(dt,du)xp(t)). We hence derive that (26)eε1tE[xp(t)]xp(0)+eε1tM3ε1-M3ε1-E0teε1s+ε1τx2(s)ds+E0teε1sx2(s-τ)ds=xp(0)+eε1tM3ε1-M3ε1-E0teε1s+ε1τx2(s)ds+E-τt-τeε1s+ε1τx2(s)dsxp(0)+eε1tM3ε1-M3ε1+-τ0eε1s+ε1τx2(s)ds. This implies immediately that limsupt+E[xp(t)]M3/ε1. Now, for any ε>0 and M=(M3/ε1)1/p/ε1/p, then by Chebyshev’s inequality, (27)𝒫{x(t)>M}=𝒫{xp(t)>Mp}E[xp(t)]Mp. Hence limsupt+𝒫{x(t)>M}ε. This implies liminft+𝒫{x(t)M}1-ε.

Next, we claim that for arbitrary ε>0, there is constant β>0 such that liminft+𝒫{x(t)β}1-ε.

Obviously, (28)limκ0+𝕐[1κ(1+γ(u))κ-1κ]λ(du)=𝕐ln(11+γ(u))λ(du)=-𝕐ln(1+γ(u))λ(du). Hence if g*>0, we can find a sufficiently small κ>0 such that (29)r(t)-𝕐γ(u)λ(du)-𝕐[1κ(1+γ(u))κ-1κ]λ(du)>0. Define y=1/x for x>0. Then, by Itô’s formula (see, e.g., [23, Theorem 2.5]), (30)dy(t)=y(t)[𝕐(11+γ(u)-1+γ(u))-r(t)+b(t)x(t-τ)+σ2(t)y-2θ+𝕐(11+γ(u)-1+γ(u))λ(du)]+a(t)dt-σ(t)y1-θdw(t)+y(t)𝕐(11+γ(u)-1)N~(dt,du). Define z=yκ, where κ satifies (29). In view of Itô’s formula, (31)dz(t)=κyκ-2(t)×{-y2(t)[(1κ(1+γ(u))κ-1κ)r(t)-𝕐γ(u)λ(du)-𝕐(1κ(1+γ(u))κ-1κ)λ(du)+b(t)x(t-τ)(1κ(1+γ(u))κ-1κ)]+a(t)y(t)+σ2(t)y2-2θ(t)+0.5σ2(t)(κ-1)y2-2θ(t)(1κ(1+γ(u))κ-1κ)}dt-σ(t)κyκ-θ(t)dw(t)+yκ(t)𝕐(1(1+γ(u))κ-1)N~(dt,du). Now, let η>0 be sufficiently small satisfying (32)r*-ε-𝕐γ(u)λ(du)-𝕐[1κ(1+γ(u))κ-1κ]λ(du)>ηκ. Define V=eηtz=eηtyκ. By virtue of Itô’s formula, (33)dV(t)=κeηtyκ-2(t)×{-y2(t)[(1κ(1+γ(u))κ-1κ)r(t)-𝕐γ(u)λ(du)-𝕐(1κ(1+γ(u))κ-1κ)λ(du)+b(t)x(t-τ)-ηκ(1κ(1+γ(u))κ-1κ)]+a(t)y(t)+0.5σ2(t)y2-2θ(t)+0.5σ2(t)κy2-2θ(t)(1κ(1+γ(u))κ-1κ)}dt-σ(t)κeηtyκ-θ(t)dw(t)+eηtyκ(t)𝕐(1(1+γ(u))κ-1)N~(dt,du)κeηtyκ-2(t){-y2(t)[(1κ(1+γ(u))κ-1κ)r*-ε-𝕐γ(u)λ(du)-𝕐(1κ(1+γ(u))κ-1κ)×λ(du)(1κ(1+γ(u))κ-1κ)]-ηκ+a(t)y(t)+σ2(t)y2-2θ(t)+0.5σ2(t)(κ-1)y2-2θ(t)(1κ(1+γ(u))κ-1κ)}dt-σ(t)κeηtyκ-θ(t)dw(t)+eηtyκ(t)𝕐(1(1+γ(u))κ-1)N~(dt,du)=eηtK(y(t))dt-σ(t)κeηtyκ-θ(t)dw(t)+eηtyκ(t)𝕐(1(1+γ(u))κ-1)N~(dt,du), for tT. Note that K(y) is upper bounded in R+; namely, G=supxR+K(y)<+. Consequently, (34)dV(t)Geηtdt-σ(t)κeηtxθ-κ(t)dw(t)+eηtyκ(t)𝕐(1(1+γ(u))κ-1)N~(dt,du), for sufficiently large t. Integrating both sides of the above inequality and then taking expectations give limsupt+E[x-κ(t)]G/η=M4. So for any ε>0, set β=ε1/κ/M41/κ. Then the desired assertion follows from the Chebyshev’s inequality . This completes the whole proof.

Remark 6.

Obviously, if assumptions (A1)–(A3) hold, θ1, inftR¯+[a(t)-b(t+τ)]0, limt+r(t) exists, and limt+r(t)0, then Theorems 3 and 5 establish the sufficient and necessary conditions for stochastic permanence and extinction of model (3).

Remark 7.

In line with g*=r*-𝕐[γ(u)-ln(1+γ(u))]λ(du) in Theorem 3 and g*=r*-𝕐[γ(u)-ln(1+γ(u))]λ(du) in Theorem 5, where 𝕐[γ(u)-ln(1+γ(u))]λ(du)>0 (see Lemma 2.2 in ), we found that the jump process exists considerable level of detriment to permanence and leads to the extinction of the population, which conforms to biological significance.

Remark 8.

If θ=1, τ=0, b(t)=0, and γ(u)=0 which means the jump process degenerates to zero, then our result about extinction and permanence coincide with the ones in paper . Moreover, in view of Theorem 5, we found that time delay has no impact on the permanence.

4. Examples and Numerical Simulations

In this section, we will use the Euler scheme (see, e.g., ) to illustrate the analytical findings.

Here, we choose  r(t)=0.2+0.01sint, a(t)=0.15, b(t)=0.1, σ2(t)=0.64, θ=0.8, 𝕐=(0,+), λ(𝕐)=1, ξ(t)=0.3et-0.5, t[0,0.5], and step size Δt=0.001. The only difference between conditions of Figures 1(a) and 1(b) is that the representations of γ(u) are different. In Figure 1(a), we choose γ(u)=0.7935, and then g*=-0.01<0. In view of Theorem 3, population x(t) will go to extinction. In Figure 1(b), we consider γ(u)=0.7045, and then g*=0.02>0. Making use of Theorem 5, the population x(t) will be stochastic permanence. By the numerical simulations, we can find that the jump process can affect the properties of the population model significantly.

The horizontal axis and the vertical axis in this represent the time t and the population size x(t) (step size Δt=0.001).

5. Conclusions and Remarks

In this paper, we investigate the permanence and extinction of a stochastic delay logistic model with jumps. Sufficient and necessary conditions for extinction are established as well as stochastic permanence.

Besides, some interesting topics deserve further consideration. One may propose some more realistic but complex models, such as introducing the colored noise into the model . Another significant problem is devoted to multidimensional stochastic delay model with jumps, and these investigations are in progress.

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Acknowledgments

This paper is supported by the National Natural Science Foundation of China (11271101), the NNSF of Shandong Province in China (ZR2010AQ021), and the Scientific Research Foundation of Harbin Institute of Technology at Weihai (HIT (WH) 201319).

Freedman H. I. Wu J. Periodic solutions of single-species models with periodic delay SIAM Journal on Mathematical Analysis 1992 23 689 701 Kuang Y. Smith H. L. Global stability for infinite delay Lotka-Volterra type systems Journal of Differential Equations 1993 103 2 221 246 2-s2.0-0001293981 10.1006/jdeq.1993.1048 He X. Gopalsamy K. Persistence, attractivity, and delay in facultative mutualism Journal of Mathematical Analysis and Applications 1997 215 1 154 173 2-s2.0-0031260789 10.1006/jmaa.1997.5632 Golpalsamy K. Stability and Oscillations in Delay Differential Equations of Population Dynamics 1992 Dordrecht, The Netherlands Kluwer Academic May R. M. Stability and Complexity in Model Ecosystems 1973 Princeton University Press Wei F. Wang K. The existence and uniqueness of the solution for stochastic functional differential equations with infinite delay Journal of Mathematical Analysis and Applications 2007 331 1 516 531 2-s2.0-33947105991 10.1016/j.jmaa.2006.09.020 Meng Q. Jiang H. Robust stochastic stability analysis of Markovian switching genetic regulatory networks with discrete and distributed delays Neurocomputing 2010 74 1–3 362 368 2-s2.0-78649470827 10.1016/j.neucom.2010.03.029 Li X. Gray A. Jiang D. Mao X. Sufficient and necessary conditions of stochastic permanence and extinction for stochastic logistic populations under regime switching Journal of Mathematical Analysis and Applications 2011 376 1 11 28 2-s2.0-78650055169 10.1016/j.jmaa.2010.10.053 Chunyan J. Daqing J. Ningzhong S. O'Regan D. Existence, uniqueness, stochastic persistence and global stability of positive solutions of the logistic equation with random perturbation Mathematical Methods in the Applied Sciences 2007 30 1 77 89 2-s2.0-33845594571 10.1002/mma.778 Liu M. Wang K. Population dynamical behavior of Lotka-Volterra cooperative systems with random perturbations Discrete and Continuous Dynamical Systems 2013 33 2495 2522 Bao J. Mao X. Yin G. Yuan C. Competitive LotkaVolterra population dynamics with jumps Nonlinear Analysis: Theory, Methods and Applications 2011 74 17 6601 6616 2-s2.0-80051597036 10.1016/j.na.2011.06.043 Bao J. Yuan C. Stochastic population dynamics driven by Lévy noise Journal of Mathematical Analysis and Applications 2012 391 2 363 375 2-s2.0-84858280723 10.1016/j.jmaa.2012.02.043 Liu M. Wang K. Dynamics of a Leslie-Gower Holling-type II predatorprey system with Levy jumps Nonlinear Analysis: Theory, Methods & Applications 2013 85 204 213 Liu M. Wang K. Stochastic Lotka-Volterra systems with Levy noise jumps Journal of Mathematical Analysis and Applications 2014 410 750 763 Applebaum D. Levy Processes and Stochastics Calculus 2009 2nd Cambridge University Press Situ R. Theory of Stochastic Differential Equations with Jumps and Applications 2005 Springer Jacob N. Wang Y. Yuan C. Stochastic differential delay equations with jumps, under nonlinear growth condition Stochastics 2009 81 571 588 Bao J. Yuan C. Comparison theorem for stochastic differential delay equations with jumps Acta Applicandae Mathematicae 2011 116 2 119 132 2-s2.0-80855138674 10.1007/s10440-011-9633-7 Yuan C. Mao X. Stability of stochastic delay hybrid systems with jumps European Journal of Control 2010 16 6 595 608 2-s2.0-79952156834 10.3166/ejc.16.595-608 Oksendal B. Sulem A. Applied Stochastic Control of Jump Diffusions 2005 Berlin, Germany Springer Ikeda N. Watanabe S. Stochastic Differential Equations and Diffusion Processes 1981 Amsterdam, The Netherlands North-Holland Publishing Company Lipster R. A strong law of large numbers for local martingales Stochastics 1980 3 217 228 Kunita H. Itô's stochastic calculus: Its surprising power for applications Stochastic Processes and their Applications 2010 120 5 622 652 2-s2.0-77949485245 10.1016/j.spa.2010.01.013 Liu M. Wang K. Analysis of a stochastic autonomous mutualism model Journal of Mathematical Analysis and Applications 2013 402 392 403 Li X. Mao X. Population dynamical behavior of non-autonomous Lotka-Volterra competitive system with random perturbation Discrete and Continuous Dynamical Systems 2009 24 2 523 545 2-s2.0-67650711514 Lu C. Ding X. H. Persistence and extinction for a stochastic logistic model with infinite delay Electronic Journal of Differential Equations 2013 262 1 16 Liu M. Wang K. Persistence and extinction in stochastic non-autonomous logistic systems Journal of Mathematical Analysis and Applications 2011 375 2 443 457 2-s2.0-78149415331 10.1016/j.jmaa.2010.09.058 Protter P. Talay D. The Euler scheme for Lévy driven stochastic differential equations The Annals of Probability 1997 25 1 393 423 2-s2.0-0031514913 Mao X. Yuan C. Stochastic Differential Equations with Markovian Switching 2006 Imperial College Press