Skip to main content

ORIGINAL RESEARCH article

Front. Phys., 27 July 2023
Sec. Social Physics
This article is part of the Research Topic Impacts of Heterogeneity on Biological Complex Systems View all 7 articles

Dynamics of an unstirred chemostat model with Beddington–DeAngelis functional response

  • School of Mathematics and Statistics, Shaanxi Normal University, Xi’an, China

This paper deals with an unstirred competitive chemostat model with the Beddington–DeAngelis functional response. With the help of the linear eigenvalue theory and the monotone dynamical system theory, we establish a relatively clear dynamic classification of this system in terms of the growth rates of two species. The results indicate that there exist several critical curves, which may classify the dynamics of this system into three scenarios: 1) extinction; 2) competitive exclusion; and 3) coexistence. Comparing with the classical chemostat model [26], our theoretical results reveal that under the weak–strong competition cases, the role of intraspecific competition can lead to species coexistence. Moreover, the simulations suggest that under different competitive cases, coexistence can occur for suitably small diffusion rates and some intermediate diffusion rates. These new phenomena indicate that the intraspecific competition and diffusion have a great influence on the dynamics of the unstirred chemostat model of two species competing with the Beddington–DeAngelis functional response.

1 Introduction

It is well known that the chemostat is a laboratory apparatus used for the continuous culture of microorganisms, while the chemostat models are extensively applied in ecology to simulate the growth of single-celled algal plankton in oceans and lakes [14]. Most of the earlier chemostat models assume the well-stirring of culture, which leads to chemostat models generally described by ordinary differential equation models (see, e.g., [2, 4, 5]). However, this idealized mixing is quite different from the real environment in which microbial populations live. Since the ability of microorganisms to move in a random fashion plays an important role in determining the survival and extinction of populations, many unstirred chemostat models have sprung up in which populations and resources are distributed in spatially variable habitats; please refer to [610] for small sampling of such works.

There are various types of response functions; among them, Holling types I–IV [11] are usually introduced to model the growth of microorganisms. Particularly, the various chemostat models with Holling type II functional response have been extensively studied (see, e.g., [4, 10, 12, 13]). As far as we know, for the unstirred competitive chemostat models with Holling type II functional response, So and Waltman [14] first obtained the local coexistence by standard bifurcation theorems. Later, Hsu and Waltman [6] obtained the asymptotic behavior of solutions by the theory of uniform persistence in an infinite-dimensional dynamical system and the theory of strongly order-preserving semi-dynamical system. To explore the effect of diffusion, Shi et al. [8] further studied this model and confirmed that stable coexistence solutions only occur at the intermediate diffusion rates. In addition, a diffusive predator–prey chemostat model with Holling type II functional response was studied by Nie et al. [7], and their analytical and numerical results show that a relatively small diffusion is conducive to the coexistence of species.

However, in nature, it is known that there is not only competition between two species but also mutual interference in species. Therefore, it is necessary to consider mutual interference in species. To this end, Beddington [15] and DeAngelis et al. [16] (simplified as B.–D.) proposed the following B.–D. functional response:

f1S,u=Sk1+S+β1u,f2S,v=Sk2+S+β2v,

where ki > 0 (i = 1, 2) are the Michaelis–Menten constants, S represents the density of the resources, u and v represent the density of two species, respectively, and βi > 0 (i = 1, 2) model the mutual interference between two species.

As illustrated by Harrision [17], the B.–D. functional response with intraspecific interference competition was superior to well-known Holling type II functional response in modeling the resource uptake of species. Therefore, there appear successively many works to describe the population dynamics by using the B.–D. functional response. For instance, Jiang et al. [18] discussed a competition model with the B.–D. functional response, and they applied the fixed-point index theory to obtain the sufficient conditions for the existence of positive solutions. In addition, a predator–prey model with a heterogeneous environment and the B.–D. functional response was constructed by Zhang and Wang [19], and the existence of positive stationary solutions was obtained by using the fixed-point index theory. We also refer the recent works [2022] about population models with the B.–D. functional response.

Particularly, the unstirred chemostat models with the B.–D. functional response have also received considerable attention in the past decades. Wang et al. [23] obtained the sufficient conditions for the existence of positive steady-state solutions and studied the effect of parameter β1 on coexistence states by the fixed-point index theory, the perturbation technique, and the bifurcation theory. Meanwhile, Nie and Wu [24] studied the unstirred chemostat model with the B.–D. functional response and inhibitor, and the uniqueness, multiplicity, and stability of the coexistence solutions were obtained by the degree theory in cones, bifurcation theory, and perturbation technique. More works on chemostat models with the B.–D. functional response can be found in [2527] and the references therein.

Mathematically speaking, these sufficient conditions for the existence of coexistence solutions are usually established in terms of the principal eigenvalues of the corresponding linearized eigenvalue problems at trivial or semi-trivial steady states (see, e.g., [18, 19, 23, 24]). It is worth noting that these principal eigenvalues depend heavily on the model parameters, which motivates us to explore how these model parameters affect the existence of coexistence solutions and establish the dynamics classification of this system in terms of these model parameters. Moreover, it should be noted that studying the asymptotic analysis of steady states of chemostat models is non-trivial, and some new techniques need to be introduced. Overall, for the unstirred chemostat system with the B.–D. functional response, we are concerned with the following questions:

(1) How do parameters such as diffusion rates, growth rates, and intraspecific competition parameters affect the dynamics of the unstirred chemostat system with the B.–D. functional response?

(2) Can we establish a clear dynamic classification of the unstirred chemostat system with the B.–D. functional response in terms of these parameters?

(3) Will there arise a new phenomenon if one introduces the B.–D. functional response into the unstirred chemostat model?

The purpose of this paper is to address these problems. We hope that the approaches in this paper might provide some new insights on the dynamical behavior of the unstirred chemostat models.

This paper is organized as follows. In Section 2, we introduce an unstirred chemostat model with the B.–D. functional response and its corresponding limiting system. In Section 3, some preliminary results are given. In Section 4, we aim to investigate the dynamics of this limiting system and obtain a relatively clear dynamic classification of this limiting system in the m1m2 plane. In Section 5, the coexistence solution for this limiting system is established by a bifurcation argument. In Section 6, we study the effect of diffusion on system dynamics by numerical approaches. In Section 7, a discussion is presented from the opinion of analytic and numerical results. Finally, the proofs of some theoretical results are deferred to the Supplementary Appendix in Supplementary Section S8.

2 The model

In this paper, we consider following the unstirred chemostat system with the B.–D. functional response:

St=dSxxm1uf1S,um2vf2S,v,x0,1,t>0,ut=duxx+m1uf1S,u,x0,1,t>0,vt=dvxx+m2vf2S,v,x0,1,t>0,Sx0,t=S0,Sx1,t+γS1,t=0,t>0,ux0,t=ux1,t+γu1,t=0,t>0,vx0,t=vx1,t+γv1,t=0,t>0,Sx,0=S0x0,x0,1,ux,0=u0x,0,vx,0=v0x,0,x0,1,

where S(x, t) is the concentration of the nutrient and u(x, t) and v(x, t) represent the population density for the two competing microorganisms with location x and time t, respectively. The positive constants m1 and m2 are corresponding to the growth rates of species u and v with nutrient concentration S. d > 0 is the diffusion rate of the nutrient and microorganisms. The initial data S0(x), u0(x), and v0(x) are non-negative non-trivial continuous functions. In the reactor, the nutrients are pumped with the rate of S0 > 0 at position x = 0, and the mixed cultures containing nutrients and microorganisms are pumped out with the rate of γ > 0 at the position x = 1, which results in the Robin boundary conditions at x = 1 [6]. Here, f1(S, u), f2(S, v) satisfying Eq. 1.1 are the nutrient uptake of species u and v at nutrient concentration S. Moreover, we redefine f1(S, u), f2(S, v) as follows [10]:

f̂1S,u=f1S,u,S0,u0,0,others,f̂2S,v=f2S,v,S0,v0,0,others.

For convenience, we still denote f̂i(S,u)(i=1,2) as fi(S, u), throughout this paper.

It is worth pointing out that system (2.1) satisfies the conservation law [4]. In other words, the total biomass concentration S + u + v in the chemostat approaches asymptotically a steady state ϕ(x)=S0(1+γγx) (see [14], Lemma 2.1); that is,

limtSx,t+ux,t+vx,t=ϕxuniformlyforx0,1.

Hence, we apply the classical internal chain transitive theory [[28], Lemma 2.1] to reduce system (2.1) into the following limiting system:

ut=duxx+m1f1ϕxuv,uu,x0,1,t>0,vt=dvxx+m2f2ϕxuv,vv,x0,1,t>0,ux0,t=ux1,t+γu1,t=0,vx0,t=vx1,t+γv1,t=0,t>0,ux,0=u0x,0,vx,0=v0x,0,u0x+v0xϕxx0,1.

In this paper, we are mainly concerned with the dynamics classification of system (2.2). Based on the competition relationship of two species, system (2.2) generates a strictly monotone dynamical system in the partial competitive order induced by the cone K = {(u, v) ∈ C[0, 1] × C[0, 1]: u ≥ 0, v ≤ 0} (see [27], Proposition 1.3 in Chapter 8). Since the dynamics of Eq. 2.2 are related to the stability of non-negative steady states [29], we also focus on the following steady-state system:

duxx+m1f1ϕxuv,uu=0,x0,1,dvxx+m2f2ϕxuv,vv=0,x0,1,ux0=ux1+γu1=0,vx0=vx1+γv1=0.

The contribution of this paper is to explore the effect of these model parameters on the dynamics of system (2.2). Precisely, we first apply the linear eigenvalue theory and the monotone dynamical system theory to establish the threshold dynamics of system (2.2) in terms of growth rates and intraspecific competition parameters (see Theorems 4.1, 4.2). Moreover, we give a relatively clear dynamic classification of system (2.2) in the m1m2 plane (Figure 2). Finally, by numerical simulations, we further investigate the effect of diffusion on the dynamics of system (2.2) (Figures 35). Particularly, the numerical results show that under the different competitive cases, coexistence occurs for suitably small diffusion rates and some intermediate diffusion rates, which reveals that the dynamics of system (2.2) are relatively complicated.

3 Preliminaries

In this section, some preliminary results are presented, which are helpful in the following analysis.

We first consider a linear eigenvalue problem

dφxx+qxφ=μφ,x0,1,φx0=ϕx1+γφ1=0,

where γ is a positive constant and q(x) ∈ C[0, 1]. For fixed d > 0, it is well known that problem (3.1) admits a principal eigenvalue μ1(q(x)) [29], which corresponds to a positive eigenfunction φ1(⋅, q(x)) normalized by maxx[0,1]φ1=1. Furthermore, by the variation characterization of the principal eigenvalue [29], we have

μ1qx=infφH10,1,φ0d01φx2dx+dγφ2101qxφ2dx01φ2dx.

Moreover, the principal eigenvalue μ1(q(x)) has the following properties.

Lemma 3.1 (See [21], Lemma 2.1). The following statements on the principal eigenvalue μ1(q(x)) are true:

(i) μ1(q(x)) depends continuously and differentially on parameter d in (0, + ), and it is strictly decreasing with respect to d in (0, + ).

(ii) qn(x) → q(x) in C[0, 1] implies μ1(qn(x)) → μ1(q(x)).

(iii) q1(x) ≥ q2(x) implies that μ1(q1(x)) ≥ μ1(q2(x)), and the equality holds only if q1(x) ≡ q2(x). Particularly, μ1(0) < 0.

We consider the following single-species model:

ωt=dωxx+mfϕω,ωω,x0,1,t>0,ωx0,t=ωx1,t+γω1,t=0,t>0,ωx,0=ω0x,0,x0,1,

where d, m > 0 are constants and f(ϕω,ω)=ϕωk+ϕ+(β1)ω. For fixed d, k > 0, let μ1(mf(ϕ, 0)) be the principal eigenvalue of

dφxx+mfϕ,0φ=μφ,x0,1,t>0,φx0=φx1+γφ1=0,

where f(ϕ,0)=ϕk+ϕ. Then, we can conclude from Lemma 3.1(iii) that μ1(mf(ϕ, 0)) is strictly increasing with respect to m in (0, + ). Moreover, limm0+μ1(mf(ϕ,0))=μ1(0)<0 by Lemma 3.1(iii), and limm+μ1(mf(ϕ,0))=+ by Eq. 3.2. For fixed d, k > 0, there exists a unique critical value m such that

μ1mfϕ,0<0if0<m<m,μ1mfϕ,0=0ifm=m,μ1mfϕ,0>0ifm>m.

To stress the dependence of the unique positive steady state of system (3.3) on m and β, let us denote it by ω(⋅; m, β).

Lemma 3.2 Suppose d, m, β, k > 0. Let ω(x, t) be the solution of system (3.3). Then,

(i) if m > m, system (3.3) admits a unique positive steady state 0 < ω(⋅; m, β) < ϕ(x) for x ∈ [0, 1], and limtω(x,t)=ω(;m,β) uniformly on [0,1];

(ii) if mm, system (3.3) has no positive steady state and limtω(x,t)=0 uniformly on [0,1].

The proof of Lemma 3.2 is similar to the arguments in [6], Theorem 3.2. So, we omit it here.

We next give some asymptotic properties of the unique positive steady state ω(⋅; m, β) of system (3.3) by taking m and β as the variable parameters.

Lemma 3.3 suppose that m > m holds. The following statements about the positive solution ω(; m, β) will hold.

(i) For fixed d, k, β > 0, there exists positive solution ω(; m, β), which is continuously differentiable with respect to m in (m, + ∞), and it is point-wise strictly increasing in m ∈ (m, + ∞). Moreover,

limmm+ω;m,β=0,limm+ω;m,β=ϕxuniformlyon0,1.

(ii) For fixed d, k > 0 and m > m, there exists positive solution ω(⋅; m, β), which is continuously differentiable with respect to β in (0, + ), and it is point-wise strictly decreasing in β ∈ (0, + ). Moreover,

limβ+ω;m,β=0uniformlyon0,1.

Proof For (i), it follows from Lemma 3.2 that ω(⋅; m, β) exists if and only if m > m. Moreover, ω(⋅; m, β) is continuously differentiable with respect to m in (m, + ) refering to the arguments in [30], Lemma 5.4(ii). Differentiating the equation of ω(⋅; m, β) with respect to m and denoting Pm(x) = ∂ω(⋅; m, β)/∂m, Pm(x) satisfies

dPm+mfϕω,ωk+βϕωk+ϕ+β1ω2Pm=fϕω,ωω,x0,1,Pm0=Pm1+γPm1=0.

We define X={ψC2[0,1]:ψ(0)=ψ(1)+γψ(1)=0} and denote L1(ψ)=dψ+m[f(ϕω,ω)(k+βϕ)ω[k+ϕ+(β1)ω]2]ψ. It is easy to see that μ1(L1) < μ1(mf(ϕω, ω)) = 0 by Lemma 3.1(iii). Noting that Pm(x)X and L1(Pm) = −f(ϕω, ω)ω < 0, we have Pm(x) > 0 on [0,1] by the generalized maximum principle [23, Theorem 5], which implies that ω(⋅; m, β) is point-wise strictly increasing in m ∈ (m, + ).

Since 0 < ω(⋅; m, β) < ϕ and ω=mdf(ϕω,ω)ω<0, ω is decreasing on [0,1]. Note that the boundary conditions ω(0)=0,ω(1)=γω(1). Then, ω(x) is uniformly bounded for x ∈ [0, 1]. It follows from the Arzela–Ascoli theorem that there exist ω1, ω2C[0, 1] with 0 ≤ ω1ϕ and 0 ≤ ω2ϕ such that

limmm+ω;m,β=ω1,limm+ω;m,β=ω2inC0,1.

To prove ω1 = 0 on [0,1], we assume by contradiction that ω1≢0 on [0,1]. Since 0 < f(ϕω, ω) < 1, the standard Lp-estimate implies that ω(⋅; m, β) is uniformly bounded in W2,p(0, 1) with p ∈ (1, ) for m ∈ (m, M], where M is a fixed constant larger than m. Therefore, limm(m)+ω(;m,β)=ω1 weakly in W2, p (0, 1) and the convergence also holds in C1[0, 1] by the Sobolev embedding theorem. Then, ω1 satisfies

dω1+mfϕω1,ω1ω1=0,x0,1,ω10=ω11+γω11=0.

Since ω1≢0 on [0,1], we have ω1 > 0 on [0,1] by the strong maximum principle. It is easy to see that μ1(mf(ϕ, 0)) > μ1(mf(ϕω1, ω1)) = 0 by Lemma 3.1(iii), a contradiction to the definition of m (Eq. 3.4). Thus, ω1 = 0.

We next prove ω2 = ϕ(x) on [0,1]. We recall that ω(⋅; m, β) satisfies

dω+mfϕω,ωω=0,x0,1,ω0=ω1+γω1=0.

Dividing the first equation of Eq. 3.9 by and integrating over (0,1),

dm01|ω|2ω2dxdγm+01fϕω,ωdx=0,

which implies

001fϕω,ωdxdγm.

Note limm+ω(;m,β)=ω2 in C[0, 1]. Taking m → +, we have 01f(ϕω2,ω2)dx=0, which means ω2 = ϕ(x) on [0,1] by 0 ≤ ω2ϕ.

(ii) The monotonicity of ω(⋅; m, β) with respect to β in (0, + ) can be proved by the similar arguments as in the proof of (i) and limβ+ω(;m,β)=0uniformlyon[0,1] holds (see [23], Remark 1.2).

It is clear that system (2.2) generates a monotone dynamical system in the partial competitive order induced by the cone K = {(u, v) ∈ C[0, 1] × C[0, 1]: u ≥ 0, v ≤ 0} (see [27], Proposition 1.3 in Chapter 8). Hence, we can recall the well-known results on the monotone dynamical system as follows.

Lemma 3.4 [9]. For the monotone dynamical system,

(i) if two semi-trivial steady states are asymptotically stable, then it has at least one unstable coexistence steady state.

(ii) if two semi-trivial steady states are unstable, then it has at least one stable coexistence steady state. Furthermore, if its coexistence steady states are all linearly stable, then there is a unique coexistence steady state that is globally asymptotically stable.

(iii) if there is no coexistence steady state and if one semi-trivial solution is linearly unstable, the other semi-trivial solution is globally asymptotically stable.

4 The dynamics analysis of system (2.2)

As we already know, the local dynamics of system (2.2) are related to the stability of semi-trivial solutions [29]. Hence, we next establish the stability of semi-trivial solutions, including local and some global stability results. Recalling fi(ϕ,0)=ϕki+ϕ(i=1,2), by the similar arguments as in (3.4), we can define mi such that

μ1mifiϕ,0<0if0<mi<mi,μ1mifiϕ,0=0ifmi=mi,μ1mifiϕ,0>0ifmi>mi.

Clearly, mi(i=1,2) are dependent on ki but not on βi.

Proposition 4.1 For fixed d > 0, the following statements hold:

(i)k2k1<m2m1<1ifk1>k2>0;(ii)1<m2m1<k2k1ifk2>k1>0;(iii)m1=m2ifk1=k2>0.

Proof (i) Since

μ1m1ϕk1+ϕ=0andμ1m2ϕk2+ϕ=0,

it is easy to check that m1>m2 by k1 > k2 > 0 and Lemma 3.1(iii). Note that Eq. 4.2) is equivalent to μ1(m1ϕ/k11+ϕ/k1)=0 and μ1(m2ϕ/k21+ϕ/k2)=0. Since ϕ1+ϕ/k1>ϕ1+ϕ/k2 based on k1 > k2 > 0, we have m1k1<m2k2; that is, k2k1<m2m1. Therefore, k2k1<m2m1<1. (ii) can be obtained similarly. For (iii), it is easy to obtain m1=m2 by the fact of k1 = k2 > 0 and μ1(mifi(ϕ,0))=0.

As the consequence of Lemma 3.2, system (2.2) admits the following trivial and semi-trivial solutions: trivial solution (0,0); semi-trivial solution (ω(⋅; m1, β1), 0) exists if and only if m1>m1; semi-trivial solution (0, ω(⋅; m2, β2)) exists if and only if m2>m2. For convenience, we denote û=ω(;m1,β1), v̂=ω(;m2,β2), and next, we give an a priori estimate for the positive steady-state solution of system (2.2).

Lemma 4.1 Suppose that (u(x), v(x)) is a non-negative solution of system (2.2) with u≢0 and v≢0 on [0,1]. Then,

(i) 0<u(x)û and 0<v(x)v̂ for x ∈ [0, 1]

(ii) 0 < u(x) + v(x) < ϕ(x) for x ∈ [0, 1]

The proof of Lemma 4.1 is exactly similar to that in [10], Lemma 4.2; hence, it is omitted here.

We next establish the linear stability of (û,0) and (0,v̂). First, the linearized operator of system (2.3) at (û,0) is given by

dφxx+m1f1ϕû,û+ûf1uϕû,ûφ+m1ûf1vϕû,ûψ=μφ,x0,1,dψxx+m2f2ϕû,0ψ=μψ,x0,1,φx0=φx1+γφ1=0,ψx0=ψx1+γψ1=0,

where f1u(ϕû,û)=k1+β1ϕ[k1+ϕ+(β11)û]2<0,f1v(ϕû,û)=k1+β1û[k1+ϕ+(β11)û]2<0. By the Riesz–Schauder theory, the eigenvalues of Eq. 4.3 consist of the eigenvalues of the following two operators:

B1m1=dd2dx+m1f1ϕû,û+ûf1uϕû,û,B2m2=dd2dx+m2f2ϕû,0,

subject to the corresponding boundary conditions. It follows from Lemma 3.1 (iii) that μ1(m1(f1(ϕû,û)+ûf1u(ϕû,û)))<μ1(m1f1(ϕû,û)). Moreover, μ1(m1f1(ϕû,û))=0 with eigenfunction φ1(m1f1(ϕû,û))=û, which implies μ1(m1f1(ϕû,û)+ûf1u(ϕû,û))<0. Hence, the stability of (û,0) is determined by the sign of the principal eigenvalue μ1(m2f2(ϕû,0)) of B2(m2). More precisely, (û,0) is asymptotically stable if μ1(m2f2(ϕû,0))<0, while (û,0) is unstable if μ1(m2f2(ϕû,0))>0.

The linearized operator of system (2.3) at (0,v̂) is given by

dφxx+m1f1ϕv̂,0φ=μφ,x0,1,dψxx+m2f2ϕv̂,v̂+v̂f2vϕv̂,v̂ψ+m2v̂f2uϕv̂,v̂φ=μψ,x0,1,φx0=φx1+γφ1=0,ψx0=ψx1+γψ1=0,

where f2u(ϕv̂,v̂)=k2+β2v̂[k2+ϕ+(β21)v̂]2<0,f2v(ϕv̂,v̂)=k2+β2ϕ[k2+ϕ+(β21)v̂]2<0. We denote

B3m2=dd2dx+m1f1ϕv̂,0,B4m2=dd2dx+m2f2ϕv̂,v̂+v̂f2vϕv̂,v̂.

Similarly, (0,v̂) is asymptotically stable if μ1(m1f1(ϕv̂,0))<0, while (0,v̂) is unstable if μ1(m1f1(ϕv̂,0))>0.

Theorem 4.1 We consider d, k1, k2 > 0 fixed. Let (u(x, t), v(x, t)) be the solution of (2.2) with any non-negative non-trivial initial condition. The following statements hold:

(i) We consider β1, β2 > 0 fixed.

(i.1) If m1m1 and m2m2, then

limt+ux,t=0,limt+vx,t=0uniformly on0,1.

(i.2) If m1m1 and m2>m2, then

limt+ux,t=0,limt+vx,t=v̂uniformly on0,1.

(i.3) If m1>m1 and m2m2, then

limt+ux,t=û,limt+vx,t=0uniformly on0,1.

(ii) We consider m1>m1, m2>max{m2,max{k2k1,1}m1}, and β1 > 0 fixed. Then, (û,0) is unstable. Moreover,

(ii.1) (4.8) holds provided

0<β2β20m2k1m1k2γm1S01+γ,

(ii.2) there exists a unique β2(β20,+) such that (0,v̂) is asymptotically stable when 0<β2<β2; (0,v̂) is unstable when β2>β2, and system (2.2) admits at least one stable coexistence steady state when β2>β2.

(iii) We consider m2>m2, m1>max{m1,max{k1k2,1}m2}, and β2 > 0 fixed. Then, (0,v̂) is unstable. Moreover,

(iii.1) Eq. 4.9 holds provided

0<β1β10m1k2m2k1γm2S01+γ,

(iii.2) there exists a unique β1(β10,+) such that (û,0) is asymptotically stable when 0<β1<β1; (û,0) is unstable when β1>β1, and system (2.2) admits at least one stable coexistence steady state when β1>β1.

Proof (i) can be proved by the similar arguments as in [6], Theorems 3.5, 3.6, and we omit it here. Next, we only prove (ii), since (iii) can be proved by similar arguments.

Claim 1. For m1>m1, m2max{k2k1,1}m1, and β1 > 0 fixed, (û,0) is unstable.

Note that û satisfies

dûxx+m1f1ϕû,ûû=0,x0,1,ûx0=ûx1+γû1=0,

which implies μ1(m1f1(ϕû,û))=0. We recall that (û,0) is asymptotically stable if μ1(m2f2(ϕû,0))<0 and it is unstable if μ1(m2f2(ϕû,0))>0. Then, we conclude from m2max{k2k1,1}m1 and 0<û<ϕ on [0,1] that

m1f1ϕû,ûm2f2ϕû,0<m1f1ϕû,0m2f2ϕû,0=m1k2m2k1+m1m2ϕûk1+ϕûk2+ϕûϕû0,,

which means that μ1(m2f2(ϕû,0))>μ1(m1f1(ϕû,û))=0 by Lemma 3.2(iii). That is, (û,0) is unstable.

Claim 2. (1) For m1>m1, m2>max{m2,max{k2k1,1}m1}, and β1 > 0 fixed, (0,v̂) is asymptotically stable when 0<β2β20, where β20 is defined by Eq. 4.10.

(2) There exists a unique β2>β20 such that (0,v̂) is asymptotically stable when 0<β2<β2, and (0,v̂) is unstable when β2>β2.

For (1), we recall that (0,v̂) is asymptotically stable if μ1(m1f1(ϕv̂,0))<0 and unstable if μ1(m1f1(ϕv̂,0))>0. Similarly, we can conclude from m2>max{m2,max{k2k1,1}m1} and 0<v̂<ϕ(x)<S0(1+γ)γ on [0,1] that

m1f1ϕv̂,0m2f2ϕv̂,v̂=m1k2m2k1+m1m2ϕv̂+m1β2v̂k2+ϕ+β21v̂k1+ϕv̂ϕv̂<0,

provided 0<β2β20(m2k1m1k2)γm1S0(1+γ). It follows from Lemma 3.2(iii) that μ1(m1f1(ϕv̂,0))<μ1(m2f2(ϕv̂,v̂))=0 when 0<β2β20. That is, (0,v̂) is asymptotically stable when 0<β2β20.

For (2), since v̂(;m2,β2) is point-wise strictly decreasing in β2 ∈ (0, + ) and limβ2+v̂(;m2,β2)=0 uniformly on [0,1] (see Lemma 3.3(ii)), we can obtain from Lemma 3.2(iii) that μ1(m1f1(ϕv̂(;m2,β2),0)) is strictly increasing in β2 ∈ (0, + ). Furthermore,

limβ2+μ1m1f1ϕv̂;m2,β2,0=μ1m1f1ϕ,0>0,

based on m1>m1 (Eq. 4.1). Moreover, μ1(m1f1(ϕv̂(;m2,β2),0))<0 when 0<β2β20. Therefore, there exists a unique β2>β20 such that

μ1m1f1ϕv̂;m2,β2,0<0if0<β2<β2,=0ifβ2=β2,>0ifβ2>β2,

which means that (0,v̂) is asymptotically stable when 0<β2<β2, while it is unstable when β2>β2.

Claim 3. For m1>m1, m2>max{m2,max{k2k1,1}m1}, and β1 > 0 fixed, system (2.2) has no positive steady states when 0<β2β20.

We assume by contradiction that system (2.2) admits a positive steady state (ū,v̄), which satisfies

dūxx+m1f1ϕūv̄,ūū=0,x0,1,dv̄xx+m2f2ϕūv̄,v̄v̄=0,x0,1,ūx0=ūx1+γū1=0,v̄x0=v̄x1+γv̄1=0.

Multiplying the first equation of (4.13) by v̄ and the second equation by ū, integrating over (0,1), and then, subtracting the resulting equations, we have

01m1k2m2k1+m1m2ϕūv̄+m1β2v̄m2β1ūk1+ϕ+β11ūv̄k2+ϕ+β21v̄ūūv̄ϕūv̄dx=0.

Since ū+v̄<ϕ(x)<S0(1+γ)γ on [0,1], we can conclude from m1>m1 and m2>max{m2,max{k2k1,1}m1} that the left side of (4.14) is less than 0, when 0<β2β20(m2k1m1k2)γm1S0(1+γ). That is a contradiction.

In conclusion, we can deduce that (ii.1) holds from Claim 1, Claim 2(1), Claim 3, and Lemma 3.4(iii). In addition, (ii.2) is the direct result of Claim 1, Claim 2(2), and Lemma 3.4(ii). The proof is completed.

Remark 4.1 Theorem 4.1(i) implies that both species with sufficiently small growth rates are washed out, while competition exclusion occurs and the species with a sufficiently faster growth rate will finally win the competition. In particular, when both species admit sufficient fast growth rates, Theorem 4.1(ii.1) suggests that the species v with stronger growth ability (m2k2 is large) and weaker intraspecific competition (β2 is small) will finally win the competition. This is consistent with the biological intuition that the species with stronger growth ability and weaker intraspecific competition has more competitive advantages. Theorem 4.1(iii.1) illustrates the similar biological phenomenon.

We next investigate the local dynamics of system (2.2). Note that the stability of (û(m1,β1),0) is determined by the sign of μ1(m2f2(ϕû(m1,β1),0)), and the stability of (0,v̂(m2,β2)) is determined by the sign of μ1(m1f1(ϕv̂(m1,β1),0)). Clearly, μ1(m2f2(ϕû(m1,β1),0)) depends on m1, m2, and β1, and μ1(m1f1(ϕv̂(m2,β2),0)) depends on m1, m2, and β2. To this end, we define

σ1m1,m2,β1μ1m2f2ϕûm1,β1,0form1>m1,m2>0,β1>0,
τ1m1,m2,β2μ1m1f1ϕv̂m2,β2,0form1>0,m2>m2,β2>0.

Lemma 4.2 The principal eigenvalues σ1(m1, m2, β1) and τ1(m1, m2, β2) have the following properties:

(i) For fixed d, k1, k2 > 0 and m1>m1,

(i.1) σ1(m1, m2, β1) is strictly decreasing with respect to m1 in (m1,+),

(i.2) σ1(m1, m2, β1) is strictly increasing with respect to β1 in (0, + ),

(i.3) σ1(m1, m2, β1) is strictly increasing with respect to m2 in (m2,+); moreover,

limm2m2+σ1m1,m2,β1=μ1m2f2ϕû,0<0,limm2+σ1m1,m2,β1=+.

(ii) For fixed d, k1, k2 > 0 and m2>m2,

(ii.1) τ1(m1, m2, β2) is strictly increasing with respect to m1 in (m1,+),

(ii.2) τ1(m1, m2, β2) is strictly increasing with respect to β2 in (0, + ),

(ii.3) τ1(m1, m2, β2) is strictly decreasing with respect to m2 in (m2,+); moreover,

limm2m2+τ1m1,m2,β2=μ1m1f1ϕ,0>0,limm2+τ1m1,m2,β2=μ10<0.

Proof For (i), (i.1) can be obtained by Lemma 3.1(iii) and Lemma 3.3(i). Similarly, (i.2) is followed by Lemma 3.1(iii) and Lemma 3.3(ii). To prove (i.3), it is obvious that σ1(m1, m2, β1) is strictly increasing with respect to m2 in (m2,+) by Lemma 3.1(iii) and limm2+σ1(m1,m2,β1)=+ by (3.2). We recall μ1(m2f2(ϕ,0))=0. Then, we can conclude from Lemma 3.1(ii) (iii) that

limm2m2+σ1m1,m2,β1=μ1m2f2ϕû,0<μ1m2f2ϕ,0=0.

For (ii), (ii.1) can be obtained by Lemma 3.1(iii), and (ii.2) can be proved by Lemma 3.1(iii) and Lemma 3.3(ii). We then prove (ii.3). Since v̂(m2,β2) is point-wise strictly increasing in m2(m2,+) by Lemma 3.3(i), it follows from Lemma 3.1(iii) that τ1(m1, m2, β2) is strictly decreasing with respect to m2 in (m2,+). Noting that limm2(m2)+v̂(m2,β2)=0 by (3.5), we can conclude from Lemma 3.1(ii) that

limm2m2+τ1m1,m2,β2=μ1m1f1ϕ,0>0,

based on m1>m1 (see (4.1)). Moreover, since limm2+v̂(m2,β2)=ϕ(x) on [0,1] (see (3.5)), we can obtain from Lemma 3.1(ii) (iii) that

limm2+τ1m1,m2,β2=μ1m1f1ϕv̂m2,β2,0=μ10<0.

The proof is completed.

Clearly, both σ1(m1, m2, β1) and τ1(m1, m2, β2) depend on m1 and m2. To investigate the local dynamics of system (2.2) in the m1m2 plane, we fix β1, β2 > 0 and denote them by σ1(m1, m2) and τ1(m1, m2).

Lemma 4.3 Suppose mi>mi(i=1,2). For fixed d, β1, β2, k1, k2 > 0, there exist two continuous critical curves

Γ1:m2=m̄2m1form1m1,+,
Γ2:m2=m̃2m1form1m1,+,

where m̄2(m1) and m̃2(m1) are differentially dependent on m1 and uniquely determined by

μ1m̄2m1f2ϕû,0=0andμ1m1f1ϕv̂;m̃2m1,0=0,

respectively. Then,

(i) the semi-trivial steady state (û,0) is locally asymptotically stable if (m1,m2)(m1,+)×(m2,m̄2(m1)), neutrally stable if (m1,m2)(m1,+)×{m̄2(m1)}, and unstable if.

(ii) the semi-trivial steady state (0,v̂) is locally asymptotically stable if (m1,m2)(m1,+)×{m̃2(m1),+}, neutrally stable if (m1,m2)(m1,+)×{m̃2(m1)}, and unstable if (m1,m2)(m1,+)×(m2,m̃2(m1)).

Proof (i) By Lemma 4.2(i), we conclude that for any m1(m1,+) given, there exists a unique m̄2(m1)>m2 such that

σ1m1,m2<0ifm2<m2<m̄2m1,=0ifm2=m̄2m1,>0ifm2>m̄2m1.

Therefore, (û,0) is locally asymptotically stable for m2<m2<m̄2(m1), neutrally stable for m2=m̄2(m1), and unstable for m2>m̄2(m1).

(ii) Similarly, we can conclude from Lemma 4.2(ii) that for any m1(m1,+) given, there exists a unique m̃2(m1)>m2 such that

τ1m1,m2>0ifm2<m2<m̃2m1,=0ifm2=m̃2m1,<0ifm2>m̃2m1.

Therefore, (0,v̂) is unstable for m2<m2<m̃2(m1), neutrally stable for m2=m̃2(m1), and locally asymptotically stable for m2>m̃2(m1). The proof is completed.

Combining with Lemma 3.4 and Lemma 4.3, we obtain the following results.

Theorem 4.2 Suppose mi>mi(i=1,2). For fixed d, k1, k2, β1, β2 > 0, m̄2 and m̃2 are defined by Eq. 4.15, respectively, and the following statements hold.

(i) Suppose m̄2<m̃2. If m2<m̄2, then (û,0) is locally asymptotically stable, and (0,v̂) is unstable if it exists. If m2>m̃2, then (û,0) is unstable, and (0,v̂) is locally asymptotically stable. If m2(m̄2,m̃2), then (û,0) and (0,v̂) are both unstable, and system (2.2) admits at least one stable coexistence steady state.

(ii) Suppose m̄2>m̃2. If m2<m̃2, then (û,0) is locally asymptotically stable, and (0,v̂) is unstable. If m2>m̄2, then (û,0) is unstable, and (0,v̂) is locally asymptotically stable. If m2(m̃2,m̄2), then (û,0) and (0,v̂) are both stable, and system (2.2) admits at least one unstable coexistence steady state.

(iii) Suppose m̄2=m̃2. If m2<m̃2, then (û,0) is locally asymptotically stable, and (0,v̂) is unstable. If m2>m̃2, then (û,0) is unstable, and (0,v̂) is locally asymptotically stable.

For fixed d, k1, k2, β1, β2 > 0, Lemma 4.3 and Theorem 4.2 imply that there exist two critical curves Γ1 and Γ2 in the m1m2 plane, which divide the local dynamics of Eq. 2.2 into competitive exclusion, bi-stability, and coexistence. To further characterize classification on the dynamics of system (2.2) in the m1m2 plane, we next give some properties of critical curves Γ1 and Γ2. We recall

μ1m̄2f2ϕû;m1,β1,0=0andμ1m1f1ϕv̂;m̃2,β2,0=0.

Clearly, m̄2 depends on m1 and β1 and m̃2 depends on m1 and β2. To emphasize these parameters, we denote m̄2 and m̃2 by m̄2(m1,β1) and m̃2(m1,β2), respectively.

Proposition 4.2 We consider d, k1, k2 > 0 fixed. The critical curve m̄2(m1,β1) has the following properties.

(i) For any β1 > 0 given, m̄2(m1,β1) is strictly increasing with respect to m1 in (m1,+). Moreover,

limm1m1+m̄2m1,β1=m2.

(ii) For any m1>m1 given, m̄2(m1,β1) is strictly decreasing with respect to β1 in (0, + ∞) and

m2<m̄2m1,β1<maxk2k1,1m1andlimβ1+m̄2m1,β1=m2.

Particularly,

maxm2,mink2k1,1m1m̄2m1,β1<maxk2k1,1m1provided0<β1β10.

(iii) For any m1>m1 given,

limβ1+m̄̇2m1,β1=0,

where m̄̇2(m1,β1) is the derivative of m̄2(m1,β1) with respect to m1 in (m1,+).

Proof (i) For any β1 > 0 given, we can conclude from Lemma 4.2(i.1) (i.3) that μ1(m2f2(ϕû),0) is strictly decreasing with respect to m1 in (m1,+) and strictly increasing with respect to m2 in (m2,+). Then, it follows from μ1(m̄2(m1,β1)f2(ϕû,0))=0 and the implicit function theorem that m̄2(m1,β1) is strictly increasing with respect to m1 in (m1,+). Since μ1(m̄2(m1,β1)f2(ϕû),0)=0, we can obtain limm1(m1)+m̄2(m1,β1)=m2 based on μ1(m2f2(ϕ,0))=0 and limm1(m1)+û=0 uniformly on [0,1] (see Eq. 3.5).

(ii) For any m1>m1 given, it follows from Lemma 4.2(i.2) (i.3) that μ1(m2f2(ϕû),0) is strictly increasing with respect to β1 in (0, + ) and strictly increasing with respect to m2 in (m2,+). Then, we can obtain from μ1(m̄2(m1,β1)f2(ϕû,0))=0 and the implicit function theorem that m̄2(m1,β1) is strictly decreasing with respect to β1 in (0, + ). Then,

limβ1+m̄2m1,β1<m̄2m1,β1<limβ10+m̄2m1,β1form1>m1.

Substituting β1 → 0+ in μ1(m̄2(m1,β1)f2(ϕû,0))=0, we have μ1(m̄2(m1,0)f2(ϕu0,0))=0 by Lemma 3.1(ii). Here, u0 satisfies

du0xx+m1ϕu0u0k1+ϕu0=0,x0,1,u0x0=u0x1+γu01=0.

Then, we can deduce from [26], Theorem 2.1 that limβ10+m̄2(m1,β1)max{k2k1,1}m1. Substituting β1 → + in μ1(m̄2(m1,β1)f2(ϕû,0))=0, we have limβ1+m̄2(m1,β1)=m2 based on μ1(m2f2(ϕ,0))=0 and limβ1+û=0 uniformly on [0,1] (see (3.6)). The estimate of m̄2(m1,β1) in Eq. 4.16 is obtained.

Finally, when m1>m1,m2<m2<min{k2k1,1}m1, Theorem 4.1(iii.1) shows that (û,0) is globally asymptotically stable provided 0<β1β10. Therefore, m̄2(m1,β1)max{m2,min{k2k1,1}m1} provided 0<β1β10. Combining with Eq. 4.16, 4.17 is obtained.

(iii) Let ψ1 > 0 with ‖ψ1 = 1 be the corresponding principal eigenfunction of μ1(m̄2f2(ϕû,0))=0. Then, ψ1 satisfies

dψ1xx+m̄2f2ϕû,0ψ1=0,x0,1,ψ1x0=ψ1x1+γψ11=0.

By differentiating Eq. 4.19 with respect to m1, denoting m1=̇, we have

dψ1̇xx+m̄̇2f2ϕû,0ψ1+m̄2f2ϕû,0ψ1̇k2k2+ϕû2ûm1,β1m1ψ1=0,x0,1,ψ1̇x0=ψ1̇x1+γψ1̇1=0,

where û(m1,β1)m1 is the derivative of û(m1,β1) with respect to m1 in (m1,+). Multiplying Eq. 4.19 by ψ1̇ and Eq. 4.20 by ψ1, integrating over (0,1), and then, subtracting the two resulting equations,

m̄̇2m1,β1=m̄201k2k2+ϕû2ûm1,β1m1ψ12dx01f2ϕû,0ψ12dx.

We next show that limβ1+û(m1,β1)m1=m1limβ1+û(m1,β1)=0. We choose an increasing sequence {β1,n} with limn+β1,n=+. Then, ûnû(m1,β1,n) is the unique positive solution of

dûnxx+m1f1nϕûn,ûnûn=0,x0,1,ûnx0=ûnx1+γûn1=0,

where f1(n)(ϕûn,ûn)=ϕûnk1+ϕ+(β1,n1)ûn. Note that 0<ûn<ϕ and f1(n)(ϕûn,ûn) is uniformly bounded on [0,1]. By Lp estimates for p ∈ (1, + ) and the Sobolev embedding theorem, we may assume that limn+ûn=U1(x) in C1[0, 1] by passing to a subsequence if necessary. Since ûn is continuously differentiable with respect to m1 in (m1,+) (see Lemma 3.3(i)), we differentiate (4.22) with respect to m1, denote m1=̇, and obtain

dû̇nxx+m1f1nϕûn,ûnk1+β1,nϕûnk1+ϕ+β1,n1ûn2û̇n=f1nϕûn,ûnûn,x0,1,û̇nx0=û̇nx1+γû̇n1=0.

Since 0<ûn<ϕ, f1(n)(ϕûn,ûn)ûn, and f1(n)(ϕûn,ûn)(k1+β1,nϕ)ûn[k1+ϕ+(β1,n1)ûn]2 are uniformly bounded on [0,1], we can use Lp estimates for p ∈ (1, + ) and the Sobolev embedding theorem again to assume that limn+û̇n=U2(x) in C1[0, 1] by passing to a subsequence if necessary. Therefore, limβ1+û(m1,β1)m1=m1limβ1+û(m1,β1)=0 based on limβ1+û(m1,β1)=0 uniformly on [0,1] (see Lemma 3.3(ii)). Finally, taking β1 → + in (4.21) and noting that limβ1+m̄2(m1,β1)=m2 (see (4.16)), we have limβ1+m̄̇2(m1,β1)=0. The proof is completed.

Proposition 4.3 We consider d, k1, k2 > 0 fixed. The critical curve m̃2(m1,β2) has the following properties:

(i) For any β2 > 0 given, m̃2(m1,β2) is strictly increasing with respect to m1 in (m1,+). Moreover,

limm1m1+m̃2m1,β2=m2.

(ii) For any m1>m1 given, m̃2(m1,β2) is strictly increasing with respect to β2 in (0, + ) and

maxm2,mink2k1,1m1<m̃2m1,β2<+.

Particularly,

maxm2,mink2k1,1m1<m̃2m1,β2maxk2k1,1m1provided0<β2β20.

(iii) For any m1>m1 given,

limβ2+m̃̇2m1,β2=+,

where m̃̇2(m1,β2) is the derivative of m̃2(m1,β2) with respect to m1 in (m1,+).

Proof (i) For any β2 > 0 given, we can conclude from Lemma 4.2(ii.1) (ii.3) that μ1(m1f1(ϕv̂(;m2,β2),0)) is strictly increasing with respect to m1 in (m1,+) and strictly decreasing with respect to m2 in (m2,+). Then, m̃2(m1,β2) is strictly increasing with respect to m1 in (m1,+) by μ1(m1f1(ϕv̂(;m̃2,β2),0))=0 and the implicit function theorem. Substituting m1(m1)+ in μ1(m1f1(ϕv̂(;m̃2,β2),0))=0, we have limm1(m1)+v̂(;m̃2,β2)=0 based on μ1(m1f1(ϕ,0))=0. Therefore, limm1(m1)+m̃2(m1,β2)=m2 by limm̃2(m2)+v̂(;m̃2,β2)=0 uniformly on [0,1] (Eq. 3.5).

(ii) For any m1>m1 given, it follows from Lemma 4.2(ii.2) (ii.3) that μ1(m1f1(ϕv̂(;m2,β2),0)) is strictly increasing with respect to β2 in (0, + ) and strictly decreasing with respect to m2 in (m2,+). Then, we can conclude from μ1(m1f1(ϕv̂(;m̃2,β2),0))=0 and the implicit function theorem that m̃2(m1,β2) is strictly increasing with respect to β2 in (0, + ). Therefore,

limβ20+m̃2m1,β2<m̃2m1,β2<limβ2+m̃2m1,β2form1>m1.

Similarly, substituting β2 → 0+ in μ1(m1f1(ϕv̂(;m̃2,β2),0))=0, we can deduce from [8], Theorem 2.1 again that max{m2,min{k2k1,1}m1}limβ20+m̃2(m1,β2). Moreover, limβ2+m̃2(m1,β2)=+. Hence, the inequality about m̃2(m1,β2) in Eq. 4.24 is obtained.

When m1>m1, m2>max{m2,max{k2k1,1}m1}, Theorem 4.1(ii.1) implies that (0,v̂) is globally asymptotically stable provided 0<β2β20. Therefore, m̃2(m1,β2)max{k2k1,1}m1 provided 0<β2β20. Combining with 4.24, 4.25 holds.

(iii) Let φ1 with ‖φ1 = 1 be the corresponding principal eigenfunction of μ1(m1f1(ϕv̂(m̃2,β2),0))=0. Then, φ1 satisfies

dφ1xx+m1f1ϕv̂m̃2,β2,0φ1=0,x0,1,φ1x0=φ1x1+γφ11=0.

By differentiating Eq. 4.27 with respect to m1 and denoting m1=̇, we have

dφ1̇xx+f1ϕv̂,0φ1+m1f1ϕv̂,0φ1̇k1k1+ϕv̂2v̂m̃2,β2m2×m̃̇2m1φ1=0,x0,1,φ1̇x0=φ1̇x1+γφ1̇1=0,

where v̂(m̃2,β2)m2 is the derivative of v̂ with respect to m2 at m2=m̃2. Multiplying (4.27) by φ1̇ and (4.28) by φ1, integrating over (0,1), and then, subtracting the two resulting equations,

m̃̇2m1,β2=01f1ϕv̂,0φ12dxm101k1k1+ϕv̂2v̂m̃2,β2m2φ12dx.

Similar to Proposition 4.2(iii), we can show limβ2+v̂(m̃2,β2)m2=m2limβ2+v̂(m̃2,β2)=0 based on limβ2+v̂(m̃2,β2)=0 uniformly on [0,1] (see Lemma 3.3(ii)). Substituting β2 → + in (4.29), it is easy to obtain limβ2+m̃̇2(m1,β2)=+.

We assume k1 > k2 > 0 without loss of generality. Then, there exist six critical curves: m1=m1,m2=m2,

L1:m2=k2k1m1,L2:m1=m2,Γ1:m2=m̄2m1,Γ2:m2=m̃2m1,

in the m1m2 plane (Figure 2), which classify the dynamics of system (2.2) into extinction of both species, competitive exclusion and coexistence. Clearly, it follows from Proposition 4.1(i) that line m2=m2m1m1 is located above line L1 and below line L2 under the assumption k1 > k2 > 0. Propositions 4.2(i) and 4.3(i) suggest that Γ1 and Γ2 are increasing with respect to m1(m1,+), respectively. Moreover, limm1(m1)+m̄2(m1)=limm1(m1)+m̃2(m1)=m2, which implies that Γ1 and Γ2 intersect at point (m1,m2). In addition, due to the effect of β1 and β2, Propositions 4.2(ii) (iii) and 4.3(ii) (iii) indicate that the locations of Γ1 and Γ2 in the m1m2 plane have the following four occasions shown in Figures 2A–D. Here, we assume that Γ2 is always located above Γ1 in this region (Figure 2). We set

Π00,m1×0,m2;Π1m1,+×0,m2;Π20,m1×m2,+.

Now, we are ready to illustrate the dynamical classification of system (2.2) in the m1m2 plane under the assumption k1 > k2 > 0, by dividing the following four cases.

Case I: 0<β1β10, 0<β2β20 (Figure 2A). Then, (4.17) holds provided 0<β1β10; that is, k2k1m1<m̄2(m1,β1)<m1 under the assumption k1 > k2 > 0. Similarly, (4.25) holds provided 0<β2β20, which means k2k1m1<m̃2(m1,β1)<m1. These suggest that both Γ1 and Γ2 lie between lines L1 and L2 (Figure 2A). It follows from Theorem 4.1(i.1) that (0,0) is g.a.s in region (m1, m2) ∈ Π0. In particular, the phase portrait graph of system (2.2) with (m1, m2) = (0.2, 0.1) ∈ Π0 is illustrated in Figure 1A, which shows that (0,0) is g.a.s in this case. Then, (û,0) is g.a.s when (m1, m2) ∈ Π1 ∪ Π3 by Theorem 4.1(i.3), (iii.1). Particularly, the phase portrait graph of system (2.2) with (m1, m2) = (1, 0.1) ∈ Π1 ∪ Π3 is shown in Figure 1B. (0,v̂) is g.a.s in region (m1, m2) ∈ Π2 ∪ Π4 by Theorem 4.1(i.2) and (ii.1). Moreover, the specific phase portrait graph with (m1, m2) = (0.2, 1) ∈ Π2 ∪ Π4 is displayed in Figure 1C. Furthermore, by Lemma 4.3, (û,0) is locally asymptotically stable in region Π5 and unstable in region Π6 ∪ Π7; (0,v̂) is locally asymptotically stable in region Π6 and unstable in region Π5 ∪ Π7. Then, we can conclude from Theorem 4.2(i) that there exist stable coexistence steady states in Π7, and the specific phase portrait graph with (m1, m2) = (1, 0.545) ∈ Π7 is presented in Figure 1D.

Case II: 0<β1β10, β2>β2 (Figure 2B). Then, (4.17) holds and Γ1 still lies between lines L1 and L2 when 0<β1β10. Moreover, (4.24) holds when β2>β2; that is, m̃2(m1,β2)>max{m2,k2k1m1} under the assumption k1 > k2 > 0. This implies that Γ2 lies above lines L1 and m2=m2. Note that limm1(m1)+m̃2(m1,β2)=m2 by Proposition 4.3(i) and m1>m2 if k1 > k2 > 0 (see Proposition 4.1). Then, m̃2(m1,β2)<m1 when m1 is near m1, which means that there exists sufficiently small ϵ > 0 such that Γ2 lies below line L2 for m1(m1,m1+ϵ).

Next, we illustrate two-fold that Γ2 will first intersect L2 and then lie above L2 as m1 increases (Figure 2B). On one hand, Theorem 4.1(ii) indicates that when m1>m1 and m2>max{m2,m1}, there exists a unique β2 such that (0,v̂) is unstable when β2>β2. This means that there exist regions above L2 such that (0,v̂) is unstable when β2>β2. Note that Γ2 is strictly increasing with respect to β2 in (0, + ) by Proposition 4.3(ii). Hence, when β2>β2, we can conclude from Lemma 4.3(ii) that the critical curve Γ2 for the stability change of (0,v̂) must intersect L2 and then lie above L2 as m1 is suitably large. This implies that there exist regions above line L2 and below Γ2 such that (0,v̂) is unstable when β2>β2. On the other hand, Proposition 4.3(iii) implies limβ2+m̃̇2(m1,β2)=+, which means that the slope of Γ2 will be bigger than 1 for suitably large β2. This, in turn, suggests that for large β2, there exists m̄1>m1 such that Γ2 must intersect L2 and then locates above L2 for all m1>m̄1. Therefore, the region Π71 occurs for large β2. This region is denoted as Π71 in Figure 2B. Then, we deduce from Theorem 4.2(i) that there exist stable coexistence steady states in Π7=Π70Π71. The other regions Π0 − Π6 can be similarly defined as in Case I.

Case III: β1>β1, 0<β2β20 (see Figure 2C). Similar to Case II, (4.25) holds provided 0<β2β20; that is, line Γ2 lies between lines L1 and L2 when 0<β2β20. Moreover, (4.16) holds when β1>β1; that is, m2<m̄2(m1,β1)<m1 under the assumption k1 > k2 > 0. This implies that Γ1 lies below line L2 and above line m2=m2. Furthermore, note that limm1(m1)+m̄2(m1,β1)=m2 by Proposition 4.2(i) and m2>k2k1m1 if k1 > k2 > 0 by Proposition 4.1. Then, m̄2(m1,β1)>k2k1m1, when m1 is near m1, which means that there exists sufficiently small ɛ > 0 such that Γ1 lies above L1 for m1(m1,m1+ε).

Similarly, we next illustrate that Γ1 will first intersect L1 and then lie below L1 as m1 increases (see Figure 2C). On one hand, Theorem 4.1(iii) suggests that when m1>m1 and m2<m2<k2k1m1, there exists a unique β1 such that (û,0) is unstable when β1>β1. This means that there exist regions below L1 such that (û,0) is unstable when β1>β1. Note that Γ1 is strictly decreasing with respect to β1 in (0, + ) by Proposition 4.2(ii). Hence, when β1>β1, we can deduce from Lemma 4.3(i) that the critical curve Γ1 for the stability change of (û,0) must intersect L1 and then lie below L1 as m1 is suitably large. This implies that there exist regions below L1 and above Γ1 such that (û,0) is unstable when β1>β1.

On the other hand, Proposition 4.2(iii) gives limβ1+m̄̇2(m1,β1)=0, which implies that the slope of Γ1 will be smaller than k2k1 for suitably large β1. This, in turn, suggests that, for large β1, there exists m̄1>m1 such that Γ1 must intersect L1 and then lie below L1 for all m1>m̄1. Therefore, the region Π72 occurs for large β1. This region is denoted as Π72 in Figure 2C. Then, we obtain from Theorem 4.2(i) that there exist stable coexistence steady states in Π7=Π70Π72. The other regions Π0 − Π6 can be similarly illustrated as in Case I.

Case IV: β1>β1, β2>β2 (Figure 2D). Combining the analysis of β1>β1 in Case II and β2>β2 in Case III, we know that both the regions Π71 and Π72 exist. It follows from Theorem 4.2(i) that there exist stable coexistence steady states in Π7=Π70Π71Π72. The other regions Π0 − Π6 can be similarly illustrated as in Case I.

We next make a comparison with the results in [8]. When the intraspecific competition is relatively weak (i.e., for the case of 0<β1β10 and 0<β2β20), the competitive dynamics of system (2.2) (Figure 2A) are similar to the unstirred chemostat model with Holling type II functional response (see [8], Theorems 2.1, 2.4] and Figure 1 in [8]), which suggests that the weak intraspecific competition has little effect on the competition outcomes of species.

However, when the intraspecific competition becomes strong, some new phenomena may occur. For instance, for the standard unstirred chemostat models, competition exclusion always happens for the weak–strong competition of two species (see [26], Theorem 2.1), while coexistence may occur in the unstirred chemostat model with B.–D. functional response with the increase of intraspecific competition, under the weak–strong competition cases (see Theorem 4.1 and Figures 2B–D). More precisely, under the weak–strong competition case m2>max{k2k1,1}m1 (i.e., species v has a stronger growth ability compared to species u), the competitive ability of species v becomes weak with the increase of β2, which may result in the coexistence of the two species (see Theorem 4.1(ii.2) and Figure 2B). Similarly, under the weak–strong competition case m2<min{k2k1,1}m1 (i.e., species u has a stronger growth ability compared to species v), the competitive ability of species u becomes weak with the increase of β1, which may cause the coexistence of the two species (see Theorem 4.1(iii.2) and Figure 2C). In particular, Theorem 4.1(ii.2), (iii.2) and Figure 2D suggest that coexistence is more likely to happen, when the intraspecific competition of two species is strong.

In summary, these theoretical results indicate that for the weak–strong competition cases, if the intraspecific competition parameter of the species with stronger growth ability is suitably large, we can observe different results from [8] that coexistence may occur. This new phenomenon suggests that the intraspecific competition parameters β1 and β2 have a great influence on the competitive outcomes of two species.

FIGURE 1
www.frontiersin.org

FIGURE 1. Phase portrait graphs of system (2.2) for different growth rates m1 and m2. Here, we take β1 = 0.01, β2 = 0.01, L = 1, S0 = 1, d = 0.5, γ = 0.5, k1 = 1, and k2 = 0.4. As shown, (0,0) is globally asymptotically stable (simplified as g.a.s) in (A) with m1 = 0.2, m2 = 0.1; (û,0) is g.a.s. in (B) with m1 = 1, m2 = 0.1; (0,v̂) is g.a.s. in (C) with m1 = 0.2, m2 = 1; and there exist stable coexistence steady states in (D) with m1 = 1, m2 = 0.545.

FIGURE 2
www.frontiersin.org

FIGURE 2. Illustration of the dynamics of system (2.2) in the m1m2 plane for the case of k1 > k2 > 0. More precisely, (A) 0<β1β10 and 0<β2β20; (B) 0<β1β10 and β2>β2; (C) β1>β1 and 0<β2β20; and (D) β1>β1 and β2>β2. Then, (0,0) is g.a.s in region Π0; (û,0) is g.a.s in region Π1 ∪ Π3; (0,v̂) is g.a.s in region Π2 ∪ Π4; (û,0) is locally asymptotically stable in region Π5 and unstable in region Π6 ∪ Π7; (0,v̂) is locally asymptotically stable in region Π6 and unstable in region Π5 ∪ Π7; and there exist stable coexistence steady states in Π7. Here, we note that (B) Π7=Π70Π71, (C) Π7=Π70Π72, and (D) Π7=Π70Π71Π72.

5 Positive solution branches of system (2.2)

We define X = W2,p(0, 1) × W2,p(0, 1) and Y = Lp(0, 1) × Lp(0, 1), where p > 1. For fixed d, k1, k2, β1, β2 > 0 and m1>m1, system (2.2) admits three branches of trivial or semi-trivial solutions in the space R+×X: Γ0={(m2,0,0):m2>0},Γu={(m2,û,0):m2>0}andΓv={(m2,0,v̂):m2>m2}, where û=ω(;m1,β1) and v̂=ω(;m2,β2). In this section, we will regard m2 as the bifurcation parameter and study separately positive solutions bifurcating from the semi-trivial branches Γu and Γv by the Crandall–Rabinowitz bifurcation theorem in [31].

We first show that there exists a positive solution branch that bifurcates from the semi-trivial solution (û,0). Moreover, the bifurcation of positive solutions from (û,0) can only occur at m2=m̄2 by Lemma 4.3.

Theorem 5.1 For fixed d, k1, k2, β1, β2 > 0 and m1>m1, there is a smooth non-constant solutions curve Γ1 = {(m2(s), u(s), v(s)): s ∈ (−ϵ, ϵ)} such that (m2(s), u(s), v(s)) is a positive solution of system (2.2) for s ∈ (0, ϵ) and satisfies m2(0)=m̄2, u(s)=û+sφ0+o(s), and v(s) = 0 + o(s). Here, ψ0 > 0 is the principal eigenfunction corresponding to the eigenvalue μ1(m̄2f2(ϕû,0))=0, which satisfies

B2m̄2ψ0=0,x0,1,ψ00=ψ01+γψ01=0,

and φ0 < 0 satisfies

B1m1φ0=m1ûf1vϕû,ûψ0,x0,1,φ00=φ01+γφ01=0.

Moreover,

m20=01m̄2f2uϕû,0φ0+f2vϕû,0ψ0ψ02dx01f2ϕû,0ψ02dx,

where f2u(ϕû,0)=k2+β2û(k2+ϕû)2 and f2v(ϕû,0)=k2+β2ϕ(k2+ϕv̂)2.

Theorem 5.1. can be proved by the similar arguments as in [35], Theorem 6.2. For completeness, we defer the proof of Theorem 5.1 to the Supplementary Appendix.

Theorem 5.2 For fixed d, k1, k2, β1, β2 > 0 and m2>m2, there is a smooth non-constant solutions curve Γ2 = {(m2(s), u(s), v(s)): s ∈ (−ϵ, ϵ)} such that (m2(s), u(s), v(s)) is a positive solution of (2.2) for s ∈ (0, ϵ) and satisfies m2(0)=m̃2, u(s)=sφ̃0+o(s), and v(s)=v̂+sψ̃0+o(s)(v̂=v̂(;m̃2,β2)). Here, φ̃0>0 is the principal eigenfunction corresponding to μ1(m1f1(ϕv̂(;m̃2,β2),0))=0, which satisfies

B3m̃2φ̃0=0,x0,1,φ̃00=φ̃01+γφ̃01=0,

and ψ̃0<0 satisfies

B4m̃2ψ̃0=m̃2v̂f2uϕv̂,v̂φ̃0,x0,1,ψ̃0=ψ̃1+γψ̃1=0.

Moreover,

m20=01f1uϕv̂,0φ̃0+f1vϕv̂,0ψ̃0φ̃02dx01v̂f1vϕv̂,0ψ̃02dx,

where f1u(ϕv̂,0)=k1+β1ϕ(k1+ϕv̂)2, f1v(ϕv̂,0)=k1+β1v̂(k1+ϕv̂)2, and v̂ is the derivative of v̂ with respect to m2 at m2=m̃2.

The proof of Theorem 5.2 is similar to the arguments in Theorem 5.1, and we omit it here.

We define Ω={(m2,u,v)R+×X:u>0,v>0,(m2,u,v)satisfiessystem(2.3)}. The following results show that the two bifurcation continua Γ1 and Γ2 in Theorems 5.1 and 5.2 are connected.

Theorem 5.3 We consider d, k1, k2, β1, β2 > 0 and m1>m1 fixed. Then, there exists a connected component Γ of Ω, which bifurcates from the semi-trivial solution branch Γu at (m̄2,û,0) and meets the other semi-trivial solution branch Γv at (m̃2,0,v̂). In particular, system (2.3) admits a positive solution (u, v) if m2 lies between m̄2 and m̃2.

The proof is motivated by the methods in [13], Theorem 6.4 (see also [8], Theorem 4.10]). For readability, the proof is given in Supplementary Appendix.

6 Numerical descriptions

In this section, we will study the effect of diffusion rates d on the dynamics of system (2.2). It follows from Eqs 4.1, 4.15 that the threshold values m1, m2, m̄2, and m̃2 depend on the diffusion rates d. Since these threshold values are dependent on d and they are non-monotone in d, we cannot theoretically establish the threshold dynamics of system (2.2) in terms of d. In order to explore the effect of d on the dynamics of system (2.2), in this subsection, we resort to numerical approaches. We fix L = 1, S0 = 1, γ = 0.5, k1 = 1, k2 = 0.4, and m1 = 1 as mentioned before. Then, by presenting the bifurcation diagrams of positive equilibrium solution of system (2.2) with the bifurcation parameter d increasing, the results are divided into the following three cases.

Case I: m2>max{k2k1,1}m1. We take m2 = 2 such that m2k2=5>m1k1=1, which means that species v has a stronger growth ability compared to species u. We can call this case the weak–strong competition [13]. To identify the effect of intraspecific competition, we fix β1 = 0.01 and let β2 change from β2 = 0.01 to β2 = 1.

First, if β1 = 0.01, β2 = 0.01, there is no coexistence and the competitive exclusion principle holds (species v with a stronger growth ability will win the competition) when d is sufficiently small (Figure 3A). As d increases, both species go extinct, which is consistent with our biological intuition that the sufficiently large diffusion rates will put species at a disadvantage. These numerical observations in Figure 3A coincide with [26], Theorem 5.4.

FIGURE 3
www.frontiersin.org

FIGURE 3. Bifurcation diagrams of the positive steady-state solutions to system (2.2) at t = 2000 with the bifurcation parameter d ranging from 0.01 to 10. Here, we take L = 1, S0 = 1, γ = 0.5, k1 = 1, k2 = 0.4, m1 = 1, and m2 = 2 and (A) β1 = 0.01, β2 = 0.01 and (B) β1 = 0.01, β2 = 1.

Second, if β1 = 0.01, β2 = 1. Clearly, under this weak–strong competition case, though species v has stronger growth ability compared to species u, the increase of β2 makes the competitive ability of v weaker. This is consistent with our biological intuition that the stronger intraspecific competition will put species at a disadvantage. Moreover, coexistence may occur when d is sufficiently small (Figure 3B), which is different from [8], Theorem 5.4. As d increases, species v wins the competition. As d further increases, the sufficiently large diffusion rates drive both species to extinction.

Case II: m1>max{k1k2,1}m2. We take m2 = 0.2 such that m1=1>max{k1k2,1}m2=0.5, which implies that species u has a stronger growth ability compared to species v. Similarly, we call this case the weak–strong competition case [13]. We fix β2 = 0.01 and let β1 change from β1 = 0.01 to β1 = 1. Similar to Case I, when β1 = 0.01, β2 = 0.01, results similar to those in Figure 3A are shown in Figure 4A. When β1 = 1, β2 = 0.01, the increase of β1 makes the competitive ability of u weaker. Then, we can observe that coexistence may occur when d is sufficiently small (see Figure 4B), which is quite different from [26], Theorem 5.4.

FIGURE 4
www.frontiersin.org

FIGURE 4. Bifurcation diagrams of the positive steady-state solutions to system (2.2) at t = 2000 with the bifurcation parameter d ranging from 0.01 to 10. Here, we take L = 1, S0 = 1, γ = 0.5, k1 = 1, k2 = 0.4, m1 = 1, and m2 = 0.2 and (A) β1 = 0.01, β2 = 0.01 and (B) β1 = 1, β2 = 0.01.

Case III: k2k1m1<m2<m1. We take m2 = 0.6 such that k2k1m1=0.4<m2=0.6<m1=1. We call this case the evenly matched competition [13]. Moreover, the fact of m1k1<m2k2 suggests that though the growth ability of the two species is evenly matched, the competitive ability of species v is still slightly better than that of species u. Then, for this evenly matched competition case, we fix β1 = 0.01 and let β2 change from β2 = 0.01 to β2 = 1.

For β1 = 0.01, β2 = 0.01, as shown in Figure 5A, the diffusion rates have a significant effect on the dynamics of system (2.2). More precisely, the dynamics of system (2.2) shift between four scenarios with the bifurcation parameter d increasing; that is, 1) competitive exclusion occurs and species v wins the competition, when d is sufficiently small; 2) coexistence occurs as d increases; 3) competitive exclusion occurs again and species u wins the competition, as d further increases; and 4) both species are washed out as d continues to increase. These suggest that system (2.2) may show a trade-off among extinction, exclusion, and coexistence as d increases. Particularly, coexistence occurs at the intermediate diffusion rates, which is in line with the theoretical results in [8].

FIGURE 5
www.frontiersin.org

FIGURE 5. Bifurcation diagrams of the positive steady state solutions to system (2.2) at t = 2000 with the bifurcation parameter d ranging from 0.01 to 5. Here we take L = 1, S0 = 1, γ = 0.5, k1 = 1, k2 = 0.4, m1 = 1, and m2 = 0.6 and (A) β1 = 0.01, β2 = 0.01 and (B) β1 = 0.01, β2 = 1.

For β1 = 0.01, β2 = 1, as stated before, the increase of β2 will make the competitive ability of v weaker. Then, we can observe from Figure 5B that both species can coexist when d is sufficiently small. As d increases, competitive exclusion happens and species u wins the competition. As d further increases, the large diffusion rates drive two species to extinction.

In shorts, for different competition Cases (I)–(III), we investigate the effect of diffusion on the dynamics of system (2.2) by taking different intraspecific competition parameters β1, β2. As shown in Figures 35, the impacts of diffusion and intraspecific competition on the competitive outcomes of species are complex, which further suggests that diffusion and intraspecific competition play a key role in determining the dynamics of system (2.2).

7 Discussion

In this paper, we investigate an unstirred chemostat model with the Beddington–DeAngelis functional response (see system (2.2)). The analytical and numerical results show that the intraspecific competition and diffusion have an important biological effect on the dynamics of system (2.2).

Theoretically, we first adopt a basic strategy regarding the growth rates as variable/bifurcation parameters to study the effect of growth rates on system (2.2). The results show that there exist six critical curves

m1=m1,m2=m2,L1:m2=k2k1m1,L2:m1=m2,Γ1:m2=m̄2m1,Γ2:m2=m̃2m1

in the m1m2 plane, which may classify the dynamics of system (2.2) into extinction of both species, competitive exclusion and coexistence (see Theorems 4.1, 4.2). To further understand the effect of βi, (i = 1, 2) on the dynamics of (2.2), we explore the properties of critical curves Γ1 and Γ2 (see Propositions 4.2 and 4.3) and get a relatively clear dynamics classification of system (2.2) in the m1m2 plane (Figure 2).

Numerically, since diffusion plays a key role in determining the competition outcomes of two species, we study the effect of diffusion on the dynamics of system (2.2). More precisely, for two weak–strong competition cases, due to the effect of intraspecific competition parameters β1 and β2, the coexistence may occur at sufficiently small diffusion rates (Figures 3B, 4B), while for the evenly matched competition case, the dynamics of system (2.2) shift between different scenarios (competitive exclusion, coexistence, and extinction) when β1 and β2 are small and the coexistence only occurs at the intermediate diffusion rates (Figure 5A). When β2 is larger than β1, we observe from Figure 5B that coexistence may occur at sufficiently small diffusion rates.

In conclusion, in this paper, the dynamics classification of system (2.2) in the m1m2 plane is established by the linear eigenvalue theory and the monotone dynamical system theory (Figure 2). Due to the effect of intraspecific competition parameters β1 and β2, the dynamics of system (2.2) are more complex than that of the unstirred chemostat model with Holling type II functional response (see Figure 1 of [8]). Numerically, we study the effect of diffusion on system (2.2) and obtain rich numerical results (Figures 35). These numerical observations reveal that, under different competition cases, the effects of diffusion and intraspecific competition on the dynamics of system (2.2) are complex. This, in turn, suggests that the B.–D. functional response is more biologically realistic and superior to the well-known Holling type II functional response in modeling the resource uptake of species.

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material; further inquiries can be directed to the corresponding author.

Author contributions

The theory part and simulation were obtained by the first author WZ. The second and third authors HN and ZW guided the work. All authors contributed to the article and approved the submitted version.

Funding

This work was supported by the National Natural Science Foundation of China (12071270 and 12171296) and the Natural Science Basic Research Program of Shaanxi (No. 2023-JC-JQ-03).

Acknowledgments

The authors are very grateful to the referees and the handling co editor-in-chief for their kind and valuable suggestions leading to a substantial improvement of the manuscript.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors, and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphy.2023.1205571/full#supplementary-material

References

1. Bayen T, Cazenave-Lacroutz H, Coville J. Stability of the chemostat system including a linear coupling between species. Discrete Contin Dyn Syst Ser B (2023) 28(2023):2104–29. doi:10.3934/dcdsb.2022160

CrossRef Full Text | Google Scholar

2. Hsu SB, Jin Y. The dynamics of a two host-two virus system in a chemostat environment. Discrete Contin Dyn Syst Ser B (2021) 26:415–41. doi:10.3934/dcdsb.2020298

CrossRef Full Text | Google Scholar

3. Nie H, Hsu SB, Grover JP. Algal competition in a water column with excessive dioxide in the atmosphere. J Math Biol (2016) 72:1845–92. doi:10.1007/s00285-015-0926-8

PubMed Abstract | CrossRef Full Text | Google Scholar

4. Smith HL, Waltman P. The theory of the chemostat. Cambridge: Cambridge University, Press (1995).

Google Scholar

5. Dellal M, Bar B. Global analysis of a model of competition in the chemostat with internal inhibitor. Discrete Contin Dyn Syst Ser B (2021) 26:1129–48. doi:10.3934/dcdsb.2020156

CrossRef Full Text | Google Scholar

6. Hsu SB, Waltman P. On a system of reaction-diffusion equations arising from competition in an unstirred chemostat. SIAM J Appl Math (1993) 53:1026–44. doi:10.1137/0153051

CrossRef Full Text | Google Scholar

7. Nie H, Shi Y, Wu JH. The effect of diffusion on the dynamics of a predator-prey chemostat model. SIAM J Appl Math (2022) 3:821–48. doi:10.1137/21m1432090

CrossRef Full Text | Google Scholar

8. Shi JP, Wu YX, Zou XF. Coexistence of competing species for intermediate dispersal rates in a reaction-diffusion chemostat model. J Dynam Differential Equations (2020) 32:1085–112. doi:10.1007/s10884-019-09763-0

CrossRef Full Text | Google Scholar

9. Smith HL. Monotone dynamical systems: An introduction to the theory of competitive and cooperative systems. In: Mathematical surveys and monographs. Providence, RI: American Mathematical Society (1995).

Google Scholar

10. Wu JH. Global bifurcation of coexistence state for the competition model in the chemostat. Nonlinear Anal (2000) 39:817–35. doi:10.1016/s0362-546x(98)00250-8

CrossRef Full Text | Google Scholar

11. Holling CS. The functional response of predator to prey density and its role in mimicry and population regulation. Mem Entonmol Soc Can (1965) 45:1–60.

Google Scholar

12. Zhang W, Nie H, Wu JH. A reaction-diffusion-advection chemostat model in a flowing habitat: Mathematical analysis and numerical simulations. Internat J Bifur Chaos Appl Sci Engrg (2023) 33:1245–74. doi:10.1142/s0218127423500736

CrossRef Full Text | Google Scholar

13. Zhang W, Nie H, Wu JH. Dynamics of a reaction-diffusion-advection model with two species competing in a flow reactor. Discrete Contin Dyn Syst Ser B (2023) 28:3453–86. doi:10.3934/dcdsb.2022226

CrossRef Full Text | Google Scholar

14. So WH, Waltman P. A nonlinear boundary value problem arising from competition in the chemostat. Appl Math Comput (1989) 32:169–83. doi:10.1016/0096-3003(89)90092-1

CrossRef Full Text | Google Scholar

15. Beddington JR. Mutual interference between parasites or predators and its effect on searching efficiency. J Anim Ecol (1975) 44:331–40. doi:10.2307/3866

CrossRef Full Text | Google Scholar

16. DeAngelis DL, Goldstein RA, O’Neill RV. A model for trophic interaction. Ecology (1975) 56:661–892.

CrossRef Full Text | Google Scholar

17. Harrision GW. Comparing predator-prey models to Luckinbill’s experiment with didinium and paramecium. Ecology (1995) 76:357–69.

Google Scholar

18. Jiang HL, Wu JH, Wang LJ, Guo GH. Qualitative analysis for a competition model with B-D functional response and numerical simulation. Numer Methods Partial Differential Equations (2014) 30:1575–94. doi:10.1002/num.21848

CrossRef Full Text | Google Scholar

19. Zhang GH, Wang XL. Extinction and coexistence of species for a diffusive intraguild predation model with B-D functional response. Discrete Contin Dyn Syst Ser B (2018) 23:3755–86. doi:10.3934/dcdsb.2018076

CrossRef Full Text | Google Scholar

20. Meng Q, Yang L. Steady state in a cross-diffusion predator-prey model with the Beddington-DeAngelis functional response. Nonlinear Anal Real World Appl (2019) 45:401–13. doi:10.1016/j.nonrwa.2018.07.012

CrossRef Full Text | Google Scholar

21. He X, Zheng SN. Protection zone in a diffusive predator-prey model with Beddington-DeAngelis functional response. J Math Biol (2017) 75:239–57. doi:10.1007/s00285-016-1082-5

PubMed Abstract | CrossRef Full Text | Google Scholar

22. Feng XZ, Sun C, Yang WB, Li CT. Dynamics of a predator-prey model with nonlinear growth rate and B-D functional response. Nonlinear Anal Real World Appl (2023) 70:103766. in press. doi:10.1016/j.nonrwa.2022.103766

CrossRef Full Text | Google Scholar

23. Wang YE, Wu JH, Guo GH. Coexistence and stability of an unstirred chemostat model with Beddington-DeAngelis function. Comput Math Appl (2010) 60:2497–507. doi:10.1016/j.camwa.2010.08.057

CrossRef Full Text | Google Scholar

24. Nie H, Wu JH. Coexistence of an unstirred chemostat model with Beddington-DeAngelis functional response and inhibitor. Nonlinear Anal Real World Appl (2010) 11:3639–52. doi:10.1016/j.nonrwa.2010.01.010

CrossRef Full Text | Google Scholar

25. Feng XZ, Sun SP, Zhang TQ, An XM. The effect of parameters on positive solutions and asymptotic behavior of an unstirred chemostat model with B-D functional response. Adv Difference Equ (2018) 23:181–204. doi:10.1186/s13662-018-1587-x

CrossRef Full Text | Google Scholar

26. Zhang S, Tan D, Chen LS. Chaotic behavior of a chemostat model with Beddington-DeAngelis functional response and periodically impulsive invasion. Chaos Solitons Fractals (2006) 29:474–82. doi:10.1016/j.chaos.2005.08.026

CrossRef Full Text | Google Scholar

27. Zhou X, Song X, Shi X. Analysis of competitive chemostat models with the Beddington-DeAngelis functional response and impulsive effect. Appl Math Model (2007) 31:2299–312. doi:10.1016/j.apm.2006.08.010

CrossRef Full Text | Google Scholar

28. Hirsch MW, Smith HL, Zhao XQ. Chain transitivity, attractivity, and strong repellors for semidynamical systems. J Dynam Differential Equations (2001) 13:107–31. doi:10.1023/a:1009044515567

CrossRef Full Text | Google Scholar

29. Cantrell R, Cosner C. Spatial ecology via reaction-diffusion equations. Chester, UK: John Wiley and Sons Ltd (2003).

Google Scholar

30. Lou Y, Nie H, Wang YE. Coexistence and bistability of a competition model in open advective environments. Math Biosci (2018) 306:10–9. doi:10.1016/j.mbs.2018.09.013

PubMed Abstract | CrossRef Full Text | Google Scholar

31. Crandall MG, Rabinowitz PH. Bifurcation from simple eigenvalues. J Funct Anal (1971) 8:321–40. doi:10.1016/0022-1236(71)90015-2

CrossRef Full Text | Google Scholar

32. Protter MH, Weinberger HF. Maximum principles in differential equations. New York: Springer-Verlag (1984).

Google Scholar

33. Shi JP. Persistence and bifurcation of degenerate solutions. J Funct Anal (1999) 169:494–531. doi:10.1006/jfan.1999.3483

CrossRef Full Text | Google Scholar

34. Shi JP, Wang XF. On global bifurcation for quasilinear elliptic systems on bounded domains. J Differential Equations (2009) 246:2788–812. doi:10.1016/j.jde.2008.09.009

CrossRef Full Text | Google Scholar

35. L´opez-G´omez J. Global bifurcation for Fredholm operators. Rend Istit Mat Univ Trieste (2016) 48:539–64. doi:10.13137/2464-8728/13172

CrossRef Full Text | Google Scholar

Keywords: unstirred chemostat model, coexistence, competitive exclusion, bifurcation, numerical simulation

Citation: Zhang W, Nie H and Wang Z (2023) Dynamics of an unstirred chemostat model with Beddington–DeAngelis functional response. Front. Phys. 11:1205571. doi: 10.3389/fphy.2023.1205571

Received: 14 April 2023; Accepted: 06 July 2023;
Published: 27 July 2023.

Edited by:

Xiaoming Zheng, Central Michigan University, United States

Reviewed by:

Xueyong Zhou, Xinyang Normal University, China
Raid Naji, University of Baghdad, Iraq

Copyright © 2023 Zhang, Nie and Wang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Zhiguo Wang, zgwang@snnu.edu.cn

Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.