# Existence and stability of positive steady-state solutions for a Lotka-Volterrasystem with intraspecific competition

- Meihua Wei
^{1}Email author, - Jinyong Chang
^{2}and - Lan Qi
^{1}

**2014**:159

https://doi.org/10.1186/1687-1847-2014-159

© Wei et al.; licensee Springer. 2014

**Received: **12 February 2014

**Accepted: **14 May 2014

**Published: **30 May 2014

## Abstract

In this paper, we investigate the existence and stability of the positivesteady-state solutions for a Lotka-Volterra system with intraspecific competition byusing the Lyapunov-Schmidt reduction technique. To do this, we must firstly obtainthe semi-trivial steady states as their base, which extend the method used in theprevious studies. Our results show that the two competition species withintraspecific competition can coexist for bigger regions of the diffusion rate*μ* and also complete the existing works.

**MSC:** 35K57.

## Keywords

## 1 Introduction

The maintenance of biodiversity has received increasing attention from ecologists andmathematicians. Moreover, resource competition is thought as an important factor indriving evolutionary diversification, in which intraspecific competition for resourcesplays a major role; see [1–4]. The classical Lotka-Volterra competition system [5–8] gives a better description of the population competition, whose dynamicalbehaviors have been studied extensively; see [9–16] and the references therein. In this paper a competitive Lotka-Volterradiffusion model of two slightly different species is studied, namely, the two speciesare identical except for their intraspecific competition rates. For this system, oncethe diffusion is involved biologically, it would be very interesting to find out whetherand when the two species competing for the same limited resources to survive cancoexist.

where Ω is a bounded domain in ${R}^{N}$ with smooth boundary *∂* Ω,*μ* is the diffusion rate, $u(x,t)$ and $v(x,t)$ denote the densities of two competing species, and$a(x)$ represents the intrinsic growth rate of species.

*θ*given in the next section) as$t\to \mathrm{\infty}$ that the slower diffuser can drive the faster one toextinction, and thus the two species cannot coexist. In [18], it is further stated that the slower diffuser may fail with a time periodicfunction $a(x,t)$ instead of $a(x)$. In [19], the model with small variations of intrinsic growth rates is given by

*μ*increases, which implies that the two species only coexist for smallregions of

*μ*if they could. For $a(x)+\tau g(x)$ taking the form $a+\tau g(x)$ in [20], the authors further illustrate that the mutant can always survive andinvade, but the original species can only coexist in some cases. Furthermore, in [21] the system with different interspecific competition rates is given as

Then a new structure of coexistence states is obtained and the two species can coexisteven for bigger regions of *μ*.

where *τ* is a small positive constant (*i.e.*,$\tau \ll 1$) and $g(x)$, $h(x)$ are smooth functions, which indicates that the two speciesonly have different intraspecific competition rates. Our main purpose is to study theexistence and stability of the positive steady-state solutions (that is, the coexistencestates) of (1.2) by using the Lyapunov-Schmidt reduction technique, the implicitfunction theorem combined with finite covering theorem and the perturbation theory forcompleting the previous studies. To do this, however, we must firstly get thesemi-trivial steady-state solutions of (1.2) according to [22], which is different from the corresponding results already known, and extendthe method used in the existing works. The main results we obtained show that the twocompetition species with different intraspecific competition can also coexist for biggerregions of the diffusion rate *μ*, but it is also important to caution thatthe conclusions for stability obtained in this paper are contrary to the ones analyzedin [21].

The rest of this paper is organized as follows. In Section 2, from [22] we firstly give the semi-trivial steady-state solutions of (1.2), and thenanalyze the stability of the solutions. In Section 3, on the basis of thesemi-trivial solutions obtained in Section 2, we investigate the existence andnonexistence of positive steady-state solutions of (1.2) by using the Lyapunov-Schmidtreduction technique, the implicit function theorem and the finite covering theorem. InSection 4, combining stability theory with perturbation theory, we further discussthe stability of positive steady-state solutions in detail.

## 2 Existence and stability of semi-trivial steady-state solutions

In this section we firstly establish the existence of semi-trivial steady-statesolutions of (1.2) by [22], whose proof is omitted. Then we simply analyze the stability of thesemi-trivial solutions.

**Lemma 2.1** (see [22])

*Suppose that*

*λ*

*is a positive parameter and thateither*$\beta (x)>0$

*on an open subset of*

*∂*Ω,

*orthe boundary condition is a Dirichlet condition on partof*

*∂*Ω,

*or*${\int}_{\mathrm{\Omega}}m(x)\phantom{\rule{0.2em}{0ex}}dx<0$

*holds*.

*The principaleigenvalue*${\sigma}_{1}$

*of*

*is positive if and only if*$0<{\lambda}^{+}<\lambda $,

*where*${\lambda}^{+}$

*is the positive principal eigenvalue of*

*If*$\beta (x)\equiv 0$*and*${\int}_{\mathrm{\Omega}}m(x)\phantom{\rule{0.2em}{0ex}}dx>0$, *then*${\sigma}_{1}>0$*for all*$\lambda >0$.

where $d(x)\in {C}^{1+\alpha}(\overline{\mathrm{\Omega}})$, $d(x)\ge {d}_{0}>0$, $\beta (x)\in {C}^{1+\alpha}(\overline{\mathrm{\Omega}})$, $\beta (x)\ge 0$, $f(x,u)$ is Lipschitz in *u* and is a measurable function in*x* which is bounded if *u* is restricted to a bounded set and$f(x,0)=0$.

**Lemma 2.2** (see [22])

*Suppose that*$f(x,u)=q(x,u)u$

*with*$q(x,u)$

*of class*${C}^{2}$

*in*

*u*

*and*${C}^{\alpha}$

*in*

*x*,

*and thereexists*$K>0$

*such that*$q(x,u)<0$

*for*$u>K$.

*If the principaleigenvalue*${\sigma}_{1}$

*is positive in the problem*

*then* (2.1) *has a minimal positive equilibrium*${u}^{\ast}$, *and all solutions to* (2.1) *that areinitially positive on an open subset of* Ω *are eventually bounded byorbits which increase toward*${u}^{\ast}$*as*$t\to \mathrm{\infty}$.

**Lemma 2.3** (see [22])

*Suppose that the hypotheses of Lemma * 2.2 *are satisfied andthat*$f(x,u)=q(x,u)u$*with*$q(x,u)$*strictly decreasingin* *u* *for*$u\ge 0$. *Then the minimal positiveequilibrium*${u}^{\ast}$*is the only positive equilibrium for* (2.1).

- (H)
$a(x)$ is Hölder continuous on $\overline{\mathrm{\Omega}}$ and ${\int}_{\mathrm{\Omega}}a(x)\phantom{\rule{0.2em}{0ex}}dx>0$.

respectively have unique positive equilibriums $\tilde{u}(x,\mu ,\tau )$ and $\tilde{v}(x,\mu ,\tau )$, denoted by $\tilde{u}(\mu ,\tau )$ and $\tilde{v}(\mu ,\tau )$, for (H) and $\tau \ll 1$. Hence, we can find the following result.

**Theorem 2.1** *Assume that* (H) *holds and*$\tau \ll 1$. *Then* (1.2) *has semi*-*trivialsteady*-*state solutions*$(\tilde{u}(\mu ,\tau ),0)$*and*$(0,\tilde{v}(\mu ,\tau ))$.

**Theorem 2.2** *Suppose that the hypotheses of Theorem * 2.1 *arevalid*. *If*$\stackrel{\u02c6}{\lambda}(\mu ,\tau )<0$ ($\stackrel{\u02c6}{\lambda}(\mu ,\tau )>0$), *then*$(\tilde{u}(\mu ,\tau ),0)$*is asymptotically stable* (*unstable*).*If*$\tilde{\lambda}(\mu ,\tau )<0$ ($\tilde{\lambda}(\mu ,\tau )>0$), *then*$(0,\tilde{v}(\mu ,\tau ))$*is asymptotically stable* (*unstable*).

**Remark 2.1** In this section, we only give a simple conclusion for the stability ofsemi-trivial steady-state solutions of (1.2), whose detailed proofs are analogous toSection 3 of [21].

## 3 Existence and nonexistence of positive steady-state solutions

The research on the steady-state solutions in a competition-diffusion system is always ahot issue. In this section we establish the existence and nonexistence of positivesteady states of (1.2). For this purpose, we must make use of the semi-trivial solutionsobtained in Section 2 for the following results.

where $p>N$ such that ${W}^{2,p}\hookrightarrow {C}^{1}(\overline{\mathrm{\Omega}})$.

According to (2.2), it is obvious that (3.1) has nontrivial nonnegative solutions${\mathrm{\Upsilon}}_{\mu}=\{(s\theta (x,\mu ),(1-s)\theta (x,\mu )):s\in [0,1]\}$ for $\tau =0$. Then for $\tau \ll 1$ we will look for the solutions of (3.1) near${\mathrm{\Upsilon}}_{\mu}$, namely, the steady states of (1.2).

**Theorem 3.1**

*Suppose that*$G(\mu )$

*and*$H(\mu )$

*have no common roots*,

*andset*${\mu}_{1}$

*and*${\mu}_{2}$

*are the consecutive and simple rootsof*$G(\mu )H(\mu )$.

- (i)
*If*$G(\mu )H(\mu )<0$*in*$({\mu}_{1},{\mu}_{2})$,*then there exists*$\delta >0$*such that for*$\mu \in ({\mu}_{1}+\delta ,{\mu}_{2}-\delta )$*such that*(1.2)*has no positive steady*-*state solutions other than semi*-*trivial ones for*$\tau \ll 1$. - (ii)
*If*$G(\mu )H(\mu )>0$*in*$({\mu}_{1},{\mu}_{2})$,*then for any sufficiently small*$\delta >0$*and*$\tau \ll 1$, (1.2)*has the positive steady*-*state solutions*$(u(\mu ,\tau ),v(\mu ,\tau ))$*satisfying*$(\mu ,u(\mu ,\tau ),v(\mu ,\tau ))$*in the neighborhood of*$({\mu}_{1}-\delta ,{\mu}_{2}+\delta )\times {\mathrm{\Upsilon}}_{\mu}$*besides semi*-*trivial ones*.*Here*$\begin{array}{r}u(\mu ,\tau )=s(\mu ,\tau )[\theta (\mu )+{\tilde{y}}_{1}(\mu ,\tau )],\\ v(\mu ,\tau )=[1-s(\mu ,\tau )][\theta (\mu )+{\tilde{z}}_{1}(\mu ,\tau )]\end{array}$(3.2)

*with*$s(\mu ,0)=s(\mu ):=\frac{H(\mu )}{G(\mu )+H(\mu )}$, ${\tilde{y}}_{1}(\mu ,0)=0$*and*${\tilde{z}}_{1}(\mu ,0)=0$. *Moreover*, *the positive steady*-*statesolution branches*$(u(\mu ,\tau ),v(\mu ,\tau ))$*connect with the semi*-*trivialones*$(u(\underline{\mu}(\tau ),\tau ),v(\underline{\mu}(\tau ),\tau ))$*and*$(u(\overline{\mu}(\tau ),\tau ),v(\overline{\mu}(\tau ),\tau ))$, *and the smooth functions*$\underline{\mu}(\tau )$*and*$\overline{\mu}(\tau )$*are defined on*$\{\tau :\tau \ge 0\}$*such that*$\underline{\mu}(0)={\mu}_{1}$*and*$\overline{\mu}(0)={\mu}_{2}$.

*Proof*For $\tau \ll 1$, our purpose is to find the positive solutions of (3.1)near ${\mathrm{\Upsilon}}_{\mu}$. Following the Lyapunov-Schmidt reduction technique [25], we know that $X={X}_{1}\oplus {X}_{2}$ on the basis of the spaces defined above. Then we can set

where $s\in [0,1]$ and $(y,z)\in {X}_{2}$ near $(0,0)$.

*L*, and

*L*is a Fredholm operator of index zerosince

*X*is compactly imbedded in

*Y*. It is easy to check that$(\theta ,-\theta )$ is in the kernel of

*L*. Due to$\theta >0$, we know that 0 is a simple eigenvalue of

*L*,which leads to

*L*is an isomorphic mapping from ${X}_{2}$ to

*Y*, then, by applying the implicit theorem tothe second equation of (3.7), it can be solved to get a unique solution$(y,z)=(y(\mu ,\tau ,s),z(\mu ,\tau ,s))$ near $(0,0)$. Furthermore, combining the finite covering theorem, thereexists ${\delta}_{2}>0$ such that

*P*, we can get a smooth function$\xi (\mu ,\tau ,s)$ satisfying

Moreover, we know that the solutions $s=0$ and $s=1$ correspond to the semi-trivial solutions$(0,\tilde{v})$ and $(\tilde{u},0)$ of (3.1), respectively. Consequently, it remains to solve${\xi}_{1}(\mu ,\tau ,s)=0$.

*τ*, set $\tau =0$ and combine with (3.6) and (3.8), which leads to

For $G(\mu )H(\mu )<0$ in $({\mu}_{1},{\mu}_{2})$, for example, we let $G(\tilde{\mu})H(\tilde{\mu})<0,\tilde{\mu}\in ({\mu}_{1},{\mu}_{2})$, and then we obtain ${\xi}_{1}(\tilde{\mu},0,s)\ne 0$ when $\tilde{\mu}\in ({\mu}_{1},{\mu}_{2})$ and $s\in [0,1]$. By using the finite covering theorem, there exists${\delta}_{3}>0$ taken smaller if necessary such that${\xi}_{1}(\mu ,\tau ,s)=0$ has no solution for $(\mu ,\tau ,s)\in ({\mu}_{1}+{\delta}_{3},{\mu}_{2}-{\delta}_{3})\times (-{\delta}_{3},{\delta}_{3})\times (-{\delta}_{3},1+{\delta}_{3})$. This results in statement (i) of Theorem 3.1.

and then denote ${y}_{1}(\mu ,\tau ,s(\mu ,\tau ))$ and ${z}_{1}(\mu ,\tau ,s(\mu ,\tau ))$ by ${\tilde{y}}_{1}(\mu ,\tau )$ and ${\tilde{z}}_{1}(\mu ,\tau )$, respectively. From (3.8), it follows that${\tilde{y}}_{1}(\mu ,0)=0$ and ${\tilde{z}}_{1}(\mu ,0)=0$. Therefore, for $(\mu ,\tau ,s)\in ({\mu}_{1}-{\delta}_{4},{\mu}_{2}+{\delta}_{4})\times (-{\delta}_{4},{\delta}_{4})\times (-{\delta}_{4},1+{\delta}_{4})$, the equation $\xi (\mu ,\tau ,s)=0$ has solutions given by $s=0$, $s=1$ and $s=s(\mu ,\tau )$, which shows that (1.2) has not only semi-trivialsteady-state solution branches, but also positive branches given by (3.2) which meet thesemi-trivial ones whose form is discussed below.

Clearly, $s(\mu ,\tau )=0$ and $s(\mu ,\tau )=1$ correspond to the semi-trivial stationary solution of(1.2). For $\mu ={\mu}_{1}$, there is either $G({\mu}_{1})=0$ or $H({\mu}_{1})=0$ due to that $G(\mu )$ and $H(\mu )$ have no common roots. Without loss of generality, we take$G({\mu}_{1})=0$, and then $1-s({\mu}_{1},0)=1-s({\mu}_{1})=0$. Moreover, ${[1-s(\mu ,\tau )]}_{\mu}{|}_{(\mu ,\tau )=({\mu}_{1},0)}={[1-s(\mu ,0)]}_{\mu}{|}_{\mu ={\mu}_{1}}=\frac{{G}^{\prime}({\mu}_{1})}{H({\mu}_{1})}\ne 0$ since ${\mu}_{1}$ is the simple root of $G(\mu )H(\mu )$. Thus, by the implicit function theorem, we can obtain aunique solution of $1-s(\mu ,\tau )=0$ defined by a smooth function $\mu =\underline{\mu}(\tau )$ for $\tau \in [0,{\delta}_{5})$ with ${\delta}_{5}>0$ small enough and $\underline{\mu}(0)={\mu}_{1}$, which implies that the corresponding semi-trivialsteady-state solution of (1.2) can be described by $(u(\underline{\mu}(\tau ),\tau ),v(\underline{\mu}(\tau ),\tau ))=(\tilde{u}(\underline{\mu}(\tau ),\tau ),0)$. Similarly, for the case $H({\mu}_{1})=0$, we can get only the solution of $s(\mu ,\tau )=0$ still given by $\mu =\underline{\mu}(\tau )$, which corresponds to the semi-trivial steady-statesolution $(u(\underline{\mu}(\tau ),\tau ),v(\underline{\mu}(\tau ),\tau ))=(0,\tilde{v}(\underline{\mu}(\tau ),\tau ))$ of (1.2). Moreover, for $\mu ={\mu}_{2}$, similar results hold true, and the unique solution of$1-s(\mu ,\tau )=0$ or $s(\mu ,\tau )=0$ is denoted by $\mu =\overline{\mu}(\tau )$ with $\overline{\mu}(0)={\mu}_{2}$. Thus the relevant solution of (1.2) is given by$(u(\overline{\mu}(\tau ),\tau ),v(\overline{\mu}(\tau ),\tau ))=(\tilde{u}(\overline{\mu}(\tau ),\tau ),0)$ or $(u(\overline{\mu}(\tau ),\tau ),v(\overline{\mu}(\tau ),\tau ))=(0,\tilde{v}(\overline{\mu}(\tau ),\tau ))$. □

**Remark 3.1** The proof of Theorem 3.1 shows the existence of loops or branchesof positive steady states just as [21], whose details are omitted here.

## 4 Stability of positive steady-state solutions

in (4.1) with smooth functions ${\varphi}_{1}(\mu ,\tau )$ and ${\psi}_{1}(\mu ,\tau )$. Furthermore, the emphasis is on the sign of${\lambda}_{1}(\mu ,\tau )$. Hence, for small *τ*, we divide thediscussions of the sign of ${\lambda}_{1}(\mu ,\tau )$ into three cases, that is, *μ* close to${\mu}_{1}$, *μ* close to ${\mu}_{2}$ and *μ* bounded away from${\mu}_{1}$ and ${\mu}_{2}$. To do this, we must first prove the following lemmas.

**Lemma 4.1**

*For*$\tau \ll 1$,

*the principal eigenvalue*${\lambda}_{1}(\mu ,\tau )$

*of*(4.1)

*satisfies*

*Proof*Multiply the first equation of (4.1) by

*v*and integrate overΩ to get

*u*and integratingby parts, we can obtain

From (4.4) and (4.5), it suffices to show that (4.3) is valid. □

Furthermore, on the basis of Lemma 4.1, we can discuss the sign of${\lambda}_{1}(\mu ,\tau )$.

**Lemma 4.2**

*For*$\mu \in [{\mu}_{1}+\u03f5,{\mu}_{2}-\u03f5]$, $\u03f5>0$,

*we have the following conclusion*:

*Proof*It is obvious that $(u(\mu ,\tau ),v(\mu ,\tau ))\to (s(\mu )\theta (\mu ),[1-s(\mu )]\theta (\mu ))$ and $(\varphi (\mu ,\tau ),\psi (\mu ,\tau ))\to (\theta (\mu ),-\theta (\mu ))$ as $\tau \to {0}^{+}$ from (3.2) and (4.2). Then, for $\tau \to {0}^{+}$, it follows that

Thus, we can obtain (4.6) from (4.3), (4.7) and (4.8). □

For the discussions below, without loss of generality, we can assume the case$G({\mu}_{1})=0$ (the case $H({\mu}_{1})=0$ is similar), and then $1-s(\underline{\mu}(\tau ))=0$ and $(u(\underline{\mu}(\tau ),\tau ),v(\underline{\mu}(\tau ),\tau ))=(\tilde{u}(\underline{\mu}(\tau ),\tau ),0)$.

**Lemma 4.3**

*Assume that*$G({\mu}_{1})=0$.

*Then we have the following conclusion*:

*Proof*According to Theorem 3.1, we know that $(\underline{\mu}(\tau ),u(\underline{\mu}(\tau ),\tau ),v(\underline{\mu}(\tau ),\tau ))=(\underline{\mu}(\tau ),\tilde{u}(\underline{\mu}(\tau ),\tau ),0)$ is the bifurcation point of the positive solutions to(1.2). Then ${\lambda}_{1}(\underline{\mu}(\tau ),\tau )=0$ and the eigenfunction $\psi (\underline{\mu}(\tau ),\tau )$ satisfies

by the mean value theorem for $\tilde{\mu}$ between *μ* and $\underline{\mu}(\tau )$.

*μ*, we obtain

as $\tau \to {0}^{+}$ and $\mu \to {\mu}_{1}$. From (4.3), (4.11) and (4.12), we can get the relation(4.9). □

In the same way, for *μ* close to ${\mu}_{2}$, the sign of ${\lambda}_{1}(\mu ,\tau )$ is decided by the result below, whose details are omittedhere.

**Lemma 4.4**

*Assume that*$G({\mu}_{2})=0$.

*Then we have the following conclusion*:

Finally, based on Lemmas 4.2-4.4, the main result of this section is presented asfollows.

**Theorem 4.1**

*Suppose that*$G(\mu )$

*and*$H(\mu )$

*have no common roots*, ${\mu}_{1}<{\mu}_{2}$

*are two consecutive and simple rootsof*$G(\mu )H(\mu )$

*and*$G(\mu )H(\mu )>0$

*in*$({\mu}_{1},{\mu}_{2})$.

*Then*,

*for*$\tau \ll 1$

*and*$\mu \in (\underline{\mu}(\tau ),\overline{\mu}(\tau ))$,

*we have*:

- (i)
*If*$G(\mu )>0$*and*$H(\mu )>0$*in*$({\mu}_{1},{\mu}_{2})$,*then the positive steady states*$(u(\mu ,\tau ),v(\mu ,\tau ))$*of*(1.2)*are asymptotically stable*; - (ii)
*If*$G(\mu )<0$*and*$H(\mu )<0$*in*$({\mu}_{1},{\mu}_{2})$,*then*$(u(\mu ,\tau ),v(\mu ,\tau ))$*is unstable*.

*Proof* Now, we only consider the statement (i) (the statement (ii) can be provedsimilarly). The key point in the proof of (i) is to obtain ${\lambda}_{1}(\mu ,\tau )<0$ by contradiction. Suppose that there exist${\tau}_{i}\to {0}^{+}$ and ${\mu}_{i}\in (\underline{\mu}({\tau}_{i}),\overline{\mu}({\tau}_{i}))$ such that ${\lambda}_{1}({\mu}_{i},{\tau}_{i})\ge 0$ for $i=1,2,\dots $ . By passing to the limit, we have$\underline{\mu}({\tau}_{i})\to {\mu}_{1}$, $\overline{\mu}({\tau}_{i})\to {\mu}_{2}$ and ${\mu}_{i}\to \tilde{\mu}$ as $i\to \mathrm{\infty}$ with $\tilde{\mu}\in [{\mu}_{1},{\mu}_{2}]$.

from Lemma 4.2 and $G(\mu )>0$, $H(\mu )>0$ in $({\mu}_{1},{\mu}_{2})$. Thus, for large *i*, we have${\lambda}_{1}({\mu}_{i},{\tau}_{i})<0$, which contradicts the assumption.

which also leads to ${\lambda}_{1}({\mu}_{i},{\tau}_{i})<0$ for large *i* combining with${\tau}_{i}\to {0}^{+}$ and ${\mu}_{i}\in (\underline{\mu}({\tau}_{i}),\overline{\mu}({\tau}_{i}))$. The contradiction completes the proof. □

**Remark 4.1** Theorem 4.1 is contrary to the stability results in [21]. But it is shown that the two competition species with differentintraspecific competition rates can also coexist for bigger regions of *μ*,which is similar to [21].

## Declarations

### Acknowledgements

The authors would like to express sincere thanks to the anonymous referee for his/hercarefully reading the manuscript and valuable comments on this paper. The work issupported by the Education Committee Foundation of Shaanxi Province, and the NationalNatural Science Foundation of China (No. 11271236).

## Authors’ Affiliations

## References

- Bertness MD:
**Intraspecific competition and facilitation in a northern acorn barnaclepopulation.***Ecology*1989, 70: 257–268. 10.2307/1938431View ArticleGoogle Scholar - Bürger R:
**A multilocus analysis of intraspecific competition and stabilizing selection on aquantitative trait.***J. Math. Biol.*2005, 50: 355–396. 10.1007/s00285-004-0294-2MathSciNetView ArticleMATHGoogle Scholar - Svanbäck R, Bolnick D:
**Intraspecific competition drives increased resource use diversity within a naturalpopulation.***Proc. R. Soc. B*2007, 274: 839–844. 10.1098/rspb.2006.0198View ArticleGoogle Scholar - Lankau RA: Rapid evolutionary change and the coexistence of species.
*Annu. Rev. Ecol. Evol. Syst.*2011, 42: 335–354. 10.1146/annurev-ecolsys-102710-145100View ArticleGoogle Scholar - Hastings A: Spatial heterogeneity and ecological models.
*Ecology*1990, 71: 426–428. 10.2307/1940296View ArticleGoogle Scholar - Takeuchi Y:
*Global Dynamical Properties of Lotka-Volterra Systems*. World Scientific, Singapore; 1996.View ArticleMATHGoogle Scholar - Du YH: Realization of prescribed patterns in the competition model.
*J. Differ. Equ.*2003, 193: 147–179. 10.1016/S0022-0396(03)00056-1View ArticleMathSciNetMATHGoogle Scholar - Cantrell RS, Cosner C, Lou Y:
**Multiple reversals of competitive dominance in ecological reserves via externalhabitat degradation.***J. Dyn. Differ. Equ.*2004, 16: 973–1010. 10.1007/s10884-004-7831-yMathSciNetView ArticleMATHGoogle Scholar - Győri I: Global attractivity in delay Lotka-Volterra differential equations.
*Fourth International Conference on Dynamic Systems and Applications*2003, 21–24. Atlanta, 21–24 May 2003Google Scholar - Pao CV:
**Global asymptotic stability of Lotka-Volterra competition systems with diffusionand time delays.***Nonlinear Anal., Real World Appl.*2004, 5: 91–104. 10.1016/S1468-1218(03)00018-XMathSciNetView ArticleMATHGoogle Scholar - Kuto K, Yamada Y:
**Positive solutions for Lotka-Volterra competition systems with largecross-diffusion.***Appl. Anal.*2010, 89: 1037–1066. 10.1080/00036811003627534MathSciNetView ArticleMATHGoogle Scholar - Hetzer G, Nguyen T, Shen W:
**Coexistence and extinction in the Volterra-Lotka competition model with nonlocaldispersal.***Commun. Pure Appl. Math.*2012, 11: 1699–1722.MathSciNetMATHGoogle Scholar - Wu DY:
**Global attractivity of a discrete Lotka-Volterra competition system with infinitedelays and feedback controls.***Adv. Differ. Equ.*2013., 2013: Article ID 14Google Scholar - He XQ, Ni WM:
**The effects of diffusion and spatial variation in Lotka-Volterracompetition-diffusion system I: heterogeneity vs. homogeneity.***J. Differ. Equ.*2013, 254: 528–546. 10.1016/j.jde.2012.08.032MathSciNetView ArticleMATHGoogle Scholar - He XQ, Ni WM:
**The effects of diffusion and spatial variation in Lotka-Volterracompetition-diffusion system II: the general case.***J. Differ. Equ.*2013, 254: 4088–4108. 10.1016/j.jde.2013.02.009MathSciNetView ArticleMATHGoogle Scholar - Meng XZ, Liu R, Zhang TH:
**Adaptive dynamics for a non-autonomous Lotka-Volterra model with size-selectivedisturbance.***Nonlinear Anal., Real World Appl.*2014, 16: 202–213.MathSciNetView ArticleMATHGoogle Scholar - Dockery J, Hutson V, Mischaikow K, Pernarowski M: The evolution of slow dispersal rates: a reaction-diffusion model.
*J. Math. Biol.*1998, 37: 61–83. 10.1007/s002850050120MathSciNetView ArticleMATHGoogle Scholar - Hutson V, Mischaikow K, Poláčik P: The evolution of dispersal rates in a heterogeneous time-periodic environment.
*J. Math. Biol.*2001, 43: 501–533. 10.1007/s002850100106MathSciNetView ArticleMATHGoogle Scholar - Hutson V, Lou Y, Mischaikow K, Poláčik P: Competing species near the degenerate limit.
*SIAM J. Math. Anal.*2003, 35: 453–491. 10.1137/S0036141002402189MathSciNetView ArticleMATHGoogle Scholar - Hetzer G, Nguyen T, Shen W:
**Effects of small spatial variation of the reproduction rate in a two speciescompetition model.***Electron. J. Differ. Equ.*2010., 2010: Article ID 160Google Scholar - Lou Y, Martínez S, Poláčik P: Loops and branches of coexistence states in a Lotka-Volterra competition model.
*J. Differ. Equ.*2006, 230: 720–742. 10.1016/j.jde.2006.04.005View ArticleMathSciNetMATHGoogle Scholar - Cantrell RS, Cosner C:
*Spatial Ecology via Reaction-Diffusion Equations*. Wiley, Chichester; 2003.MATHGoogle Scholar - Deimling K:
*Nonlinear Functional Analysis*. Springer, Berlin; 1985.View ArticleMATHGoogle Scholar - Hess P:
*Periodic Parabolic Boundary Value Problems and Positivity*. Longman, Harlow; 1985.MATHGoogle Scholar - Golubitsky M, Schaeffer D I. In
*Singularities and Groups in Bifurcation Theory*. Springer, New York; 1985.View ArticleGoogle Scholar - Kato T:
*Perturbation Theory for Linear Operators*. Springer, New York; 1985.Google Scholar

## Copyright

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative CommonsAttribution License (http://creativecommons.org/licenses/by/4.0), which permitsunrestricted use, distribution, and reproduction in any medium, provided the originalwork is properly credited.