JAM Journal of Applied Mathematics 1687-0042 1110-757X Hindawi Publishing Corporation 10.1155/2014/314729 314729 Research Article Global Stability for a Three-Species Food Chain Model in a Patchy Environment Li Hongli Jiang Yaolin Zhang Long Teng Zhidong Song Xinyu 1 College of Mathematics and System Sciences Xinjiang University Urumqi 830046 China xju.edu.cn 2014 682014 2014 10 06 2014 03 08 2014 6 8 2014 2014 Copyright © 2014 Hongli Li et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

We investigate a three-species food chain model in a patchy environment where prey species, mid-level predator species, and top predator species can disperse among n different patches ( n 2 ) . By using the method of constructing Lyapunov functions based on graph-theoretical approach for coupled systems, we derive sufficient conditions under which the positive equilibrium of this model is unique and globally asymptotically stable if it exists.

1. Introduction

Coupled systems on networks are used to describe a wide variety of physical, natural, and artificial complex dynamical systems, such as neural networks, biological systems, and the spread of infectious diseases in heterogeneous populations (see  and the references therein). A mathematical description of a network is a directed graph consisting of vertices and directed arcs connecting them. At each vertex, the dynamics are given by a system of differential equations called vertex system. The directed arcs indicate interconnections and interactions among vertex systems.

Diffusion in patchy environment is one of the most prevalent phenomena of nature. Since the spatiotemporal heterogeneity incurs great impacts on the species’ diversity, structure, and genetical polymorphism, scientists of biology, ecology, and biomathematics paid great attention on the population dynamics with diffusion. The stability of equilibrium is the precondition of applications of dispersal models in practice. Therefore, there are a great amount of literatures on this topic (see  and the references therein). In , Kuang and Takeuchi considered a predator-prey model in which preys disperse among two patches and proved the uniqueness and global stability of a positive equilibrium by constructing a Lyapunov function.

Recently, a graph theoretic approach was proposed to construct Lyapunov functions for some general coupled systems of ordinary differential equations on networks, and the global stability was explored in [11, 12]. We refer to [13, 14] for recent applications.

In , Li and Shuai considered the following predator-prey model where prey species disperse among n patches ( n 2 ) : (1) x ˙ i = x i ( r i - b i x i - e i y i ) + j = 1 n d i j ( x j - α i j x i ) , y ˙ i = y i ( - γ i - δ i y i + ε i x i ) , where i = 1,2 , , n . They provided a systematic method for constructing a global Lyapunov function for the coupled systems on networks and then gave some sufficient conditions of stability for system (1). In fact, there may be more species in some habitats and they can construct a food chain; in this case it is more realistic to consider a multiple species predator-prey system. Based on this fact, in this paper, we investigate the following three-species food chain model in a patchy environment: (2) x ˙ i = x i ( r i - σ i x i - e i y i ) + j = 1 n a i j x j - j = 1 n a j i x i , y ˙ i = y i ( - θ i - δ i y i + ε i x i - ρ i z i ) + j = 1 n b i j y j - j = 1 n b j i y i , i = 1,2 , , n , z ˙ i = z i ( - γ i - α i z i + η i y i ) + j = 1 n c i j z j - j = 1 n c j i z i , where x i , y i , and z i denote the densities of prey species, mid-level predator species, and top predator species, respectively; all the parameters are nonnegative constants, and ε i , e i , η i , and ρ i are positive. Constant a i j 0 is the dispersal rate of prey species from patch j to patch i , constant b i j 0 is the dispersal rate of mid-level predator species from patch j to patch i , and constant c i j 0 is the dispersal rate of top predator species from patch j to patch i . We refer the reader to [10, 15] for interpretations of predator-prey models and parameters.

This paper is organized as follows. In Section 2, we introduce some preliminaries on graph theory which will be used in Section 3. In Section 3, the global stability of the positive equilibrium of system (2) is proved. Finally, a conclusion is given in Section 4.

2. Preliminaries

Since the coupled system considered in this paper is built on a directed graph, the following basic concepts and theorems on graph theory can be found in .

A directed graph or digraph G = ( V , E ) contains a set V = { 1,2 , , n } of vertices and a set E of arcs ( i , j ) leading from initial vertex i to terminal vertex j . A subgraph H of G is said to be spanning if H and G have the same vertex set. A digraph G is weighted if each arc ( j , i ) is assigned a positive weight a i j . Here a i j > 0 if and only if there exists an arc from vertex j to vertex i in G . The weight w ( H ) of a subgraph H is the product of the weights on all its arcs.

Given a weighted digraph G with n vertices, define the weight matrix A = ( a i j ) n × n whose entry a i j equals the weight of arc ( j , i ) if it exists and 0 otherwise. For our purpose, we denote a weighted digraph as ( G , A ) . A digraph G is strongly connected if, for any pair of distinct vertices, there exists a directed path from one to the other. A weighted digraph ( G , A ) is strongly connected if and only if the weight matrix A is irreducible. The Laplacian matrix of ( G , A ) is defined as (3) L = ( k 1 a 1 k - a 12 - a 1 n - a 21 k 2 a 2 k - a 2 n - a n 1 - a n 2 k n a n k ) . Let c i denote the cofactor of i th diagonal element of L . The following results are listed as follows from .

Proposition 1 (see [<xref ref-type="bibr" rid="B12">12</xref>]).

Assume n 2 . Then, (4) c i = T T i w ( T ) , i = 1,2 , , n , where T i is the set of all spanning trees T of ( G , A ) that are rooted at vertex i and w ( T ) is weight of T . In particular, if ( G , A ) is strongly connected, then c i > 0 for 1 i n .

Theorem 2 (see [<xref ref-type="bibr" rid="B12">12</xref>]).

Assume n 2 . Let c i be given in Proposition 1. Then the following identity holds: (5) i , j = 1 c i a i j G i ( x i ) = i , j = 1 c i a i j G j ( x j ) , where G i ( x i ) , 1 i n , are arbitrary functions.

3. Main Results

In this section, the stability for the positive equilibrium of a three-species food chain model in a patchy environment is considered.

Theorem 3.

Assume that a positive equilibrium E * = ( x 1 * , y 1 * , z 1 * , , x n * , y n * , z n * ) exists for system (2) and the following assumptions hold.

Dispersal matrixes A = ( a i j ) n × n , B = ( b i j ) n × n , and C = ( c i j ) n × n are irreducible.

There exist nonnegative constants λ and β such that ε i a i j x j * = λ e i b i j y j * and ( e i ρ i / η i ) z j * = β e i b i j y j * for all 1 i , j n .

Then the positive equilibrium E * of system (2) is unique and globally asymptotically stable in R + 3 n .

Proof.

From equilibrium equations of (2), we obtain (6) σ i x i * + e i y i * - j = 1 n a i j x j * x i * + j = 1 n a j i = r i , ε i x i * - δ i y i * - ρ i z i * + j = 1 n b i j y j * y i * - j = 1 n b j i = θ i , - α i z i * + η i y i * + j = 1 n c i j z j * z i * - j = 1 n c j i = γ i .

Next, we show that E * is globally asymptotically stable in R + 3 n , and thus it is unique. Let (7) V i ( x i , y i , z i ) = ε i ( x i - x i * - x i * ln x i x i * ) + e i ( y i - y i * - y i * ln y i y i * ) + e i ρ i η i ( z i - z i * - z i * ln z i z i * ) . Note that 1 - x + ln x 0 for x > 0 and equality holds if and only if x = 1 . Differentiating V i along the solution of system (2), we obtain (8) V ˙ i = ε i ( x i - x i * ) ( r i - σ i x i - e i y i + j = 1 n a i j x j x i - j = 1 n a j i ) + e i ( y i - y i * ) ( - θ i - δ i y i + ε i x i - ρ i z i j = 1 n 0000000000000 + j = 1 n b i j y j y i - j = 1 n b j i ) + e i ρ i η i ( z i - z i * ) ( - γ i - α i z i + η i y i + j = 1 n c i j z j z i - j = 1 n c j i ) = ε i ( x i - x i * ) × [ - σ i ( x i - x i * ) - e i ( y i - y i * ) + j = 1 n a i j ( x j x i - x j * x i * ) ] + e i ( y i - y i * ) × [ - δ i ( y i - y i * ) + ε i ( x i - x i * ) - ρ i ( z i - z i * ) j = 1 n 000 + j = 1 n b i j ( y j y i - y j * y i * ) ] + e i ρ i η i ( z i - z i * ) × [ - α i ( z i - z i * ) + η i ( y i - y i * ) + j = 1 n c i j ( z j z i - z j * z i * ) ] j = 1 n ε i a i j x j * ( x j x j * - x i x i * - x i * x j x i x j * + 1 ) + j = 1 n e i b i j y j * ( y j y j * - y i y i * - y i * y j y i y j * + 1 ) + j = 1 n e i ρ i η i c i j z j * ( z j z j * - z i z i * - z i * z j z i z j * + 1 ) = j = 1 n ε i a i j x j * ( x j x j * - x i x i * - x i * x j x i x j * + 1 000000000 + ln x i x i * - ln x j x j * + ln x i * x j x i x j * ) + j = 1 n e i b i j y j * ( y j y j * - y i y i * - y i * y j y i y j * + 1 0000000 i 0000 + ln y i y i * - ln y j y j * + ln y i * y j y i y j * ) + j = 1 n e i ρ i η i c i j z j * ( z j z j * - z i z i * - z i * z j z i z j * + 1 000000000 i 0000 + ln z i z i * - ln z j z j * + ln z i * z j z i z j * ) j = 1 n ε i a i j x j * ( x j x j * - x i x i * + ln x i x i * - ln x j x j * ) + j = 1 n e i b i j y j * ( y j y j * - y i y i * + ln y i y i * - ln y j y j * ) + j = 1 n e i ρ i η i c i j z j * ( z j z j * - z i z i * + ln z i z i * - ln z j z j * ) = j = 1 n e i b i j y j * 000 × [ ( λ x j x j * + λ ln x j * x j + β z j z j * + β ln z j * z j i 0000000 + y j y j * + ln y j * y j ) 00000 i 0 - ( λ x i x i * + λ ln x i * x i + β z i z i * + β ln z i * z i 000000000 i + y i y i * + ln y i * y i ) ] = j = 1 n e i b i j y j * [ G j ( x j , y j , z j ) - G i ( x i , y i , z i ) ] , where (9) G i ( x i , y i , z i ) = λ x i x i * + λ ln x i * x i + β z i z i * + β ln z i * z i + y i y i * + ln y i * y i . Consider a weight matrix W = ( w i j ) n × n with entry w i j = e i b i j y j * , and denote the corresponding weighted digraph as ( G , W ) . Let (10) L - = ( k 1 w 1 k - w 12 - w 1 n - w 21 k 2 w 2 k - w 2 n - w n 1 - w n 2 k n w n k ) . Note that L - is the Laplacian matrix of ( G , W ) . Since ( b i j ) n × n is irreducible, we know that ( w i j ) n × n is irreducible; by Proposition 1, we obtain that c i > 0 for all i . Then, from Theorem 2, the following identity holds: (11) i = 1 n c i j = 1 n e i b i j y j * ( G j ( x j , y j , z j ) - G i ( x i , y i , z i ) ) = 0 . Set (12) V ( x 1 , y 1 , z 1 , x 2 , y 2 , z 2 , , x n , y n , z n ) = i = 1 n c i V i ( x i , y i , z i ) . Using (8) and (11), we obtain (13) V ˙ = i = 1 n c i V ˙ i i = 1 n c i j = 1 n e i b i j y j * ( G j ( x j , y j , z j ) - G i ( x i , y i , z i ) ) = 0 for all ( x 1 , y 1 , z 1 , x 2 , y 2 , z 2 , , x n , y n , z n ) R + 3 n . Therefore, V = i = 1 n c i V i as defined in Theorem 3.1 of  is a Lyapunov function for the system (2); namely, V ˙ 0 for all ( x 1 , y 1 , z 1 , x 2 , y 2 , z 2 , , x n , y n , z n ) R + 3 n ; V ˙ = 0 implies that x i = x i * , y i = y i * , and z i = z i * for all i . By LaSalle Invariance Principle , E * is globally asymptotically stable in R + 3 n ; this also implies that E * is unique in R + 3 n . This completes the proof of Theorem 3.

Remark 4.

A = ( a i j ) n × n , B = ( b i j ) n × n , and C = ( c i j ) n × n are dispersal matrices; a typical assumption we impose on these matrices is that they are irreducible. In biological terms, this means individuals in each patch can disperse between any two patches directly or indirectly.

Applying the similar proof as that for Theorem 3, we have the following corollaries.

Corollary 5.

Consider the model (14) x ˙ i = x i ( r i - σ i x i - e i y i ) , y ˙ i = y i ( - θ i - δ i y i + ε i x i - ρ i z i ) + j = 1 n b i j y j - j = 1 n b j i y i , i = 1,2 , , n , z ˙ i = z i ( - γ i - α i z i + η i y i ) , where all the parameters are nonnegative constants, ε i , e i , η i , and ρ i are positive, and the dispersal matrix B = ( b i j ) n × n is irreducible. Then, if a positive equilibrium E * exists in (14), it is unique and globally asymptotically stable in the positive cone R + 3 n .

Corollary 6.

Consider the model (15) x ˙ i = x i ( r i - σ i x i - e i y i ) + j = 1 n a i j x j - j = 1 n a j i x i , y ˙ i = y i ( - θ i - δ i y i + ε i x i - ρ i z i ) , i = 1,2 , , n , z ˙ i = z i ( - γ i - α i z i + η i y i ) , where all the parameters are nonnegative constants, ε i , e i , η i , and ρ i are positive, and the dispersal matrix A = ( a i j ) n × n is irreducible. Then, if a positive equilibrium E ~ * exists in (15), it is unique and globally asymptotically stable in the positive cone R + 3 n .

4. Conclusion

In this paper, we generalize the model of the n -patch predator-prey model of  to a three-species food chain model where prey species, mid-level predator species, and top predator species can disperse among n different patches ( n 2 ) . Our proof of global stability of the positive equilibrium utilizes a graph-theoretical approach to the method of Lyapunov function.

Biologically, Theorem 3 implies that if a three-species food chain system is dispersing among strongly connected patches (which is equivalent to the irreducibility of the dispersal matrixes) and if the system is permanent (which guarantees the existence of positive equilibrium), then the numbers of prey species, mid-level predator species, and top predator species in each patch will eventually be stable at some corresponding positive values given the well-coupled dispersal (condition ( H 2 ) of Theorem 3).

Corollaries 5 and 6 imply that if a three-species food chain system is dispersing among strongly connected patches (only one species can disperse among n different patches ( n 2 ) ) and if the system is permanent, then the numbers of prey species, mid-level predator species, and top predator species in each patch will eventually be stable at some corresponding positive values.

Theorem 3 requires the extra condition ( H 2 ) ; the global stability for the positive equilibrium of system (2) without condition ( H 2 ) is still unclear. It remains an interesting problem for a three-species food chain model in patchy environment.

Conflict of Interests

The authors declare that there is no conflict of interests.

Acknowledgments

This work is supported by National Natural Science Foundation of China (Grant nos. 11371287, 11361059), the International Science and Technology Cooperation Program of China (Grant no. 2010DFA14700), Xinjiang Introduction Plan Project of High Level Talents, the Natural Science Foundation of Xinjiang (Grant no. 2012211B07), and the Scientific Research Programmes of Colleges in Xinjiang (XJEDU2013I03).

Bishop C. M. Neural Networks for Pattern Recogniton 1995 Oxford, UK Oxford University Press MR1385195 Dai Y. Cai Y. Xu X. Synchronization criteria for complex dynamical networks with neutral-type coupling delay Physica A 2008 387 18 4673 4682 10.1016/j.physa.2008.03.024 2-s2.0-44149125249 May R. M. Stability and Complexity in Model Ecosystems 2001 Princeton, NJ, USA Princeton University Press Thieme H. R. Mathematics in Population Biology 2003 Princeton, NJ, USA Princeton University Press MR1993355 Awrejcewicz J. Bifurcation and Chaos in Coupled Oscillators 1991 New Jersey, NJ, USA World Scientific Publishing 10.1142/1342 MR1178489 Barabási A. L. Jeong H. Néda Z. Ravasz E. Schubert A. Vicsek T. Evolution of the social network of scientific collaborations Physica A 2002 311 3-4 590 614 10.1016/S0378-4371(02)00736-7 MR1943379 2-s2.0-0037102646 Beretta E. Takeuchi Y. Global stability of single-species diffusion Volterra models with continuous time delays Bulletin of Mathematical Biology 1987 49 4 431 448 10.1016/S0092-8240(87)80005-8 MR908159 ZBL0627.92021 2-s2.0-0023083590 Hastings A. Dynamics of a single species in a spatially varying environment: the stabilizing role of high dispersal rates Journal of Mathematical Biology 1982 16 1 49 55 10.1007/BF00275160 MR693372 2-s2.0-0020420302 Lu Z. Y. Takeuchi Y. Global asymptotic behavior in single-species discrete diffusion systems Journal of Mathematical Biology 1993 32 1 67 77 10.1007/BF00160375 MR1256831 Kuang Y. Takeuchi Y. Predator-prey dynamics in models of prey dispersal in two-patch environments Mathematical Biosciences 1994 120 1 77 98 10.1016/0025-5564(94)90038-8 MR1281745 ZBL0793.92014 2-s2.0-0028359362 Guo H. Li M. Y. Shuai Z. A graph-theoretic approach to the method of global Lyapunov functions Proceedings of the American Mathematical Society 2008 136 8 2793 2802 10.1090/S0002-9939-08-09341-6 MR2399043 ZBL1155.34028 2-s2.0-70349388314 Li M. Y. Shuai Z. Global-stability problem for coupled systems of differential equations on networks Journal of Differential Equations 2010 248 1 1 20 10.1016/j.jde.2009.09.003 MR2557892 ZBL1190.34063 2-s2.0-70349785039 Li M. Y. Shuai Z. Wang C. Global stability of multi-group epidemic models with distributed delays Journal of Mathematical Analysis and Applications 2010 361 1 38 47 10.1016/j.jmaa.2009.09.017 MR2567280 ZBL1175.92046 2-s2.0-70349384940 Shu H. Fan D. Wei J. Global stability of multi-group {SEIR} epidemic models with distributed delays and nonlinear transmission Nonlinear Analysis: Real World Applications 2012 13 4 1581 1592 10.1016/j.nonrwa.2011.11.016 MR2890994 2-s2.0-84856551488 Freedman H. I. Deterministic Mathematical Models in Population Ecology 1980 New York, NY, USA Marcel Dekker MR586941 La Salle J. P. The Stability of Dynamical Systems 1976 Philadelphia, Pa, USA SIAM CBMS-NSF Regional Conference Series in Applied Mathematics