Hindawi

Discrete Dynamics in Nature and Society

Volume 2018, Article ID 6519696, 9 pages
https://doi.org/10.1155/2018/6519696

Research Article
Dynamics and Patterns of a Diffusive Prey-Predator System with
a Group Defense for Prey

1 2
Honglan Zhu and Xuebing Zhang
1
Business School, Huaiyin Institute of Technology, Huaian, China
2
College of Mathematics and Statistics, Nanjing University of Information Science and Technology, Nanjing, China

Correspondence should be addressed to Xuebing Zhang; zxb1030@163.com

Received 11 September 2017; Accepted 26 November 2017; Published 8 January 2018

Academic Editor: Chris Goodrich

Copyright © 2018 Honglan Zhu and Xuebing Zhang. 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 study a diffusive prey-predator system with a group defense for prey. Under Neumann boundary condition, we analyze local
and stability of nonnegative constant steady states and the existence and nonexistence of nonconstant steady states. These results
also exhibit the critical role of the system parameters leading to the formation of spatiotemporal patterns.

1. Introduction straight line through the origin and is unbounded. Thus,

more reasonable response functions should be nonlinear
The predator-prey system first proposed by [1, 2] is one of the and bounded. In 1913, Michaelis and Menten proposed the
fundamental ecological systems in both ecology and mathe- response function
matical ecology. Based on different settings, various types of
predator-prey models described by differential systems have 𝑚𝑥𝑦
𝑃 (𝑥, 𝑦) = , (3)
been proposed and the dynamics of these systems are studied 𝑎+𝑥
[3–6]. The basic form of these models is as follows:
where 𝑚 > 0 denotes the maximal growth rate of the species
𝑑𝑥 𝑥 and 𝑎 > 0 is the half-saturation constant. It is now referred to
= 𝑟𝑥 (1 − ) − 𝑃 (𝑥, 𝑦) ,
𝑑𝑡 𝐾 as a Michaelis-Menten function or a Holling type II function.
(1)
𝑑𝑦 Another class of response function is
= −𝑠𝑦 + 𝑐𝑃 (𝑥, 𝑦) ,
𝑑𝑡
𝑚𝑥2 𝑦
where 𝑟 is the intrinsic growth rate and 𝐾 is the environ- 𝑃 (𝑥, 𝑦) = (4)
𝑎 + 𝑏𝑥 + 𝑥2
mental carrying capacity of prey population, and the function
𝑃(𝑥) is the functional response; the constant 𝑐(>0) is the which is called a sigmoidal response function, while the
ratio of biomass conversion and 𝑠 is the natural death rate of simplification
predator species. The simplest functional response is Lotka-
Volterra function which is described as 𝑚𝑥2 𝑦
𝑃 (𝑥, 𝑦) = (5)
𝑘 𝑎 + 𝑥2
{
{𝑎𝑥𝑦, 0 ≤ 𝑥 ≤ 𝑎 ,
𝑃 (𝑥, 𝑦) = { (2) is known as a Holling type III function. Some authors [7, 8]
{ 𝑘
𝑘𝑦, 𝑥≥ , considered system (1) with following response function:
{ 𝑎
which is also called Holling type I function. However, the 𝑚𝑥𝑦
𝑃 (𝑥, 𝑦) = (6)
curve defined by the Lotka-Volterra response function is a 𝑎 + 𝑥2
2 Discrete Dynamics in Nature and Society

which is called Holling type IV function. Besides, Bedding- which implies that (8) is a closed system and has no flux across
ton-DeAngelis type 𝑃(𝑥, 𝑦) = 𝛽𝑥𝑦/(𝛼 + 𝑥 + 𝑚𝑦) and more the boundary 𝜕Ω.
complicated functional response 𝑃(𝑥, 𝑦) = 𝛽𝑥2 𝑦/(𝑥2 + 𝑚𝑦2 ) The structure of this paper is arranged as follows. In
are also considered by some researchers [9, 10]. Section 2, we estimate the a priori bounds of positive
Recently, some works consider the case when animals solutions of (7). In Section 3, the local and global stabilities
join together in herds in order to provide a self-defense of nonnegative constant steady states of (7) are discussed. In
from predators. In [11], the authors argued that it is more Section 4, we give a priori estimate for the positive solutions
appropriate to model the response functions of prey that of (8) by using maximum principle and Harnack inequality.
exhibit herd behavior in terms of the square root of the In Section 5, we give a nonexistence result of nonconstant
prey population. Inspired by this thought, the authors in [12] solutions of (8). In Section 6, we consider the existence
choose response function 𝑃(𝑥) = √𝑥 to reflect this fact. of nonconstant positive solutions of (8). Finally, to support
When motion is allowed, [13] considered the spatiotemporal our theoretical predictions, some numerical simulations are
behavior of a prey-predator system with a group defense given.
for prey by means of extensive computer simulations. The
proposed model is as follows: 2. Basic Dynamics and a Priori Bound
𝜕𝑢 𝑢
= 𝑑1 Δ𝑢 + 𝑟1 𝑢 (1 − ) − 𝜌𝑢𝛼 V, Theorem 1. For system (7), one has the following.
𝜕𝑡 𝐾
(a) If 𝑢0 (𝑥) ≥ 0, V0 (𝑥) ≥ 0, then system (7) has a unique
(𝑥, 𝑡) ∈ Ω × (0, +∞) ,
solution (𝑢(𝑡, 𝑥), V(𝑡, 𝑥)) such that 𝑢(𝑡, 𝑥) > 0, V(𝑡, 𝑥) >
𝜕V 0 for 𝑡 ∈ (0, +∞) and 𝑥 ∈ Ω.
= 𝑑2 ΔV + 𝛽𝑢𝛼 V − 𝑟2 V, (𝑥, 𝑡) ∈ Ω × (0, +∞) (7)
𝜕𝑡 (b) Any solution (𝑢, V) of (7) satisfies
𝑢 (𝑥, 0) = 𝑢0 (𝑥) , lim sup 𝑢 (𝑡, 𝑥) ≤ 𝐾,
𝑡→+∞
V (𝑥, 0) = V0 (𝑥) , (9)
𝛽 𝑟
𝑥 ∈ Ω, ∫ V (𝑥, 𝑡) 𝑑𝑥 ≤ (1 + 1 ) 𝐾 |Ω| .
Ω 𝜌 𝑟2
where 𝑢 and V denote, respectively, the densities of prey and
predator species. 𝑟 is the growth rate of prey species, 𝐾 is its Proof. (a) Define
carrying capacity, 𝑟2 is the mortality rate of predator species,
𝑢
𝜌 is the search efficiency of predator for prey, 𝛽 is the biomass 𝑓 (𝑢, V) = 𝑟1 𝑢 (1 − ) − 𝜌𝑢𝛼 V,
conversion coefficient, and 𝛼 ∈ (0, 1) represents a kind 𝐾 (10)
of aggregation efficiency. The local dynamics for nonspatial 𝑔 (𝑢, V) = 𝛽𝑢𝛼 V − 𝑟2 V.
model was studied, such as Hopf bifurcation and existence of
extinction domain. For model (7), the authors only give some Then 𝑓V = −𝜌𝑢𝛼 ≤ 0 and 𝑔𝑢 = 𝛽𝛼𝑢𝛼−1 V ≥ 0 in R2+ =
numerical simulations to find some spatiotemporal features. {𝑢 ≥ 0, V ≥ 0}. Hence, (7) is a mixed qusi-monotone system.
Reference [14] considers the direction and the stability of the Consider following system:
bifurcating periodic solutions for model (7) with 𝛼 = 1/2
under Neumann boundary conditions. Reference [15] inves- 𝑑𝑢 𝑢
tigated the global dynamics of nonspatial model including = 𝑟1 𝑢 (1 − ) − 𝜌𝑢𝛼 V,
𝑑𝑡 𝐾
the nonexistence of periodic orbits and the existence and
𝑑V
uniqueness of limit cycles. We refer readers to [16–21] as = 𝛽𝑢𝛼 V − 𝑟2 V, (11)
some other related works on predator-prey model with herd 𝑑𝑡
behavior. 𝑢 (0) = 𝑢0 ,
It is noted that up to now no one has studied the existence
and nonexistence of positive steady state solutions of (7). V (0) = V0 .
Therefore, the main aim of this article is to study the existence
and nonexistence of nonconstant positive solutions of the Assume 𝑢(𝑡; 𝑢0 , V0 ), V(𝑡; 𝑢0 , V0 ) are the unique solution to
following elliptic system: system (11). Let
𝑢 max 𝑢0 (𝑥) = 𝜙𝑀,
−𝑑1 Δ𝑢 = 𝑟1 𝑢 (1 − ) − 𝜌𝑢𝛼 V, in Ω,
𝐾 Ω
(12)
−𝑑2 ΔV = 𝛽𝑢𝛼 V − 𝑟2 V, in Ω, (8) max V0 (𝑥) = 𝜓𝑀.
Ω
𝜕𝑢 𝜕V
= = 0, on 𝜕Ω, Obviously, (𝑢(𝑡, 𝑥), V(𝑡, 𝑥)) = (0, 0) and (𝑢(𝑡), V(𝑡)) = (𝑢(𝑡;
𝜕] 𝜕] 𝜙𝑀, 𝜓𝑀), V(𝑡; 𝜙𝑀, 𝜓𝑀)) are a pair of lower-solution and
where ] is the outward unit normal vector on 𝜕Ω, and we upper-solution to system (7). Therefore, according to the
impose a homogeneous Neumann type boundary condition, Theorem 8.3.3 in [22] or Theorem 5.3.2 in [23], system (7)
Discrete Dynamics in Nature and Society 3

has a unique globally defined solution (𝑢(𝑥, 𝑡), V(𝑥, 𝑡)) which see that the possible nonnegative constant steady states of (7)
satisfies are
𝐸0 = (0, 0) ,
0 ≤ 𝑢 (𝑥, 𝑡) ≤ 𝑢 (𝑡; 𝜙𝑀, 𝜓𝑀) ,
(13) 𝐸1 = (𝐾, 0) , (18)
0 ≤ V (𝑥, 𝑡) ≤ V (𝑡; 𝜙𝑀, 𝜓𝑀) .
∗ ∗ ∗
𝐸 = (𝑢 , V ) ,
The strong maximum principle implies that 𝑢(𝑥, 𝑡), V(𝑥, 𝑡) >
0 when 𝑡 > 0 for all 𝑥 ∈ Ω. where 𝑢∗ = (𝑟2 /𝛽)1/𝛼 , V∗ = 𝑟1 𝑢∗(1−𝛼) (1−𝑢∗ /𝐾)/𝜌. Obviously,
(b) By the first equation of (7), we easily obtain the fact the positive constant steady state 𝐸∗ exists if 𝑟2 < 𝛽𝐾𝛼 holds.
that 𝑟1 𝑢(1−𝑢/𝐾)−𝜌𝑢𝛼 V ≤ 𝑟1 𝑢(1−𝑢/𝐾) in [0, +∞)×Ω; the first
result follows easily from the simple comparison argument Notation 1. Let 0 = 𝜇0 < 𝜇1 < 𝜇2 < ⋅ ⋅ ⋅ < 𝜇𝑛 <
for parabolic problems, and thus there exists 𝑇 ∈ (0, +∞) ⋅ ⋅ ⋅ → ∞ be the eigenvalues of −Δ on Ω under homogeneous
such that 𝑢(𝑡, 𝑥) ≤ 𝐾 + 𝜖 in [𝑇, +∞) × Ω for an arbitrary Neumann boundary condition. We define the following space
constant 𝜖 > 0. decomposition:
For the estimate of V(𝑥, 𝑡), let 𝑈(𝑡) = ∫Ω 𝑢(𝑥, 𝑡)𝑑𝑥, 𝑉(𝑡) = (i) 𝑆(𝜇𝑛 ) is the space of eigenfunctions corresponding to
∫Ω V(𝑥, 𝑡)𝑑𝑥; then 𝜇𝑖 for 𝑛 = 0, 1, 2, . . ..
(ii) 𝑋𝑖𝑗 fl {c ⋅ 𝜙𝑖𝑗 : c ∈ R2 }, where {𝜙𝑖𝑗 } are orthonormal
𝑑𝑈 basis of 𝑆(𝜇𝑛 ) for 𝑗 = 1, 2, . . . , dim[𝑆(𝜇𝑛 )].
= ∫ 𝑢𝑡 𝑑𝑥
𝑑𝑡 Ω (iii) X fl {u = (𝑢, V) ∈ [𝐶1 (Ω)]2 : 𝜕𝑢/𝜕n = 𝜕V/𝜕n = 0},
dim[𝑆(𝜇𝑗 )]
𝑢 and so X = ⊕∞
𝑖=1 X𝑖 , where X𝑖 = ⊕𝑗=1 X𝑖𝑗 .
= 𝑑1 ∫ Δ𝑢 𝑑𝑥 + ∫ [𝑟1 𝑢 (𝑢 − ) − 𝜌𝑢𝛼 V] 𝑑𝑥 (14)
Ω Ω 𝐾
𝑢 Let 𝐸̃ be a nonnegative constant steady state of (7);
= ∫ [𝑟1 𝑢 (𝑢 − ) − 𝜌𝑢𝛼 V] 𝑑𝑥, then the linearization of (7) at a constant solution 𝐸̃ can be
Ω 𝐾
expressed by
𝑑𝑉
= ∫ V𝑡 𝑑𝑥 = 𝑑2 ∫ ΔV 𝑑𝑥 + ∫ (𝛽𝑢𝛼 V − 𝑟2 V) 𝑑𝑥 u𝑡 = (𝐷Δ + 𝐽) u, (19)
𝑑𝑡 Ω Ω Ω
(15)
where 𝐷 = diag(𝑑1 , 𝑑2 ), u = (𝑢(𝑥, 𝑡), V(𝑥, 𝑡))𝑇 , and
= ∫ (𝛽𝑢𝛼 V − 𝑟2 V) 𝑑𝑥.
Ω 2𝑟1 𝑢∗
𝑟1 − − 𝜌𝛼𝑢∗(𝛼−1) V∗ −𝜌𝑢∗𝛼
𝐽=( 𝐾 ). (20)
Multiplying (14) by 𝛽/𝜌 and adding it to (15), we have −𝛽V∗ 𝑢∗(𝛼−1) 𝛽𝑢∗𝛼 − 𝑟2
𝛽 𝛽𝑟 𝑢 In view of Notation 1, we can induce the eigenvalues of system
( 𝑈 + 𝑉) = −𝑟2 𝑉 + 1 ∫ 𝑢 (𝑢 − ) 𝑑𝑥 (19) confined on the subspace X𝑖 . If 𝜆 is an eigenvalue of (19)
𝜌 𝑡 𝜌 Ω 𝐾
on X𝑖 , it must be an eigenvalue of the matrix −𝜇𝑛 𝐷 + 𝐽 for
𝛽 each 𝑛 ≥ 0. It is easy to see that 𝜆 satisfies the characteristic
≤ −𝑟2 ( 𝑈 + 𝑉) (16) equation
𝜌
𝛽𝑟2 𝛽𝑟1 2𝑟1 𝑢∗
+( + ) 𝜌 (𝐾 + 𝜖) |Ω| . 𝜆2 + (𝑑1 𝜇𝑛 + 𝑑2 𝜇𝑛 − 𝑟1 + + 𝜌𝛼𝑢∗(𝛼−1) V∗ ) 𝜆
𝜌 𝜌 𝐾
2𝑟1 𝑢∗ (21)
Integration of the inequality leads to + 𝑑2 𝜇𝑛 (𝑑1 𝜇𝑛 − 𝑟1 + + 𝜌𝛼𝑢∗(𝛼−1) V∗ )
𝐾
𝛽 + 𝜌𝛽𝑢∗(2𝛼−1) V∗ = 0.
∫ V (𝑥, 𝑡) 𝑑𝑥 = 𝑉 (𝑡) < 𝑈 (𝑡) + 𝑉 (𝑡)
Ω 𝜌
(17) Theorem 2. (i) The trivial equilibrium 𝐸0 = (0, 0) is unstable.
𝛽 𝑟 (ii) If 𝛽𝐾𝛼 < 𝑟2 , then 𝐸1 = (𝐾, 0) is globally asymptotically
≤ (1 + 1 ) (𝐾 + 𝜖) |Ω| . stable.
𝜌 𝑟2
(iii) If 1 − 𝛼 − ((2 − 𝛼)/𝐾)𝑢∗ < 0, then 𝐸∗ is locally
asymptotically stable.

Proof. (i) For 𝐸0 = (0, 0), the eigenvalues are

3. Stability of the Nonnegative Constant
𝜆 1𝑛 = 𝑟1 − 𝑑1 𝜇𝑛 ,
Steady States of (7) (22)
𝜆 2𝑛 = −𝑑2 𝜇𝑛 .
In this section, we will analyze the stability of nonnegative
constant steady states of (7). By the direct computation, we Obviously, 𝐸0 is unstable.
4 Discrete Dynamics in Nature and Society

(ii) For 𝐸1 = (𝐾, 0), the eigenvalues are 4. The Prior Estimate
𝜆 1𝑛 = −𝑟1 − 𝑑1 𝜇𝑛 , In this section, we will give some a priori estimates of positive
(23) solutions to (8). Firstly, we give two known lemmas.
𝜆 2𝑛 = 𝛽𝐾𝛼 − 𝑟2 − 𝑑2 𝜇𝑛 .
Lemma 3 (Harnack inequality (cf. [24])). Let 𝜔 ∈ 𝐶2 (Ω) ∩
𝛼
If 𝛽𝐾 < 𝑟2 , then 𝜆 1𝑛 and 𝜆 2𝑛 are all negative. Therefore 𝐶1 (Ω) be a positive classical solution to
𝐸1 is locally asymptotically stable. Indeed, 𝐸1 is globally
asymptotically stable. Δ𝜔 (𝑥) + 𝑐 (𝑥) 𝜔 (𝑥) = 0 in Ω,
On account of Theorem 1, we have lim sup𝑡→+∞ maxΩ 𝑢(⋅, (31)
𝜕𝜔
𝑡) ≤ 𝐾, and thus there exists 𝑇1 ∈ (0, +∞) such that, for an = 0 on 𝜕Ω.
𝜕𝑛
arbitrary constant 0 < 𝜖 < (𝑟2 /𝛽)1/𝛼 − 𝐾,
Then there exists a positive constant 𝐶 such that
𝑢 (⋅, 𝑡) ≤ 𝐾 + 𝜖, 𝑡 ≥ 𝑇1 . (24)
max 𝜔 ≥ 𝐶 min 𝜔. (32)
Ω Ω
It follows from the second equation of (7) that
Lemma 4 (maximum principle (cf. [25])). Suppose that 𝑔 ∈
V𝑡 − 𝑑2 ΔV ≤ V (𝛽 (𝐾 + 𝜖)𝛼 − 𝑟2 ) , 𝑡 ≥ 𝑇1 . (25) 𝐶(Ω × R).
(i) Assume that 𝜔 ∈ 𝐶2 (Ω) ∩ 𝐶1 (Ω) satisfies
Therefore, lim sup𝑡→+∞ maxΩ V(⋅, 𝑡) ≤ 0, and there exists 𝑇2 >
𝑇1 such that Δ𝜔 (𝑥) + 𝑐 (𝑥) 𝜔 (𝑥) ≥ 0 in Ω,
𝜕𝜔 (33)
V (⋅, 𝑡) ≤ 𝜖, 𝑡 ≥ 𝑇2 . (26) ≤ 0 on 𝜕Ω.
𝜕𝑛
It follows from the first equation of (7) that If 𝜔(𝑥0 ) = maxΩ 𝜔, then 𝑔(𝑥0 , 𝜔(𝑥0 )) ≥ 0.
(ii) Assume that 𝜔 ∈ 𝐶2 (Ω) ∩ 𝐶1 (Ω) satisfies
𝑢
𝑢𝑡 − 𝑑1 Δ𝑢 ≥ 𝑟1 𝑢 (1 − ) − 𝜌𝑢𝛼 𝜖, 𝑡 > 𝑇2 , 𝑥 ∈ Ω. (27)
𝐾 Δ𝜔 (𝑥) + 𝑐 (𝑥) 𝜔 (𝑥) ≤ 0 in Ω,
𝜕𝜔 (34)
On account of and the arbitrariness of 𝜖 > 0, we
≥ 0 on 𝜕Ω.
have lim inf 𝑡→+∞ minΩ 𝑢(⋅, 𝑡) ≥ 𝐾. This combined with 𝜕𝑛
lim sup𝑡→+∞ maxΩ 𝑢(⋅, 𝑡) ≤ 𝐾 allows us to derive
If 𝜔(𝑥0 ) = minΩ 𝜔, then 𝑔(𝑥0 , 𝜔(𝑥0 )) ≤ 0.
lim max |𝑢 (⋅, 𝑡) − 𝐾| = 0. (28) Lemma 5. For any positive solution (𝑢, V) of system (8),
𝑡→+∞ Ω

0 < 𝑢 (𝑥) < 𝐾,

Hence, 𝐸1 is globally asymptotically stable when 𝑟2 > 𝐾𝛼 𝛽.
(iii) When 𝐸∗ = (𝑢∗ , V∗ ) exists, the corresponding 𝛽 𝑑1 𝑟1
characteristic equation is as follows: 0 < V (𝑥) < ( + )𝐾 (35)
𝜌 𝑑2 𝑟2
2−𝛼 ∗
𝜆2 + (𝑑1 𝜇𝑛 + 𝑑2 𝜇𝑛 − 𝑟1 (1 − 𝛼 − 𝑢 )) 𝜆 for any 𝑥 ∈ Ω.
𝐾
Proof. Form Lemma 4, 𝑢(𝑥) ≤ 𝐾 and from the strong
2−𝛼 ∗ (29)
+ 𝑑2 𝜇𝑛 (𝑑1 𝜇𝑛 − 𝑟1 (1 − 𝛼 − 𝑢 )) maximum principle 𝑢(𝑥) < 𝐾 for all 𝑥 ∈ Ω. Multiplying
𝐾
the first equation of (8) by 𝛽/𝜌 and adding it to the second
+ 𝜌𝛽𝑢∗(2𝛼−1) V∗ = 0. equation, we have
𝛽 𝛽 𝑢
Obviously, we have − ( 𝑑1 Δ𝑢 + 𝑑2 ΔV) = 𝑟1 𝑢 (1 − ) − 𝑟2 V
𝜌 𝜌 𝐾
2−𝛼 ∗ 𝛽 𝑑𝑟𝛽
𝜆 1𝑛 + 𝜆 2𝑛 = 𝑟1 (1 − 𝛼 − 𝑢 ) − 𝑑1 𝜇𝑛 − 𝑑2 𝜇𝑛 ,
𝐾 ≤ 𝑟1 𝐾 + 1 2 𝐾 (36)
𝜌 𝜌𝑑2
2−𝛼 ∗ (30)
𝜆 1𝑛 𝜆 2𝑛 = 𝑑2 𝜇𝑛 (𝑑1 𝜇𝑛 − 𝑟1 (1 − 𝛼 − 𝑢 )) 𝑟2 𝛽
𝐾 − ( 𝑑 𝑢 + 𝑑2 V) .
𝑑2 𝜌 1
+ 𝜌𝛽𝑢∗(2𝛼−1) V∗ .
Then the maximum principle implies that
∗
If 1 − 𝛼 − ((2 − 𝛼)/𝐾)𝑢 < 0, then 𝜆 1𝑛 + 𝜆 2𝑛 < 0 and 𝜆 1𝑛 𝜆 2𝑛 > 𝛽 𝛽 𝑑𝑟
0. Hence, all the roots of (29) have negative real part which 𝑑 𝑢 + 𝑑2 V ≤ ( 2 1 + 𝑑1 ) 𝐾. (37)
𝜌 1 𝜌 𝑟2
means that 𝐸∗ is locally asymptotically stable when 1 − 𝛼 −
((2 − 𝛼)/𝐾)𝑢∗ < 0. Hence, V(𝑥) < (𝛽/𝜌)(𝑑1 /𝑑2 + 𝑟1 /𝑟2 )𝐾.
Discrete Dynamics in Nature and Society 5

In the following, we estimate the positive lower bound of (ii) If 𝑢0 ≢ 0, V0 ≡ 0, then this implies that 𝑢0 satisfies
positive solution of (8). (8). So 𝑢0 ≡ 𝐾 for large 𝑛. Thus

Theorem 6. Let Ω be a bounded smooth domain in 𝑅𝑛 . There 𝛽𝑢𝑛𝛼 − 𝑟2 󳨀→ 𝛽𝐾𝛼 − 𝑟2 > 0 (45)
exist two positive constants 𝐶 < 𝐶 depending possibly on 𝑑1 ,
𝑑2 , 𝐾, 𝛽, 𝛼, 𝜌, and Ω, such that such that any positive solution for large 𝑛 since 𝛽𝐾𝛼 < 𝑟2 , which derives a contradiction
(𝑢(𝑥), V(𝑥)) of system (8) satisfies again to the second integral equation of (42). This completes
the proof.
𝐶 ≤ 𝑢 (𝑥) ≤ 𝐶,

𝐶 ≤ V (𝑥) ≤ 𝐶 (38) 5. Nonexistence of Nonconstant Positive

for any 𝑥 ∈ Ω.
Proof. From Lemma 5, we obtain
𝛽 𝑑1 𝑟1
𝑢 (𝑥) , V (𝑥) ≤ 𝐶 fl max {𝐾, ( + ) 𝐾} ,
𝜌 𝑑2 𝑟2
where 𝐶 depends on 𝑑1 , 𝑑2 , 𝐾, 𝛽, 𝛼, and 𝜌.
From Lemma 3, we obtain the fact that there exists a
positive constant 𝐶2 such that
sup 𝑢 (𝑥) ≤ 𝐶2 inf 𝑢 (𝑥) ,
Ω Ω
sup V (𝑥) ≤ 𝐶2 inf V (𝑥) .
Ω Ω
Steady States
In this section, we can show the nonexistence of nonconstant
positive solutions to system (8) when the diffusion coeffi-
cients 𝑑1 and 𝑑2 are large.
(39)
Theorem 7. There exists a positive constant 𝑑∗ such that
elliptic problem (8) has no nonconstant positive solution if
min{𝑑1 , 𝑑2 } > 𝑑∗ .
Proof. Suppose that (𝑢(𝑥), V(𝑥)) is a nonconstant positive
solution of system (8). Denote 𝑢 = |Ω|−1 ∫Ω 𝑢(𝑥)𝑑𝑥 ≥ 0,
(40) V = |Ω|−1 ∫Ω 𝑢(𝑥)𝑑𝑥 ≥ 0. Then

On the contrary, suppose the result is false. Then there ∫ (𝑢 − 𝑢) 𝑑𝑥 = 0,

Ω
exists a sequence {(𝑢𝑛 , V𝑛 )} of positive solutions to system (8) (46)
such that ∫ (V − V) 𝑑𝑥 = 0.
Ω
sup 𝑢𝑛 󳨀→ 0
Ω
Define
or sup V𝑛 󳨀→ 0 (41) 𝑢−𝑢
Ω ℎ (𝑢) = (47)
𝑢𝛼 − 𝑢𝛼
as 𝑛 󳨀→ +∞.
for 𝑢 ≠ 𝑢. Indeed, we can prove that ℎ(𝑢) > 0 and ℎ󸀠 (𝑢) > 0.
By the regularity theory for elliptic equations, there exists In fact, notice
a subsequence of {(𝑢𝑛 , V𝑛 )}, which will be denoted again by
{(𝑢𝑛 , V𝑛 )}, such that {(𝑢𝑛 , V𝑛 )} → (𝑢0 , V0 ) in 𝐶2 (Ω) as 𝑛 → (1 − 𝛼) 𝑢𝛼 + 𝛼𝑢𝑢(𝛼−1) − 𝑢𝛼
+∞. Observe that 𝑢0 ≤ 𝐾 and, from (41), either 𝑢0 ≡ 0 or ℎ󸀠 (𝑢) = 2
. (48)
(𝑢𝛼 − 𝑢𝛼 )
V0 ≡ 0. Therefore, we have the following two cases:
(i) 𝑢0 ≡ 0, V0 ≢ 0; or 𝑢0 ≡ 0, V0 ≡ 0. Let
(ii) 𝑢0 ≢ 0, V0 ≡ 0.
Since {(𝑢𝑛 , V𝑛 )} is a positive solution of (8), one can obtain ℎ1 (𝑢) = (1 − 𝛼) 𝑢𝛼 + 𝛼𝑢𝑢(𝛼−1) − 𝑢𝛼 , (49)
the following integral equation by integrating (8) for 𝑢𝑛 and
V𝑛 over Ω: and we have
𝑢𝑛
∫ (𝑟1 𝑢𝑛 (1 − ) − 𝜌𝑢𝑛𝛼 V𝑛 ) 𝑑𝑥 = 0, ℎ1󸀠 (𝑢) = 𝛼 (1 − 𝛼) 𝑢(𝛼−2) (𝑢 − 𝑢) , (50)
Ω 𝐾
(42)
which implies that ℎ1 (𝑢) > min ℎ1 (𝑢) = ℎ1 (𝑢) = 0 for 𝑢 ≠ 𝑢.
∫ (𝛽𝑢𝑛𝛼 V𝑛 − 𝑟2 V𝑛 ) 𝑑𝑥 = 0. Therefore, we obtain the fact that ℎ(𝑢) > 0.
Ω
Furthermore, multiplying the first equation of (8) by 𝛽/𝜌,
(i) In this case, 𝑢0 ≡ 0; then adding it to the second equation, and integrating over Ω, we
𝛽𝑢𝑛𝛼 − 𝑟2 󳨀→ −𝑟2 < 0 (43) get
uniformly as 𝑛 → ∞ and V𝑛 > 0; then for sufficiently large 𝑛, 𝛽
we have − ∫ ( 𝑑1 Δ𝑢 + 𝑑2 ΔV) 𝑑𝑥
Ω 𝜌
∫ (𝛽𝑢𝑛𝛼 V𝑛 − 𝑟2 V𝑛 ) 𝑑𝑥 < 0, (44) (51)
Ω 𝛽 𝑢
= ∫ ( 𝑟1 𝑢 (1 − ) − 𝑟2 V) 𝑑𝑥,
which is a contradiction. Ω 𝜌 𝐾
6 Discrete Dynamics in Nature and Society

and then the Neumann boundary conditions lead to Hence, if

𝛽 𝑢 𝐾𝛽𝑟1 1
𝑟2 ∫ V 𝑑𝑥 = ∫ 𝑟1 𝑢 (1 − ) 𝑑𝑥 ≤ |Ω| . (52) min {𝑑1 , 𝑑2 } > 𝑑∗ fl max {𝐴, 𝐵} , (58)
Ω Ω 𝜌 𝐾 4𝜌 𝜇1

Thus then
1 𝐾𝛽𝑟1 ∇ (𝑢 − 𝑢) = ∇ (V − V) = 0, (59)
V= ∫ V 𝑑𝑥 ≤ . (53)
Ω Ω 4𝑟2 𝜌
and (𝑢, V) must be a constant solution.
Multiplying the first equation in (8) by 𝑢 − 𝑢, we have
𝑢 6. Existence of Nonconstant Positive
𝑑1 ∫ |∇ (𝑢 − 𝑢)|2 𝑑𝑥 = ∫ (𝑢 − 𝑢) (𝑟1 𝑢 (1 − )
Ω Ω 𝐾 Steady States
− 𝜌𝑢𝛼 V) 𝑑𝑥 = ∫ (𝑢 − 𝑢) (𝑟1 (𝑢 − 𝑢) In this subsection, we discuss the existence of nonconstant
Ω positive solutions to system (8) when the diffusion coeffi-
𝑟1 2 cients 𝑑1 and 𝑑2 vary while the parameters 𝑟1 , 𝐾, 𝛼, 𝜌, 𝛽,
− (𝑢 − 𝑢2 ) − 𝜌𝑢𝛼 V + 𝜌𝑢𝛼 V) 𝑑𝑥 and 𝑟2 are fixed by using the Leray-Schauder degree theory.
𝐾
Throughout this section, we assume that the positive constant
≤ 3𝑟1 ∫ (𝑢 − 𝑢)2 𝑑𝑥 steady state 𝐸∗ = (𝑢∗ , V∗ ) exists.
Ω
(54) For simplicity, denote u = (𝑢, V) and
V
+ 𝜌𝐾𝛼 ∫ |(𝑢 − 𝑢) (V − V)| 𝑑𝑥 + 𝜌 ∫ (𝑢 2𝑟1 𝑢∗ (𝛼−1) ∗
Ω Ω ℎ (𝑢) 𝑎1 = 𝑟1 − − 𝜌𝛼 (𝑢∗ ) V ,
𝐾
𝛼
𝜌𝐾 𝐾𝛽𝑟1 𝑎2 = 𝜌 (𝑢∗ ) ,
𝛼
− 𝑢)2 𝑑𝑥 ≤ (3𝑟1 + + ) ∫ (𝑢
2 4𝑟2 ℎ (0) Ω
(𝛼−1)
𝑎3 = 𝛽V∗ (𝑢∗ ) ,
2 𝜌𝐾𝛼
− 𝑢) 𝑑𝑥 + ∫ (V − V)2 𝑑𝑥.
2 Ω 𝑢
𝑟1 𝑢 (1 − ) − 𝜌𝑢𝛼 V
Φ (u) = ( 𝐾 ),
Multiplying the second equation in (8) by V − V, we have 𝛽𝑢𝛼 V − 𝑟2 V
𝑑2 ∫ |∇ (V − V)|2 𝑑𝑥 = ∫ (V − V) (𝛽𝑢𝛼 V − 𝑟2 V) 𝑑𝑥 𝑎1 −𝑎2 (60)
Ω Ω Φu (𝐸∗ ) = ( ),
𝑎3 0
= ∫ (V − V) (−𝑟2 (V − V) + 𝛽𝑢𝛼 V − 𝛽𝑢𝛼 V) 𝑑𝑥
Ω 𝐷 = diag {𝑑1 , 𝑑2 } ,
(55)
𝐾𝛽2 𝑟1 2 𝜕𝑢 𝜕V
≤ (𝑟2 + 𝛽𝐾𝛼 + ) ∫ (V − V)2 𝑑𝑥 𝑋 = {(𝑢, V) ∈ [𝐶1 (Ω)] : = = 0 on 𝜕Ω} ,
8𝑟2 𝜌ℎ (0) Ω 𝜕𝑛 𝜕𝑛
𝐾𝛽2 𝑟1 𝑋+ = {(𝑢, V) : 𝑢, V ≥ 0, (𝑢, V) ∈ 𝑋} ,
+ ∫ (𝑢 − 𝑢)2 𝑑𝑥.
8𝑟2 𝜌ℎ (0) Ω
Λ = {(𝑢, V) ∈ 𝑋 : 𝐶−1 < 𝑢, V < 𝐶 on Ω} .
From (54) and (55) and the Poincaré inequality, we obtain
Thus, (8) can be written as
𝑑1 ∫ |∇ (𝑢 − 𝑢)|2 𝑑𝑥 + 𝑑2 ∫ |∇ (V − V)|2 𝑑𝑥
Ω Ω 𝜕u
−𝐷Δu = Φ (u) in = 0 on Ω, (61)
𝜕𝑛
2 2
≤ 𝐴 ∫ (𝑢 − 𝑢) 𝑑𝑥 + 𝐵 ∫ (V − V) 𝑑𝑥 (56)
Ω Ω and, obviously, u is a positive solution of (61) if and only if
1
≤ (𝐴 ∫ |∇ (𝑢 − 𝑢)|2 𝑑𝑥 + 𝐵 ∫ |∇ (V − V)|2 𝑑𝑥) , F (u) fl u − (𝐼 − Δ)−1 (𝐷−1 Φ (u) + u) in 𝑋+ , (62)
𝜇1 Ω Ω

where where (𝐼 − Δ)−1 is the inverse of 𝐼 − Δ with the homogeneous

Neumann boundary condition. As F(⋅) is a compact pertur-
𝜌𝐾𝛼 𝐾𝛽𝑟1 𝐾𝛽2 𝑟1 bation of the identity operator, the Leray-Schauder degree
𝐴 = 3𝑟1 + + + ,
2 4𝑟2 ℎ (0) 8𝑟2 𝜌ℎ (0) deg(F(⋅), Λ, 0) is well-defined from Theorem 6. By direct
(57) computation, we have
𝛼𝐾𝛽2 𝑟1 𝜌𝐾𝛼
𝐵 = 𝑟2 + 𝛽𝐾 + + .
8𝑟2 𝜌ℎ (0) 2 Fu (𝐸∗ ) = 𝐼 − (𝐼 − Δ)−1 (𝐷−1 Φu (𝐸∗ ) + 𝐼) . (63)
Discrete Dynamics in Nature and Society 7

If Fu (𝐸∗ ) is invertible, the index of F is defined as The positive solutions of the problem
∗ 𝛾
index (F (⋅) , 𝐸 ) = (−1) , (64) A𝑡 (u) = u in Ω,
∗
where 𝛾 is the number of negative eigenvalues of Fu (𝐸 ). 𝜕u (71)
Note that 𝜆 is an eigenvalue of Fu (𝐸∗ ) on 𝑋𝑗 if and only if = 0 on 𝜕Ω
𝜕𝑛
it is an eigenvalue of the matrix
1 are contained in Λ. Note that u is a positive solution of system
𝐵𝑗 = 𝐼 − [𝐷−1 Φu (𝐸∗ ) + 𝐼] (8) if and only if it is a positive solution of (71) with 𝑡 = 1. u∗ is
1 + 𝜇𝑗
the unique positive constant solution of (71) for any 𝑡 ∈ [0, 1].
(65)
1 According to the choice of 𝑑∗ in Theorem 7, we have 𝐸∗ which
= [𝜇 𝐼 − 𝐷−1 Φu (𝐸∗ )] . is the only fixed point of A0 .
1 + 𝜇𝑗 𝑗

Thus Fu (𝐸∗ ) is invertible if and only if, for all 𝑗 ≥ 0, the deg (𝐼 − A0 , Λ, 0) = index (𝐼 − A0 , Λ, 𝐸∗ ) = 1. (72)
matrix 𝐵𝑗 is nonsingular. Writing
Since F = 𝐼 − 𝐻(⋅, 1) and if (8) has no other solutions
𝐻 (𝑑1 , 𝑑2 ; 𝜇) = det [𝜇𝐼 − 𝐷−1 Φu (𝐸∗ )] except the constant one 𝐸∗ , then we have

1 (66) deg (𝐼 − A1 , Λ, (0, 0)) = index (F, 𝐸∗ )

= det [𝜇𝐷 − Φu (𝐸∗ )] , (73)
𝑑1 𝑑2 𝑖
= (−1)∑𝑘=𝑗+1 𝑚(𝜇𝑘 ) = −1.
we have that if 𝐻(𝑑1 , 𝑑2 ; 𝜇) ≠ 0, then 𝐻(𝑑1 , 𝑑2 ; 𝜇) < 0 if
and only if the number of negative eigenvalues of Fu (𝐸∗ ) in On the other hand, by the homotopy invariance of the
𝑋𝑗 is odd. The following lemma gives the explicit formula of topological degree,
calculating the index.
deg (𝐼 − A0 , Λ, 0) = deg (𝐼 − A1 , Λ, 0) , (74)
Lemma 8. If 𝐻(𝑑1 , 𝑑2 ; 𝜇𝑖 ) ≠ 0 for all 𝑖 ≥ 0, then
which is a contradiction. Therefore, there exists at least one
index (F (⋅) , 𝐸∗ ) = (−1)𝛾 ,
nonconstant solution of (8).
(67)
𝛾= ∑ 𝑚 (𝜇𝑖 ) ,
𝑖≥0,𝐻(𝑑1 ,𝑑2 ;𝜇)<0 7. Numerical Simulation
where 𝑚(𝜇𝑖 ) is the algebraic multiplicity of 𝜇𝑖 . 7.1. Global Stability of Equilibrium 𝐸1 . Consider system (7)
with following parameters: 𝑑1 = 0.8, 𝑑2 = 0.9, 𝑟1 = 0.9, 𝛽 =
To facilitate our computation of deg(F(⋅), 𝐸∗ ), we only 0.1, 𝑟2 = 0.2, 𝐾 = 2, 𝜌 = 0.1, and 𝛼 = 2/3. According
need consider the sign of det[𝜇𝐷 − Φu (𝐸∗ )]. The direct to the discussions in Section 3, the steady state 𝐸1 is globally
calculation gives asymptotically stable; see Figure 1.
det [𝜇𝐷 − Φu (𝐸∗ )] = 𝑑1 𝑑2 𝜇2 − 𝑎1 𝑑2 𝜇 + 𝑎2 𝑎3 . (68)
7.2. Stability of Positive Steady State 𝐸∗ . Consider system (7)
Obviously, nonnegative roots of (68) exist if and only if 𝑎12 𝑑2 − with following parameters: 𝑑1 = 0.8, 𝑑2 = 0.9, 𝑟1 = 0.9, 𝛽 =
4𝑑1 𝑎2 𝑎3 > 0 and 𝑎1 > 0. Assume that 𝜇+ and 𝜇− are the two 0.3, 𝑟2 = 0.2, 𝐾 = 2, 𝜌 = 0.1, and 𝛼 = 2/3. According to the
roots of (68), we have the following conclusion. discussions in Section 3, the positive steady state 𝐸∗ is locally
asymptotically stable; see Figure 2.
Theorem 9. Assuming that 𝛽𝐾𝛼 > 𝑟2 ,
𝑑2 4𝑎2 𝑎3 8. Conclusions
> , 𝑎1 > 0, (69)
𝑑1 𝑎12
In this paper, we have investigated the existence/nonexistence
and there exist 𝑖, 𝑗 ∈ 𝑁, such that 0 ≤ 𝜇𝑗 < 𝜇− < 𝜇𝑗+1 ≤ 𝜇𝑖 < of nonconstant positive steady states for a diffusive predator-
prey system with a group defense for prey under Neumann
𝜇+ < 𝜇𝑖+1 and ∑𝑖𝑘=𝑗+1 𝑚(𝜇𝑘 ) is odd, then (8) has at least one
boundary conditions. The existence results provide a theo-
nonconstant positive solution. retical support for pattern formation caused by diffusion. We
also study the stability of nonnegative equilibria and obtain
Proof. For 𝑡 ∈ [0, 1], we define
the fact that 𝐸1 is globally asymptotically stable when 𝛽𝐾𝛼 <
1−𝑡 𝑡 𝑟2 . In fact, the positive steady state does not exist at this time.
∗
+ ) 𝑓 (𝑢, V)
𝑢+( If 𝛽𝐾𝛼 > 𝑟2 and 1 − 𝛼 − ((2 − 𝛼)/𝐾)𝑢∗ < 0, then the positive
A𝑡 (u) ≜ (−Δ + 𝐼) ( −1 𝑑 𝑑 1 ) , (70)
1−𝑡 𝑡 steady state 𝐸∗ is locally asymptotically stable. It is easily
V + ( ∗ + ) 𝑔 (𝑢, V) obtained that when 1 − 𝛼 − ((2 − 𝛼)/𝐾)𝑢∗ = 0, characteristic
𝑑 𝑑2
equation (29) has a pair of purely imaginary roots. Therefore,
where 𝑑∗ is defined in Theorem 7. system (7) occurs with Hopf bifurcation, as shown in Figure 3.
8 Discrete Dynamics in Nature and Society

2
0.3
1.95

Predator 
0.2
Prey u

1.9
1.85 0.1
1.8 0
10 10
400 500 500
5 300 5 400
200 200 300
X 0 0 100 X 100
t 0 0 t

Figure 1: The steady state 𝐸1 is globally asymptotically stable.

0.8 6
0.7

Predator 
5.5
Prey u

0.6
5
0.5
0.4 4.5
10 10
500 500
5 400 5 400
300 300
200 200
X 0 100 X 0 0 100
0 t t

Figure 2: The positive steady state 𝐸∗ is locally asymptotically stable.

8 8
6 6
Predator 
Prey u

4 4
2 2
0 0
4 4
300 300
2 200 2 200
100 100
X 0 0 t X 0 0 t

Figure 3: Hopf bifurcation occurs with parameters 𝑟1 = 0.9, 𝛼 = 2/3, 𝛽 = 0.1, 𝑟2 = 0.18, 𝐾 = 10; 𝜌 = 0.2.

Conflicts of Interest [2] V. Volterra, “Fluctuations in the abundance of a species con-

sidered mathematically,” Nature, vol. 118, no. 2972, pp. 558–560,
The authors declare that they have no conflicts of interest. 1926.
[3] Z. Hu, Z. Teng, and L. Zhang, “Stability and bifurcation analysis
Acknowledgments of a discrete predator-prey model with nonmonotonic func-
tional response,” Nonlinear Analysis: Real World Applications,
The work is sponsored by Natural Science Foundation of vol. 12, no. 4, pp. 2356–2377, 2011.
Jiangsu Province (CN) (BK20150420).
[4] W. Zuo and J. Wei, “Stability and Hopf bifurcation in a diffusive
predatory-prey system with delay effect,” Nonlinear Analysis:
References Real World Applications, vol. 12, no. 4, pp. 1998–2011, 2011.
[1] L. AlfredJames, Elements of Physical Biology. Applied Mathe- [5] Y. Song and J. Wei, “Local Hopf bifurcation and global periodic
matical Sciences, Williams and Wilkins Company, Baltimore, solutions in a delayed predator-prey system,” Journal of Mathe-
Maryland, 1925. matical Analysis and Applications, vol. 301, no. 1, pp. 1–21, 2005.
Discrete Dynamics in Nature and Society 9

[6] M. Baurmann, T. Gross, and U. Feudel, “Instabilities in spatially [24] C.-S. Lin, W.-M. Ni, and I. Takagi, “Large amplitude stationary
extended predator-prey systems: spatio-temporal patterns in solutions to a chemotaxis system,” Journal of Differential Equa-
the neighborhood of Turing-Hopf bifurcations,” Journal of tions, vol. 72, no. 1, pp. 1–27, 1988.
Theoretical Biology, vol. 245, no. 2, pp. 220–229, 2007. [25] Y. Lou and W.-M. Ni, “Diffusion, self-diffusion and cross-
[7] L. I. Xue-Peng, “Wen Sheng Yang. Global asymptotic stability diffusion,” Journal of Differential Equations, vol. 131, no. 1, pp.
of a diffusive predator-prey model with ratio-dependent func- 79–131, 1996.
tional response,” Applied Mathematics Computation, vol. 250,
no. 7, pp. 71–77, 2015.
[8] J.-c. Huang, “Bifurcations in a predator-prey system with
Holling type-IV functional response,” Journal of Central China
Normal University. Natural Sciences. Huazhong Shifan Daxue
Xuebao. Ziran Kexue Ban, vol. 36, no. 1, pp. 3–10, 2002.
[9] X.-P. Yan and C.-H. Zhang, “Stability and turing instability in
a diffusive predator-prey system with Beddington-DeAngelis
functional response,” Nonlinear Analysis: Real World Applica-
tions, vol. 20, pp. 1–13, 2014.
[10] Y. Yang, S. Ruan, and D. Xiao, “Global stability of an age-
structured virus dynamics model with Beddington-DeAngelis
infection function,” Mathematical Biosciences and Engineering,
vol. 12, no. 4, pp. 859–877, 2015.
[11] V. Ajraldi, M. Pittavino, and E. Venturino, “Modeling herd
behavior in population systems,” Nonlinear Analysis: Real World
Applications, vol. 12, no. 4, pp. 2319–2338, 2011.
[12] P. A. Braza, “Predator-prey dynamics with square root func-
tional responses,” Nonlinear Analysis: Real World Applications,
vol. 13, no. 4, pp. 1837–1843, 2012.
[13] E. Venturino and S. Petrovskii, “Spatiotemporal behavior of a
prey-predator system with a group defense for prey,” Ecological
Complexity, vol. 14, pp. 37–47, 2013.
[14] Y. Lv, Y. Pei, and R. Yuan, “Hopf bifurcation and global stability
of a diffusive Gause-type predator–prey models,” Computers &
Mathematics with Applications, vol. 72, no. 10, pp. 2620–2635,
2016.
[15] C. Xu, S. Yuan, and T. Zhang, “Global dynamics of a pred-
ator–prey model with defense mechanism for prey,” Applied
Mathematics Letters, vol. 62, pp. 42–48, 2016.
[16] S. P. Bera, A. Maiti, and G. Samanta, “Stochastic analysis of a
prey-predator model with herd behaviour of prey,” Lithuanian
Association of Nonlinear Analysts. Nonlinear Analysis: Modelling
and Control, vol. 21, no. 3, pp. 345–361, 2016.
[17] C. Q. Xu and S. L. Yuan, “Stability and Hopf bifurcation in a
delayed predator-prey system with herd behavior,” Abstract and
Applied Analysis, vol. 2014, Article ID 568943, 8 pages, 2014.
[18] S. M. Salman, A. M. Yousef, and A. A. Elsadany, “Stability,
bifurcation analysis and chaos control of a discrete predator-
prey system with square root functional response,” Chaos,
Solitons & Fractals, vol. 93, pp. 20–31, 2016.
[19] X. Tang and Y. Song, “Cross-diffusion induced spatiotemporal
patterns in a predator-prey model with herd behavior,” Nonlin-
ear Analysis: Real World Applications, vol. 24, pp. 36–49, 2015.
[20] X. Tang and Y. Song, “Bifurcation analysis and Turing instability
in a diffusive predator-prey model with herd behavior and
hyperbolic mortality,” Chaos, Solitons & Fractals, vol. 81, pp.
303–314, 2015.
[21] X. Tang, Y. Song, and T. Zhang, “Turing–Hopf bifurcation
analysis of a predator–prey model with herd behavior and cross-
diffusion,” Nonlinear Dynamics, vol. 86, no. 1, pp. 73–89, 2016.
[22] C. V. Pao, On Nonlinear Parabolic and Elliptic Equations ,
Plenum Press, New York, NY, USA, 1992.
[23] Q. X. Ye and Z. Y. Li, Introduction to Reaction-Diffusion
Equations, Science Press, Beijing, China, 1990.
Advances in Advances in Journal of Journal of
Operations Research
Hindawi Publishing Corporation
Decision Sciences
Hindawi Publishing Corporation
Applied Mathematics
Hindawi Publishing Corporation
Algebra
Hindawi Publishing Corporation
Probability and Statistics
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

The Scientific International Journal of

World Journal
Hindawi Publishing Corporation
Differential Equations
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Submit your manuscripts at

https://www.hindawi.com

International Journal of Advances in

Combinatorics
Hindawi Publishing Corporation
Mathematical Physics
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of Journal of Mathematical Problems Abstract and Discrete Dynamics in

Complex Analysis
Hindawi Publishing Corporation
Mathematics
Hindawi Publishing Corporation
in Engineering
Hindawi Publishing Corporation
Applied Analysis
Hindawi Publishing Corporation
Nature and Society
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

International
Journal of Journal of
Mathematics and
Mathematical
#HRBQDSDĮ,@SGDL@SHBR
Sciences

Journal of International Journal of Journal of

Hindawi Publishing Corporation Hindawi Publishing Corporation Volume 2014

http://www.hindawi.com
Function Spaces
Hindawi Publishing Corporation
Volume 201 http://www.hindawi.com http://www.hindawi.com
Stochastic Analysis
Hindawi Publishing Corporation
Optimization
Hindawi Publishing Corporation
Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014