
This research concerned with a new formulation of a spatial predator-prey model with Leslie-Gower and Holling type II schemes in the presence of prey social behavior. The aim interest here is to distinguish the influence of Leslie-Gower term on the spatiotemporal behavior of the model. Interesting results are obtained as Hopf bifurcation, Turing bifurcation and Turing-Hopf bifurcation. A rigorous mathematical analysis shows that the presence of Leslie-Gower can induce Turing pattern, which shows that this kind of interaction is very important in modeling different natural phenomena. The direction of Turing-Hopf bifurcation is studied with the help of the normal form. The obtained results are tested numerically.
Citation: Fethi Souna, Salih Djilali, Sultan Alyobi, Anwar Zeb, Nadia Gul, Suliman Alsaeed, Kottakkaran Sooppy Nisar. Spatiotemporal dynamics of a diffusive predator-prey system incorporating social behavior[J]. AIMS Mathematics, 2023, 8(7): 15723-15748. doi: 10.3934/math.2023803
[1] | Fatao Wang, Ruizhi Yang, Yining Xie, Jing Zhao . Hopf bifurcation in a delayed reaction diffusion predator-prey model with weak Allee effect on prey and fear effect on predator. AIMS Mathematics, 2023, 8(8): 17719-17743. doi: 10.3934/math.2023905 |
[2] | Heping Jiang . Complex dynamics induced by harvesting rate and delay in a diffusive Leslie-Gower predator-prey model. AIMS Mathematics, 2023, 8(9): 20718-20730. doi: 10.3934/math.20231056 |
[3] | Rongjie Yu, Hengguo Yu, Min Zhao . Steady states and spatiotemporal dynamics of a diffusive predator-prey system with predator harvesting. AIMS Mathematics, 2024, 9(9): 24058-24088. doi: 10.3934/math.20241170 |
[4] | Wei Li, Qingkai Xu, Xingjian Wang, Chunrui Zhang . Dynamics analysis of spatiotemporal discrete predator-prey model based on coupled map lattices. AIMS Mathematics, 2025, 10(1): 1248-1299. doi: 10.3934/math.2025059 |
[5] | Na Min, Hongyang Zhang, Xiaobin Gao, Pengyu Zeng . Impacts of hunting cooperation and prey harvesting in a Leslie-Gower prey-predator system with strong Allee effect. AIMS Mathematics, 2024, 9(12): 34618-34646. doi: 10.3934/math.20241649 |
[6] | Weili Kong, Yuanfu Shao . Bifurcations of a Leslie-Gower predator-prey model with fear, strong Allee effect and hunting cooperation. AIMS Mathematics, 2024, 9(11): 31607-31635. doi: 10.3934/math.20241520 |
[7] | Kexin Zhang, Caihui Yu, Hongbin Wang, Xianghong Li . Multi-scale dynamics of predator-prey systems with Holling-IV functional response. AIMS Mathematics, 2024, 9(2): 3559-3575. doi: 10.3934/math.2024174 |
[8] | Binfeng Xie, Na Zhang . Influence of fear effect on a Holling type III prey-predator system with the prey refuge. AIMS Mathematics, 2022, 7(2): 1811-1830. doi: 10.3934/math.2022104 |
[9] | Naveed Iqbal, Ranchao Wu, Yeliz Karaca, Rasool Shah, Wajaree Weera . Pattern dynamics and Turing instability induced by self-super-cross-diffusive predator-prey model via amplitude equations. AIMS Mathematics, 2023, 8(2): 2940-2960. doi: 10.3934/math.2023153 |
[10] | Yingyan Zhao, Changjin Xu, Yiya Xu, Jinting Lin, Yicheng Pang, Zixin Liu, Jianwei Shen . Mathematical exploration on control of bifurcation for a 3D predator-prey model with delay. AIMS Mathematics, 2024, 9(11): 29883-29915. doi: 10.3934/math.20241445 |
This research concerned with a new formulation of a spatial predator-prey model with Leslie-Gower and Holling type II schemes in the presence of prey social behavior. The aim interest here is to distinguish the influence of Leslie-Gower term on the spatiotemporal behavior of the model. Interesting results are obtained as Hopf bifurcation, Turing bifurcation and Turing-Hopf bifurcation. A rigorous mathematical analysis shows that the presence of Leslie-Gower can induce Turing pattern, which shows that this kind of interaction is very important in modeling different natural phenomena. The direction of Turing-Hopf bifurcation is studied with the help of the normal form. The obtained results are tested numerically.
The prey predator model (denoted P-P model) is considered one of many powerful tools for predicting the evolution of species in nature [17,18,26]. In recent years, a well-known prey behavior dented by 'herd behavior' or 'social behavior' that been successfully modeled and analyzed. When the prey makes a group defense (which know by prey's social behavior), the predators cannot reach the prey located in the center of the herd, so the predator will hunt only the prey located in the frontier of the herd [31]. Therefore, the interaction happens only on the borders of the group. There are numerous approaches to determining the number of prey on the outer line of the group. The simplest one is to consider that the prey makes a group o a square shape, then the number of the prey will be proportional to the square root of the prey density [33]. For modeling this specific behavior we presume that the resources make a group with a square shape. Hence, the density of the resources in the frontier of the group is four times the square root of the number of the resources on the herd frontier. Hence, inspired by the manner of proposing the classical Holling I interaction functional we can deduce that the functional response that describes this conduct is F(M,N)=γN√M, where γ is the hunting rate, and M is the density o the resources, N is the density of the consumers. The manner of building this interaction functional is discussed in details through the paper [1]. Further, by changing the group structure or shape the density of the resources will be changed accordingly. As epitome if we consider that the resources make a group with a sphere shape it will influence the number of the resources in the outer bound where it will be 413×323×π13×M23. For generalizing the previous results, it is considered that in [3,37] that the functional response F can be generalized by the one F2(M,N)=γNMα, where the parameter 0<α<1 indicates the structure of the resources group. In the case of the square or circle group shape this rate will become α=12, and in the case of α=23 we got the case of the sphere of cube group shape, which means that this functional response generalizes all the previous cases of the prey herd structure. Dealing with the resources that exhibit grouping behavior is not always easy for the consumers, where the from one herd to another, and from predator to another, handling with the prey in the outer bound of the group changes and takes different time (different handling time for the predator to handle with a prey) which is been investigated in the first time by Holling to propose the Holling II interaction functional [17]. This point of view is applied for modeling the intermingling between the resources and the consumers in this case through the paper [6], where a new functional response is obtained F3(M,N)=γNMα1+γthMα, which summarizes all the intermingling functions and cases on interaction. This intermingling function was the subject of an investigation on many occasions for the purpose of modeling many behaviors in nature we mention a few [10,13]. In fact, using the functional response F3 as intermingling function next to the logistic increasing of the prey and linear mortality of resources (see as an example the paper [6]), we get a Gause-type model [4], which means that the mathematical investigation is trivial and can be distinguished easily from the paper [4]. Investigating the herd behavior in mathematical models is the subject of the recent activities, we cite the researches [2,5,7,8,9,10,12,13,32], and for more reading about different mathematical modeling of some natural phenomenon we cite the papers [11,14,16,21,22,23,24,25,30]. In this research, we will use a different approach where we will incorporate the Leslie-Gower intermingling functional with the functional F3(M,N) functional responses. Incorporating Leslie-Gower's functional response form with resources social behavior is a recent step and attracts any researchers, we cite for instance the papers [15,18,27,28,35,36], hence, it is the subject of interest in this research.
Modeling the intermingling resource-consumer in the case of resources social depends on the spatial positioning, so, it is wise to consider a spatiotemporal model [34]. Our purpose in this research is to study the influence of the Leslie-Gower forme on the evolution of the two species, where we will use a comparative analysis for achieving this goal. In these regards, we consider the first model that models the intermingling resource-consumer in the case of the resources social conduct and no Leslie-Gower form. The investigated model is:
{∂∂τM(x,τ)=βM(x,τ)(1−M(x,τ)L)+δΔM(x,τ)−γMα(x,τ)N(x,τ)1+γτhMα(x,τ),x∈(0,lπ), τ>0,∂∂τN(x,τ)=−μN(x,τ)+eγMα(x,τ)N(x,τ)1+γτhMα(x,τ)+ηΔN(x,τ),x∈(0,lπ), τ>0,∂∂→nM(x,τ)=∂∂→nN(x,τ)=0,x∈(0,lπ), τ>0,M(x,0)=M0(x)≥0,N(x,0)=N0(x)≥0,x∈(0,lπ), | (1.1) |
with M(x,τ) (resp. N(x,τ)) is the number of the resources (rep. consumer) at t and position x, βM(x,τ)(1−M(x,τ)L) is the logistic increasing of the resources with increasing rate β and the crying capacity of the environment L for the resources, μ in the mortality coefficient for the consumer, e is the conversion rate of the resources into consumer, δ (resp. η) is the diffusion rate for the resources (resp. consumer). To mention the Neumann boundary conditions highlights that neither the resources or the consumers cannot move cross the borders. Our main contribution consists to cooperate the Leslie-Gower with the prey social behavior interaction function and determine its effect on the temporal behavior of the solutions. The investigated model is given as:
{∂∂τM(x,τ)=βM(x,τ)(1−M(x,τ)L)+δΔM(x,τ)−γMα(x,τ)N(x,τ)1+γτhMα(x,τ),x∈(0,lπ), τ>0,∂∂τN(x,τ)=σN(x,τ)(1−N(x,τ)M(x,τ))+ηΔN(x,τ),∂∂→nM(x,τ)=∂∂→nN(x,τ)=0,x∈(0,lπ), τ>0,M(x,0)=M0(x)≥0,N(x,0)=N0(x)≥0,x∈(0,lπ). | (1.2) |
Our purpose is to investigate with the influence of the new approximation provided in (1.2) in modeling the interaction resources-consumer in the presence of the resources grouping behavior. In fact, we will use a comparative analysis between the two models (1.1) and (1.2), where we will show that the system (1.2) have a very rich dynamics. For achieving these aims we use the sections:
In Sec. 2 we analyze (1.1), where it is proved that it can undergo Hopf bifurcation (H-bifurcation), and cannot have Turing instability (T-instability), which means that it is not possible to have Turing-Hopf bifurcation (T-H bifurcation). The third section is used to analyze the system (1.2), where we will show that the system (1.2) undergoes many types of bifurcation as T-bifurcation, H-bifurcation, and T-H bifurcation. However, the system (1.1) can undergo only Hopf bifurcation (there is no T-bifurcation, and then there is no T-H bifurcation). The normal form of T-H bifurcation is utilized for study of steady states solution near the T-H bifurcation. The obtained mathematical results are confirmed using numerical simulation.
We know that the equilibrium states for the system (1.1) are solutions of the following system
{βM(1−ML)−γMαN1+γτhMα=0,−μN+eγMαN1+γτhMα=0. | (2.1) |
Clearly, (2.1) has three equilibria E0(0,0), E1(L,0), and the unique positive equilibrium point E∗(M∗,N∗) where
M∗=(μγ(e−μth))1α,N∗=eβμ(1−M∗L), |
which exists if the following conditions hold
(H1):e>μth and M∗<L (i.e.γ>γ∗=μLα(e−μth)). | (2.2) |
The linearized system of (1.1) evaluated at (M∗,N∗) is
(∂M∂τ∂N∂τ)=(DΔ+JE∗(M∗,N∗))(uv), | (2.3) |
where DΔ=diag(δ∂2∂x2,η∂2∂x2) and
JE∗(M∗,N∗)=(β(1−2M∗L)−αγ(M∗)α−1N∗(1+γτh(M∗)α)2−γ(M∗)α1+γτh(M∗)αeγα(M∗)α−1N∗(1+γτh(M∗)α)20). | (2.4) |
Then the characteristic equation of system (2.3) takes the following form
λ2−Trkλ+Detk=0,k∈N0, | (2.5) |
where
Trk=β(1−2M∗L)−αγ(M∗)α−1N∗(1+γτh(M∗)α)2−(δ+η)(kl)2, | (2.6) |
and
Detk=δη(kl)4−(β(1−2M∗L)−αγ(M∗)α−1N∗(1+γτh(M∗)α)2)η(kl)2+Det0, | (2.7) |
with
Det0=eαγ2(M∗)2α−1N∗(1+γτh(M∗)α)3>0. |
Using the fact that
β(1−M∗L)=αγ(M∗)α−1N∗1+γτh(M∗)α, and 1+γτh(M∗)α=eγμ(M∗)α, |
it follows that (2.6) becomes
Trk=β(1−α+αμτhe)−βL(2−α+αμτhe)(μγ(e−μth))1α−(δ+η)(kl)2, | (2.8) |
and
Detk=δη(kl)4−(β(1−α+αμτhe)−βL(2−α+αμτhe)(μγ(e−μth))1α)η(kl)2+Det0. | (2.9) |
Putting
LH=(2−α+αμτhe1−α+αμτhe)(μγ(e−μth))1α. |
Clearly, for k=0 we have Det0>0. Then, if L>LH we have Tr0<0 which means that E∗ is locally stable and if L<LH we get Tr0<0, then E∗ is unstable. Now, taking L as the bifurcation parameter. Thus, we have
Theorem 2.1. Presume that (H1) holds and we put
L=Lk:=(2−α+αμτhe)(μγ(e−μth))1αl2(1−α+αμτhe)−(δ+η)k2. | (2.10) |
Then, there exists an integer k∗ for which the model (1.1) undergoes a H-bifurcation at E∗ when L=Lk for 0≤k≤k∗. Further, if k=0 the periodic solution is homogeneous (spatially), and if k=1,...,k∗ the periodic solution is nonhomogeneous (spatially) when k=1,...,k∗, where [.] is the integer part function.
Proof. Denotes
ˉk=[l√1−α+αμτheδ+η]. |
Clearly, for an integer k<ˉk we have Lk>0. We consider that Detk defined by (3.10) is a function of k. Clearly, Det0>0, therefore there exists ˜k>0 (it can be +∞ if Detk>0 for all integer k=0,1,2,...) a positive integer which represents the first integer that satisfy Det˜k>0 and Det˜k+1≤0. Taking k∗=min{ˉk,˜k}. In this case, we guarantees that for L=Lk, k=0,1,...,k∗ we have Trk=0 and Detk>0 for all k=0,1,...,k∗ which implies that (2.5) has purely imaginary roots. Letting
λk(L)=θk(L)±iωk(L),k=0,1,...,k∗ |
be the solution of Eq (2.5) verifying
θk(Lk)=0,ωk(Lk)=√Detk(Lk). |
Then, we get
θ′k(Lk)=dReλkdL|L=Lk=[l2(1−α+αμτhe)−(δ+η)k2]2(2−α+αμτhe)(μγ(e−μth))1α>0. |
It follows that the transversal condition is justified at each Lk, k=0,1,...,k∗, that is to say that system (1.1) undergoes H-bifurcation at L=Lk.
Now, we focus on proving that the system (1.1) cannot exhibit the existence of Turing instability. This phenomena is know by the diffusion driven instability, where considering the presence of spatial diffusion (with distinct diffusion rates) can destabilize a stable equilibrium, which means that if an equilibrium is stable in the absence of the spatial diffusion can become instable in the presence of the diffusion (with distinct dispersal rates).
Lemma 2.2. The model (1.1) cannot have T-instability at E∗.
Proof. Before proceeding to prove Lemma 3.6, it is necessary to assume that the conditions that guarantees the existence and the stability of E∗ in the absence of diffusion holds which consists to suppose that (H1) holds and L<LH. Immediately, we find that
β(1−2M∗L)−αγ(M∗)α−1N∗(1+γτh(M∗)α)2=β(1−α+αμτhe)−βL(2−α+αμτhe)(μγ(e−μth))1α<0. |
Consequently, it follows by Detk>0, ∀k∈N the non occurrence of T-instability.
In this section, we consider the diffusive predator-prey model with Leslie-Gower term (1.2). We discuss the existence of Hopf bifurcation, after that we derive the condition for Turing pattern which leads to the occurrence of T-H bifurcation. First of all, we prove that system (1.2) has unique positive solution
Here, we investigate the existence and positivity of solution for (1.2).
Theorem 3.1. Assume that β,L,γ,τh,δ and η are all positive, if M0(x,τ)≥0 and N0(x,τ)≥0 for (x,τ)∈[0,lπ]×[0,+∞). Hence, (1.2) has a unique positive solution verifying
0≤M(x,τ)≤M∗∗(τ),0≤N(x,τ)≤N∗∗(τ)for(x,τ)∈[0,lπ]×[0,+∞], |
such that (M∗∗(τ),N∗∗(τ)) is the unique solution of
{Mτ=βM(1−ML),Nτ=N(1−NM),M(0)=M∗∗0=supx∈[0,lπ]M0(x),N(0)=N∗∗0=supx∈[0,lπ]N0(x). | (3.1) |
Proof. Putting
f(M,N)=βM(1−ML)−γMαN1+γτhMα,g(M,N)=σN(1−NM). |
Since, fN≤0 and gM≥0 for (M,N)∈R2+={(M,N)|M≥0} and from [29] yields that f,g are mixed quasi-monotone functionals in R. Now, letting
(˜M(x,τ),˜N(x,τ))=(0,0) and (ˆM(x,τ),ˆN(x,τ))=(M∗∗(τ),N∗∗(τ)). |
From
∂ˆM∂τ−δΔˆM−f(ˆM,˜N)=0≥0=∂˜M∂τ−δΔ˜M−f(˜M,ˆN), |
∂ˆN∂τ−ηΔˆN−f(ˆM,ˆN)=0≥0=∂˜N∂τ−ηΔ˜N−f(˜M,˜N) |
and 0≤M0(x,τ)≤M∗∗0,0≤N0(x,τ)≤>N∗∗0, for (x,τ)∈[0,lπ]×[0,+∞], we can conclude that (˜M,˜N) and (ˆM,ˆN) are the upper and lower solutions of (1.2), respectively. From [29], we get the existence of a unique globally defined solution (M(x,τ),N(x,τ)) of system (1.2) satisfying 0≤M(x,τ)≤M∗∗(τ) and 0≤N(x,τ)≤N∗∗(τ).
Applying the strong maximum principle to (1.2), we obtain that M(x,τ)>0, N(x,τ)>0 for (x,τ)∈(0,lπ)×(0,+∞).
Now, we mainly focus on checking the existence of feasible steady states. We can easily verify that E1=(L,0) are always equilibrium for (1.2). Next, we investigate the existence of a positive steady state of our proposed system (1.2) which is denoted by E2=(M∗,N∗), then we have the following theorem
Theorem 3.2. Suppose that all parameters are positive, then the system (1.2) has a unique positive steady state E2=(M∗,N∗) with M∗=N∗ and 0<M∗<L (please see Figure 1).
Proof. Obviously, E2 is the solution of:
{βM∗(1−M∗L)−γM∗αN∗1+γτhM∗α=0,σ(1−N∗M∗)=0. | (3.2) |
Using the second equation of (3.2), we can notice that M∗=N∗. Substituting M∗=N∗ into the first equation of (3.2), gives h(M∗)=0, with
h(M)=γMα1+γτhMα+β(ML−1), |
and
h′(M)=αγMα−1(1+γτhMα)2+βL>0, for M∈[0,L]. |
Obviously, h(0)=−β<0 and h(L)=γLα1+γτhLα>0. Hence, h(M) is an increasing function for M∈[0,L], and bisect the horizontal axis at M∗, where 0<M∗<L. The proof of Theorem 3.2 is completed.
Now, we define the Sobolev space
X={U=(M,N)T∈H2(0,lπ)×H2(0,lπ)|∂→nM=∂→nN=0,x=0,lπ}, | (3.3) |
and its complexification, XC:=X⊕iX={x1+ix2|x1,x1∈X}, with the the inner product ⟨.,.⟩ as
<U1,U2>=lπ∫0(¯M1M2+¯N1N2)dx,Ui=(Mi,Ni)T∈XC,i=1,2. |
The linearization of (1.2) at (M,N) is
(∂M∂τ∂N∂τ)=(DΔ+Jk(M,N))(MN), | (3.4) |
where DΔ=diag(δ∂2∂x2,η∂2∂x2) and
Jk(M,N)=(β(1−2ML)−αγMα−1N(1+γτhMα)2−γMα1+γτhMασ(NM)2σ(1−2NM)). | (3.5) |
Clearly, the problem
−Δϕ=μϕ,x∈(0,lπ);ϕ′(0)=ϕ′(lπ)=0, |
has the eigenvalues ξk=(kl)2(k=0,1,2,...) where the corresponding normalized eigenfunctions defining in the Sobolev space X are
ξk(x)=cosklx‖cosklx‖={√1lπ,k=0,√2lπcosklx,k≥0. | (3.6) |
Now, let
U(x,τ)=+∞∑k=0(akbk)cos(kπlx)eλkτ, | (3.7) |
be the nontrivial solution of the system (3.4) yields the existence of k∈N0 such that λ satisfy
det(λI−Ck−Jk(M,N))=0, |
where, I is 2×2 identity matrix, and Ck=−(kl)2diag(δ,η). After a straightforward calculation we obtain the characteristic equation of (3.4) as
λ2−Trkλ+Detk=0,k∈N0, | (3.8) |
where
Trk=β(1−2ML)−αγMα−1N(1+γτhMα)2+σ(1−2NM)−(δ+η)(kl)2, | (3.9) |
and
Detk=δη(kl)4−[ηβ(1−2ML)−ηαγMα−1N(1+γτhMα)2+δσ(1−2NM)](kl)2 | (3.10) |
+σ[(β(1−2ML)−αγMα−1N(1+γτhMα)2)(1−2NM)+γMα1+γτhMα(NM)2]. |
For the semi trivial steady state E1 we obtain
{Trk|E1=−β+σ−(δ+η)(kl)2,for k≥0,Detk|E1=δη(kl)4−(σδ−ηβ)(kl)2−σβ,for k≥0. | (3.11) |
Obviously, we have Det0|E1=−σβ<0 in the absence of the spatial diffusion, which means that E1 is always unstable.
Now, let's concentrate on analyzing the stability and the bifurcation properties of E2. From Theorem 3.2, (1.2) has a unique equilibrium denoted by E2=(M∗,N∗) where, M∗=N∗ and 0<M∗<L. Submitting (M∗,N∗) into (3.9) and (3.10) and using the fact that
βM∗(1−M∗L)=γ(M∗)α1+γτh(M∗)α and 1+γτh(M∗)α=γ(M∗)αβ(1−M∗L), | (3.12) |
then, we obtain
{Trk(σ)=σ0−σ−(δ+η)(kl)2,for k≥0,Detk(σ)=δη(kl)4−(ησ0−σδ)(kl)2+σC∗,for k≥0, | (3.13) |
where
C∗=αβ2γM∗(1−M∗L)2+12βM∗>0, | (3.14) |
and
σ0=β(1−2M∗L)−αβ2γ(M∗)α(1−M∗L)2. | (3.15) |
Now, putting
β∗=γ(1−2M∗L)(M∗)αα(1−M∗L)2. | (3.16) |
Remark 3.3. If L2<M∗<L, i.e. β∗<0, we obtain that Trk(σ)<0 and Detk(σ)>0, so E2 is always locally asymptotically stable (with or without diffusion). Thus, there is no Hopf bifurcation.
Now, assume that the following condition holds
(H2): 0<M∗<L2 and 0<β<β∗. |
So, we have the following theorem
Theorem 3.4. If η=δ=0, then:
(i) The positive steady state E2 of system (1.2) is locally asymptotically stable if σ>σ0 and unstable if σ<σ0 (see Figure 2).
(ii) The local system of the diffusion system (1.2) undergoes Hopf bifurcation for σ=σ0, where σ0 is defined in (3.15) (please see Figure 3).
Proof. Under the condition (H2) and from (3.13), if k=0 one can immediately get that E2 is locally asymptotically stable when σ>σ0 Figure 2 and unstable when σ<σ0. For σ=σ0, Eq (3.8) has a pair of purely imaginary roots ±i√σC∗. Let λ(σ)=υ(σ)±iω(σ) be the root of (3.8) satisfying υ(σ0)=0, ω(σ0)=±i√σC∗. A simple calculate gives following transversality condition:
ddσυ(σ)|σ=σ0=−12<0. | (3.17) |
Therefore, we conclude that the non diffusive system associated with (1.2) undergoes H-bifurcation at E2 when σ=σ0 (see Figure 3).
Now, concentrating on studying the occurrence of time-periodic solutions for (1.2) generated by H-bifurcation. throughout the rest of this paper, we assume that (H2) holds i.e. σ0>0 and taking σ as the bifurcation parameter. Recall that H-bifurcation appears if
Trk=0, Detk>0 and ∂∂σλ(σ)|σ=σ0≠0. |
Obviously
σ=σk=σ0−(δ+η)(kl)2, k∈N0 | (3.18) |
where
k∗=max{n∈N | Detk(σk)>0 and σk>0 for k=0,1,...,n−1} | (3.19) |
are the critical points for H-bifurcation. The values of the H-bifurcation are highlighted as follows.
Theorem 3.5.
(i) The diffusive model (1.2) undergoes H-bifurcation at E2 when σ=σk, for k=0,1,...,k∗−1, where σk and k∗ are defined in (3.18) and (3.19), respectively (see Figure 4). Further, for k=0 we get a homogeneous periodic solution and a non homogeneous periodic solution for k=1,2,...,k∗−1.
(ii) The eventual H-bifurcation points (σk)0≤k≤k∗−1 satisfying the following relationships
σk∗−1<...<σk+1<σk<σk−1<...<σ1<σ0. |
Proof.
(i) From the definition of the integer k∗, we can easily affirm that when σ=σk, Trk(σk)=0 and Detk(σk)>0 for k=0,1,...,k∗−1, which follows purely imaginary roots of Eq (3.8). Letting
λk(σ)=Ak(σ)±iBk(σ),k=0,1,...,k∗−1 |
be the roots of Eq (3.8) which verifies
Ak(σk)=0,Bk=√Detk(σk). |
It follows that if σ is in the neighborhood σk, the solutions of the characteristic equation (3.8) take the following form
Ak(σ)±iBk(σ)=Trk(σ)±√Tr2k(σ)−4Detk(σ)2 |
with
Ak(σ)=Trk(σ)2,Bk(σ)=√Detk(σ)−Tr2k(σ)4 |
and we have
A′(σk)=−12<0. |
This yields to the verification of the transversality condition for each σk where k=0,1,...,k∗−1.
(ii) Now, we are in the position to prove the second affirmation. After a simple calculate we find
σk+1−σk=−(1+2n)(δ+η)l2<0. |
This implies that (σk) is strictly decreasing sequence for all k=0,1,...,k∗−1. The proof of Theorem 3.5 is completed.
In the following, we mainly prove that under certain sufficient condition, the system (1.2) may exhibits T-instability unlike the situation in the system without Leslie-Gower term (please see Sec. 2). Mentioning that T-instability holds when the equilibrium point is locally stable in the non diffusive system and becomes unstable in the case of diffusive system (i.e. Detk<0 for some value of integer k). Notice that E2 is locally stable if the condition (H2) is satisfied and σ>σ0. In this case we have
Tr0(σ)<0, and Det0(σ)>0. |
In order to study the occurrence of the Turing instability, we define the functional Θ as
Θ((kl)):=Detk(σ)=δη((kl)2)2−(ησ0−σδ)(kl)2+σC∗, |
which considered as a quadratic polynomial in (kl)2 and σ0 is defined in (3.15).
Lemma 3.6. If
ηδ<σσ0, | (3.20) |
the system (1.2) cannot undergo T-instability.
Proof. Clearly, under the condition (3.20) we have Θ((kl)2)>0 which means that system (1.2) has no diffusion driven instability.
In the next, we presume that
(H3):ηδ>σσ0. |
Obviously, if Θ((kl)2)<0, then Eq (3.8) has one of the two roots is positive. If F(η,δ)=ησ0−σδ>0, Θ((kl)) has a minimum at
(kl)2min=ησc−σδ2ηδ>0. |
Evaluating Θ at this minimum, we get
min(kl)2Θ((kl)2)=σC∗−(ησ0−σδ)24ηδ, | (3.21) |
where C∗ is defined by (3.14). Next we show that under (H3), min(kl)2Θ((kl)2)<0 for some values η/δ>0, which it known by the condition of T-instability. Defining the ratio ξ=η/δ and
Π(η,δ)=(ησ0−σδ)2−4ηδσC∗=σ20η2−2σ(σ0+2C∗)ηδ+σ2δ2. |
Then, Π(η,δ)=0 and F(η,δ)=0 are equivalent to
σ20ξ2−2σ(σ0+2C∗)ξ+σ2=0, | (3.22) |
and
ξ=ξ∗=σσ0. | (3.23) |
Notice that
4σ2(σ0+2C∗)2−4σ20σ2=16σ2C∗(C∗+σ0)>0, |
which means that Eq (3.22) has two has two positive real roots
ξ1=σ(σ0+2C∗)+2σ√C∗(C∗+σ0)σ20,ξ2=σ(σ0+2C∗)−2σ√C∗(C∗+σ0)σ20. | (3.24) |
Easily, one can see that 0<ξ2<ξ∗<ξ1 and if η/δ>ξ1, we have min(kl)2Θ((kl)2)<0 and F(η,δ)>0. Here, the positive equilibrium E2 becomes unstable, which means that T-instability occurs.
Now, defining the set
RT:={(η,δ):δ>0, η>0 and η/δ>ξ1}. |
Hence, we get the results
Theorem 3.7. Presume that (H2) holds and σ>σ0 (for having the stability of the positive equilibrium). Then there exists an unbounded set RT such that for any (η,δ)∈RT, the equilibrium E2 becomes unstable, that is, Turing instability (for illustrations we refer Figure 5).
In this subsection, our aim is to investigate the occurrence of T-H bifurcation. This type of bifurcation happen if there are two integers k1 and k2 where for k=k1, (1.2) has H-bifurcation and for k=k2 the system (1.2) undergoes T-bifurcation, this kind of bifurcation is a bi-dimensional bifurcation which means that we need to choose two bifurcation parameters. Hence, we choose σ and δ as bifurcation parameters. Assuming that σ0>0, it is well known that Tr0(σ)=0 and Det0(σ)<0 are necessary conditions for Hopf bifurcation to occur. From (3.13) and (3.15), Tr0=0 is equivalent to
σ=σH(δ)=σ0=β(1−2M∗L)−αβ2γ(M∗)α(1−M∗L)2, | (3.25) |
which is the line of H-bifurcation in δ−σ plan, where the frequency of the oscillations is
ωH=Im(λ)=√σC∗. |
Besides, T-instability occurs when ηδ>ξ1, where ξ is given in (3.24). It follows that the critical value of the T-bifurcation for the parameter σ is
σ=σT(δ)=ησ20δ((σ0+2C∗)+2√C∗(C∗+σ0))=ησ20(√C∗+√C∗+σ0)2. | (3.26) |
Now, we prove the existence of intersection point between the H-bifurcation curve σH and T-instability curve σT in δ−σ plane. Defining the following function
h(x)=ησ20x((σ0+2C∗)+2√C∗(C∗+σ0)),x>0. |
Clearly, h is monotonously decreases with the increasing of x. In addition, we have limx→0+h(x)=+∞. Therefore, we conclude that the H-bifurcation line σH cuts the T-bifurcation curve σT at (δT−H,σT−H)=(ˆδ,σ0) where
ˆδ=ησ0(√C∗+√C∗+σ0)2. | (3.27) |
Now, we will examine the transversality condition. Fixing δ, and taking σ as parameter, letting λ(σ) the roots of (3.8), hence:
ddσReλ(σ)|σH=ddσReλ(σ)|σT=−12<0, |
then, we get
Theorem 3.8. Presume that the conditions (H2) and (H3) are verified, then:
(i) The H-bifurcation line σH cuts the T-bifurcation curve σT in δ−σ-parameter space at the unique point (ˆδ,σ0), where σ0 and ˆδ are defined in (3.15) and (3.27) (for illustrations we refer Figure 6).
(ii) At (δ,σ)=(ˆδ,σ0) the characteristic equation (3.8) has a simple zero root.
In order to illustrate numerically the obtained result in Theorem 3.8, T-bifurcation curve and H-bifurcation curves are plotted in δ−σ plane (please see Figure 5). we fix the parameters β=0.9, L=10, γ=0.2, τh=0.5, α=2/3, σ=0.5 and η=1. It follows that M∗=N∗=4.97<L=10, σ0=0.4478<σ=0.5, α∗=0.432<α=0.667, C∗=5.6316 and ˆδ=0.049. From Figure 6, we can see that the intersection point (1) divide the δ−σ plan into four regions. In D1, E2 is stable. D2 represents the T-bifurcation region. D3 is the domain in which the pure H-bifurcation occurs. In D4, both T-instability and H-bifurcation occur. In this situation, the diffusive system (1.2) produces a complex spatiotemporal patterns, where the two instabilities T- bifurcation and H-bifurcation coincide.
Here, we mainly focus on the calculate of the normal form of T-H bifurcation in order to determine the properties and the spatiotemporal dynamics of (1.2) at E2=(M∗,N∗) near the T-H bifurcation point (ˆδ,σ0). At first, we set μ=(μ1,μ2)∈R2) where μ1=δ−ˆδ, μ2=σ−σ0. Then, we we apply the translation ¯M=M−M∗, ¯N=N−N∗ to (1.2) and introduce a new parameter μ=(μ1,μ2)∈R2). we denote ¯M by M and ¯N by N. Therefore, the diffusive system (1.2) becomes
{∂∂τM(x,τ)=β(M(x,τ)+M∗)(1−(M(x,τ)+M∗)L)+(ˆδ+μ1)ΔM(x,τ)−γ(M(x,τ)+M∗)α(N(x,τ)+M∗)1+γτh(M(x,τ)+M∗)α,x∈(0,lπ), τ>0,∂∂τN(x,τ)=(σ0+μ2)(N(x,τ)+M∗)(1−N(x,τ)+M∗M(x,τ)+M∗)+ηΔN(x,τ),x∈(0,lπ), τ>0,∂∂→nM(x,τ)=∂∂→nN(x,τ)=0,x∈(0,lπ), τ>0,M(x,0)=M0(x)≥0,N(x,0)=N0(x)≥0,x∈(0,lπ). | (4.1) |
For system (4.1) and according to [19], also we get
D(μ)=(ˆδ+μ100η),L(μ)=(σ0π(σ0+μ2)−(σ0+μ2)), | (4.2) |
and
F(φ,μ)=(β(φ1+M∗)(1−(φ1+M∗)L)−γ(φ1+M∗)α(φ2+M∗)1+γτh(φ1+M∗)α−σ0φ1+a1φ2(σ0+μ2)(φ2+M∗)(1−φ2+M∗φ1+M∗)−(σ0+μ2)φ1+(σ0+μ2)φ2), | (4.3) |
where
π=−γM∗1+γτhM∗,andφ=(φ1,φ2)T∈X. |
It follows that
D(0)=(ˆδ00η),D1(μ)=(2μ1000),L(0)=(σ0πσ0−σ0),L1(μ)=(002μ2−2μ2), |
and
Q(φ,χ)=(a11φ1χ1+a12(φ1χ2+φ2χ1)+a13φ2χ2a21φ1χ1+a22(φ1χ2+φ2χ1)+a23φ2χ2), |
C(φ,χ,ν)=(b11φ1χ1ν1+b12(φ1χ1ν2+φ1χ2ν1+φ2χ1ν1)+b13(φ1χ2ν2+φ2χ1ν2+φ2χ2ν1)+b14φ2χ2ν2b21φ1χ1ν1+b22(φ1χ1ν2+φ1χ2ν1+φ2χ1ν1)+b23(φ1χ2ν2+φ2χ1ν2+φ2χ2ν1)+b24φ2χ2ν2), |
where
χ=(χ1,χ2)T∈X,ν=(ν1,ν2)T∈X. |
The coefficients aij and bij are given as
a11=α(1−α)γ(M∗)α−1+(2γ2α2τh+α(1−α))(M∗)α−2(1+γτh(M∗)α)3, a12=−αγ(M∗)α−1(1+γτh(M∗)α)2, |
a21=−2σ0M∗, a22=2σ0M∗, a23=−2σ0M∗, |
b11=−α(α−1)(α−2)γ(M∗)α−2(1+γτh(M∗)α)2+6α2γ2τh(α−1)(M∗)α−2(1+γτh(M∗)α)−6α3γ3τh(M∗)3α−2(1+γτh(M∗)α)4, |
b12=α(1−α)γ(M∗)α−2(1+γτh(M∗)α)+2α2γ2τh(M∗)2α−2(1+γτh(M∗)α)3, |
b_{21} = 6\frac{\sigma_{0}}{(M^{\ast})^{2}}, \ b_{22} = -4\frac{\sigma_{0}}{(M^{\ast})^{2}}, \ b_{23} = 2\frac{\sigma_{0}}{(M^{\ast})^{2}}, |
a_{13} = b_{13} = b_{14} = b_{24} = 0. |
Now, we get the corresponding characteristic matrices as
\mathbb{D}_{k}(\lambda) = \left( \begin{array}{cc} \lambda +\delta \mu_{k}-\sigma_{0} & -\pi \\ -\sigma_{0} & \lambda +\eta \mu_{k}+\sigma_{0} \end{array} \right),\quad k\in \mathbb{N}. |
Clearly, \lambda = \pm i\omega with \omega = \sqrt{Det_{0}} , are eigenvalues of \mathbb{D}_{0}(\lambda) , and \lambda = 0 is a simple eigenvalues for \mathbb{D}_{\hat{\delta}}(\lambda) , while other eigenvalues have negative real parts. From Theorem 3.8 and by using a simple calculate we can obtain
\varphi_{1} = \left( \begin{array}{cc} 1 \\ \frac{\sigma_{0}}{\hat{\delta}\mu_{\hat{\delta}}+\sigma_{0}} \end{array} \right),\quad \varphi_{2} = \left( \begin{array}{cc} 1 \\ \frac{\sigma_{0}-i\omega}{-\pi} \end{array} \right), |
and
\chi_{1} = \left( \begin{array}{cc} \frac{\hat{\delta}\mu_{\hat{\delta}}+\sigma_{0}}{(1+\hat{\delta})\mu_{\hat{\delta}}} \\ \frac{(\mu_{\hat{\delta}}-\sigma_{0})(\hat{\delta}\mu_{\hat{\delta}}+\sigma_{0})}{(1+\hat{\delta})\mu_{\hat{\delta}}\sigma_{0}} \end{array} \right),\quad \chi_{2} = \left( \begin{array}{cc} \frac{-\pi\sigma_{0}}{-\pi\sigma_{0}+(\omega+i\sigma_{0})^{2}} \\ \frac{-\pi(i\omega-\sigma_{0})}{-\pi\sigma_{0}+(\omega+i\sigma_{0})^{2}} \end{array} \right). |
Then, by the procedure developed in [19,20], the normal form restricted on central manifold at T-H bifurcation singularity is
\begin{equation} \left\{ \begin{array}{lccccc} \dot{Z}_{1} = m_{1}(\mu)Z_{1}+m_{200}Z_{1}^{2}+m_{011}Z_{2}\overline{Z}_{2} \\ \quad +m_{300}Z_{1}^{3}+m_{111}Z_{1}Z_{2}\overline{Z}_{2}+\text{h.o.t.}, \\ \dot{Z}_{2} = i\omega Z_{2}+n_{2}(\mu)Z_{2}+n_{110}Z_{1}Z_{2} \\ \quad +n_{210}Z_{1}^{2}Z_{2}+n_{021}Z_{2}^{2}\overline{Z}_{2}+\text{h.o.t.}, \\ \dot{\overline{Z}}_{2} = -i\omega\overline{Z}_{2}+\overline{n}_{2}(\mu)\overline{Z}_{2}+\overline{n}_{110}Z_{1}\overline{Z}_{2} \\ \quad +\overline{n}_{210}Z_{1}^{2}\overline{Z}_{2}+\overline{n}_{021}Z_{2}\overline{Z}_{2}^{2}+\text{h.o.t.}. \end{array} \right. \end{equation} | (4.4) |
where the calculation of m_{1}(\mu), \ m_{200}, \ m_{011}, \ m_{300}, \ m_{111}, \ n_{2}(\mu), \ n_{110}, \ n_{210}, \ n_{021} are given in "Appendix".
By using the new parameter transformation Z_{1} = r, \ Z_{2} = \rho\cos \vartheta-i\rho\sin \vartheta , then we get
\begin{equation} \left\{ \begin{array}{lc} \dot{r} = m_{1}(\mu)r+m_{300}r^{3}+m_{111}r\rho^{2}, \\ \dot{\rho} = Re(n_{2}(\mu))\rho+Re(n_{210})\rho r^{2}+Re(n_{021})\rho^{2}. \end{array} \right. \end{equation} | (4.5) |
Here, we will provide some figure for illustrating the obtained results. In fact we investigate the following cases:
Figure 7: In this figure we set \beta = 0.5 , L = 15 , \gamma = 1.5 , \alpha = 0.7 , \tau_h = 1.7 , \sigma = 3.1 , \delta = 0.02 , \eta = 0.04 , l = 1 and the initial data M(x, 0) = 0.5+0.1\cos(3x) , N(x, 0) = 0.5+0.1\cos(2x) . Here, we obtain the stability of the nonhomogeneous steady state.
Figure 8: In this figure we consider the following values \beta = 3.1 , L = 50 , \gamma = 1.5 , \alpha = 0.66 , \tau_h = 0.01 , \sigma = 1.01 , \delta = 0.01 , \eta = 0.04 , l = 1 and the initial data M(x, 0) = 0.5+0.1\cos(3x) , N(x, 0) = 0.5+0.1\cos(2x) . Here, we arrive to the stability of the nonhomogeneous steady state.
Figure 9: In this graphical representation we set \beta = 1.1 , L = 50 , \gamma = 1.5 , \alpha = 0.66 , \tau_h = 0.01 , \sigma = 0.51 , \delta = 0.01 , \eta = 0.04 , l = 1 and the data M(x, 0) = 0.5+0.1\cos(3x) , N(x, 0) = 0.5+0.1\cos(2x) . We arrive at stability of the nonhomogeneous steady state.
Figure 10: Here we choose the set of values \beta = 1.1 , L = 50 , \gamma = 1.5 , \alpha = 0.66 , \tau_h = 0.01 , \sigma = 0.1 , \delta = 0.01 , \eta = 0.04 , l = 1 and the initial data M(x, 0) = 0.5+0.1\cos(3x) , N(x, 0) = 0.5+0.1\cos(2x) . Here, we arrive to stability of nonhomogeneous periodic solutions.
Figure 11: Here we choose the set of values \beta = 1.51 , L = 10 , \gamma = 1.5 , \alpha = 0.66 , \tau_h = 0.01 , \sigma = 0.051 , \delta = 0.01 , \eta = 0.04 , l = 1 and the data M(x, 0) = 0.5+0.1\cos(3x) , N(x, 0) = 0.5+0.1\cos(2x) . Here, we arrive to stability of nonhomogeneous periodic solutions.
Figure 12: Here, we choose the set of values values \beta = 1.51 , L = 10 , \gamma = 1.5 , \alpha = 0.66 , \tau_h = 0.01 , \sigma = 0.08 , \delta = 0.09 , \eta = 0.01 , l = 1 and the data M(x, 0) = 0.5+0.1\cos(3x) , N(x, 0) = 0.5+0.1\cos(2x) . Here, we arrive to stability of nonhomogeneous periodic solutions.
In this research, we investigated a spatiotemporal P-P model with Leslie-Gower for modeling the saturation of the predator increasing in terms of the density of the prey. The reason behind considering such approximation is to highlight that the evolution of the consumers is affected directly by the density of the resources. The similarity points between the two models, in the absence of the Leslie-Gower scheme interaction functional (1.1), and the presence of this last (1.2), as the occurrence of H-bifurcation in the absence and the presence of diffusion in two studied models. The disagreement between the two considered models consists of the presence of T-instability for (1.1) and the existence of this last in the diffusive model (1.2). Our study was focused on distinguishing the influence of this case of interaction on the value of T-H bifurcation. As it is been highlighted in Figure 4, \sigma (increasing rate for predator) generated by considering the Leslie-Gower scheme interaction functional has a big effect on the existence of T-patterns, and hence it influences the existence of T-H bifurcation. No one can neglect the role of herd behavior in modeling much natural behavior, and the considered model can fit many cases in different species as fish population (sardines), gnus and buffalos which intermingle with the lions, hyenas, which highlights the importance of considering such as approximation.
In fact, there are many scenarios that can behold as the persistence of the two categories with nonhomogeneous patches as it is been shown in Figures 7 and 8, or in nonhomogeneous and periodic patters as the Figures 10–12. These scenarios generated by the presence of the Leslie-Gower scheme functional response (more precisely the T-H bifurcation), which shows the huge importance of considering such as approximation.
This study is supported via funding from Prince Sattam bin Abdulaziz University project number (PSAU/2023/R/1444). The author S. Djilali is partially supported by the PRFU project No.C00L03UN130120200004, DGRSDT, Algeria.
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Calculations of m_{1}(\mu), \ m_{200}, \ m_{011}, \ m_{300}, \ m_{111}, \ n_{2}(\mu), \ n_{110}, \ n_{210}, \ n_{021} . Here, we are in the position to give the expressions of m_{1}(\mu), \ m_{200}, \ m_{011}, \ m_{300}, \ m_{111}, \ n_{2}(\mu), \ n_{110}, \ n_{210}, \ n_{021} . We will put only the formulas of these parameters. For more details about the method of calculation we refer the authors to [19,20].
m_{1}(\mu) = \frac{1}{2}\chi_{1}\left(L_{1}(\mu)\varphi_{1}-\mu_{\hat{\delta}}D_{1}(\mu)\varphi_{1}\right), |
m_{200} = m_{011} = m_{110} = 0, |
n_{2}(\mu) = \frac{1}{2}\chi_{2}\left(L_{1}(\mu)\varphi_{2}-0D_{1}(\mu)\varphi_{2}\right), |
m_{300} = \frac{1}{4}\chi_{1}C_{\varphi_{1}\varphi_{1}\varphi_{1}}+\frac{1}{\omega}\chi_{1}Re\left[iQ_{\varphi_{1}\varphi_{2}}\chi_{2}\right]Q_{\varphi_{1}\varphi_{1}}+\chi_{1}Q_{\varphi_{1}\left(h_{200}^{0}+\frac{1}{\sqrt{2}}h_{200}^{2\hat{\delta}}\right)}. |
m_{111} = \chi_{1}C_{\varphi_{1}\varphi_{1}\overline{\varphi}_{1}}+\frac{2}{\omega}\chi_{1}Re\left[iQ_{\varphi_{1}\varphi_{2}}\chi_{2}\right]Q_{\varphi_{1}\overline{\varphi}_{1}}+\chi_{1}\left(Q_{\varphi_{1}\left(h_{011}^{0}+\frac{1}{\sqrt{2}}h_{200}^{\hat{\delta}}\right)}+Q_{\varphi_{2}h_{101}^{\hat{\delta}}}+Q_{\overline{\varphi}_{2}h_{110}^{\hat{\delta}}}\right), |
n_{210} = \frac{1}{2}\chi_{2}C_{\varphi_{1}\varphi_{1}\varphi_{2}}+\frac{1}{2i\omega}\chi_{2}\left(2Q_{\varphi_{1}\varphi_{1}}\chi_{1}Q_{\varphi_{1}\varphi_{2}}+\left(-Q_{\varphi_{2}\varphi_{2}}\chi_{2}+Q_{\varphi_{2}\overline{\varphi}_{2}}\overline{\chi}_{2}\right)Q_{\varphi_{1}\varphi_{1}}\right)+\chi_{2}\left(Q_{\varphi_{1}h_{110}^{\hat{\delta}}}+Q_{\varphi_{2}h_{200}^{0}}\right), |
n_{021} = \frac{1}{2}\chi_{2}C_{\varphi_{2}\varphi_{2}\overline{\varphi}_{2}}+\frac{1}{4i\omega}\chi_{2}\left(\frac{2}{3}Q_{\overline{\varphi}_{2}\overline{\varphi}_{2}}\overline{\chi}_{2}Q_{\varphi_{2}\varphi_{2}}+\left(-2Q_{\varphi_{2}\varphi_{2}}\chi_{2}+4Q_{\varphi_{2}\overline{\varphi}_{2}}\overline{\chi}_{2}\right)Q_{\varphi_{2}\overline{\varphi}_{2}}\right)+\chi_{2}\left(Q_{\varphi_{2}h_{011}^{0}}+Q_{\overline{\varphi}_{2}h_{020}^{0}}\right), |
where
h_{200}^{0} = -\frac{1}{2}L^{-1}(0)Q_{\varphi_{1}\varphi_{1}}+\frac{1}{2i\omega}\left(\varphi_{2}\chi_{2}-\overline{\varphi}_{2}\overline{\chi}_{2}\right)Q_{\varphi_{1}\varphi_{1}}, |
h_{200}^{2\hat{\delta}} = -\frac{1}{2\sqrt{2}}\left[L(0)+diag(-4\mu_{\hat{\delta}}-4\hat{\delta}\mu_{\hat{\delta}}\right]^{-1}\times Q_{\varphi_{1}\varphi_{1}}, |
h_{011}^{0} = -L^{-1}(0)Q_{\varphi_{2}\overline{\varphi}_{2}}+\frac{1}{i\omega}\left(\varphi_{2}\chi_{2}-\overline{\varphi}_{2}\overline{\chi}_{2}\right)Q_{\varphi_{2}\overline{\varphi}_{2}}, |
h_{020}^{0} = \frac{1}{2}\left[2i\omega-L(0)\right]^{-1}Q_{\varphi_{2}\varphi_{2}}-\frac{1}{2i\omega}\left(\varphi_{2}\chi_{2}-\frac{1}{3}\overline{\varphi}_{2}\overline{\chi}_{2}\right)Q_{\varphi_{2}\varphi_{2}}, |
h_{200}^{\hat{\delta}} = \left[i\omega I-\left(L(0)-diag(-\mu_{\hat{\delta}}-\hat{\delta}\mu_{\hat{\delta}})\right)\right]^{-1}\times Q_{\varphi_{1}\varphi_{2}}-\frac{1}{i\omega}\varphi_{1}\chi_{1}Q_{\varphi_{1}\varphi_{2}}, |
and
h_{002}^{0} = \overline{h_{020}^{0}}, \quad h_{101}^{\hat{\delta}} = \overline{h_{101}^{\hat{\delta}}},\quad h_{200}^{2\hat{\delta}} = 0. |
[1] | V. Ajraldi, M. Pittavino, E. Venturino, Modelling herd behaviour in population systems. Nonlinear Anal. Real World Appl., 12 (2011), 2319–2338. https://doi.org/10.1016/j.nonrwa.2011.02.002 |
[2] |
P. A. Braza, Predator-prey dynamics with square root functional responses, Nonlinear Anal. Real World Appl., 13 (2012), 1837–1843. https://doi.org/10.1016/j.nonrwa.2011.12.014 doi: 10.1016/j.nonrwa.2011.12.014
![]() |
[3] |
I. M. Bulai, E. Venturino, Shape effects on herd behavior in ecological interacting population models, Math. Comput. Simulat., 141 (2017), 40–55. https://doi.org/10.1016/j.matcom.2017.04.009 doi: 10.1016/j.matcom.2017.04.009
![]() |
[4] | I. Boudjema, S. Djilali, Turing-Hopf bifurcation in Gauss-type model with cross diffusion and its application, Nonlinear Stud., 25 (2018), 665–687. |
[5] |
S. Djilali, Herd behavior in a predator-prey model with spatial diffusion: bifurcation analysis and Turing instability, J. Appl. Math. Comput., 58 (2018), 125–149. https://doi.org/10.1007/s12190-017-1137-9 doi: 10.1007/s12190-017-1137-9
![]() |
[6] |
S. Djilali, Impact of prey herd shape on the predator-prey interaction, Chaos, Solitons Fract., 120 (2019), 139–148. https://doi.org/10.1016/j.chaos.2019.01.022 doi: 10.1016/j.chaos.2019.01.022
![]() |
[7] | S. Djilali, Effect of herd shape in a diffusive predator-prey model with time delay, J. Appl. Anal. Comput., 9 (2019), 638–654. |
[8] |
S. Djilali, S. Bentout, Spatiotemporal patterns in a diffusive predator-prey model with prey social behavior, Acta. Appl. Math., 169 (2020), 125–143. https://doi.org/10.1007/s10440-019-00291-z doi: 10.1007/s10440-019-00291-z
![]() |
[9] |
S. Djilali, Pattern formation of a diffusive predator-prey model with herd behavior and nonlocal prey competition, Math. Meth. Appl. Sci., 43 (2020), 2233–2250. https://doi.org/10.1002/mma.6036 doi: 10.1002/mma.6036
![]() |
[10] |
S. Djilali, Spatiotemporal patterns induced by cross-diffusion in predator-prey model with prey herd shape effect, Int. J. Biomath., 13 (2020), 2050030. https://doi.org/10.1142/S1793524520500308 doi: 10.1142/S1793524520500308
![]() |
[11] | S. Djilali, B. Ghanbari, S. Bentout, A. Mezouaghi, Turing-Hopf bifurcation in a diffusive Mussel-Algae model with time-fractional-order derivative, Chaos, Solitons Fract., 138 (2020) 109954. https://doi.org/10.1016/j.chaos.2020.109954 |
[12] |
B. Ghanabri, S. Djilali, Mathematical and numerical analysis of a three-species predator-prey model with herd behavior and time fractional-order derivative, Math. Meth. Appl. Sci., 43 (2020), 1736–1752. https://doi.org/10.1002/mma.5999 doi: 10.1002/mma.5999
![]() |
[13] |
B. Ghanabri, S. Djilali, Mathematical analysis of a fractional-order predator-prey model with prey social behavior and infection developed in predator population, Chaos, Solitons Fract., 138 (2020), 109960. https://doi.org/10.1016/j.chaos.2020.109960 doi: 10.1016/j.chaos.2020.109960
![]() |
[14] |
B. Ghanbari, S. Kumar, R. Kumar, A study of behaviour for immune and tumor cells in immunogenetic tumour model with non-singular fractional derivative, Chaos, Solitons Fract., 133 (2020), 109619. https://doi.org/10.1016/j.chaos.2020.109619 doi: 10.1016/j.chaos.2020.109619
![]() |
[15] |
J. Gine, C. Valls, Nonlinear oscillations in the modified Leslie-Gower model, Nonlinear Anal. Real World Appl., 51 (2020), 103010. https://doi.org/10.1016/j.nonrwa.2019.103010 doi: 10.1016/j.nonrwa.2019.103010
![]() |
[16] |
E. F. D. Goufo, S. Kumar, S. B. Mugisha, Similarities in a fifth-order evolution equation with and with no singular kernel, Chaos, Solitons Fract., 130 (2020), 109467. https://doi.org/10.1016/j.chaos.2019.109467 doi: 10.1016/j.chaos.2019.109467
![]() |
[17] |
C. S. Holling, The functional response of invertebrate predator to prey density, Mem. Entomol. Soc. Canada, 98 (1966), 5–86. https://doi.org/10.4039/entm9848fv doi: 10.4039/entm9848fv
![]() |
[18] |
C. A. Ibarra, J. Flores, Dynamics of a Leslie-Gower predator-prey model with Holling type II functional response, Allee effect and a generalist predator, Math. Comput. Simul., 188 (2021), 1–22. https://doi.org/10.1016/j.matcom.2021.03.035 doi: 10.1016/j.matcom.2021.03.035
![]() |
[19] |
W. Jiang, Q. An, J. Shi, Formulation of the normal forms of Turing-Hopf bifurcation in reaction-diffusion systems with time delay, J. Differ. Equ., 268 (2020), 6067–6102. https://doi.org/10.1016/j.jde.2019.11.039 doi: 10.1016/j.jde.2019.11.039
![]() |
[20] |
W. Jiang, H. Wang, X. Cao, Turing instability and turing-hopf bifurcation in diffusive Schnakenberg systems with gene expression time delay, J. Dyn. Differ. Equ., 31 (2019), 2223–2247. https://doi.org/10.1007/s10884-018-9702-y doi: 10.1007/s10884-018-9702-y
![]() |
[21] |
S. Kumar, A new analytical modelling for fractional telegraph equation via Laplace transform, Appl. Math. Model., 38 (2014), 3154–3163. https://doi.org/10.1016/j.apm.2013.11.035 doi: 10.1016/j.apm.2013.11.035
![]() |
[22] |
S. Kumar, A. Kumar, D. Baleanu, Two analytical methods for time-fractional nonlinear coupled Boussinesq-Burger's equations arise in propagation of shallow water waves, Nonlinear Dyn., 85 (2016), 699–715. https://doi.org/10.1007/s11071-016-2716-2 doi: 10.1007/s11071-016-2716-2
![]() |
[23] |
S. Kumar, M. M. Rashidi, New analytical method for gas dynamics equation arising in shock fronts, Comput. Phys. Commun., 185 (2014), 1947–1954. https://doi.org/10.1016/j.cpc.2014.03.025 doi: 10.1016/j.cpc.2014.03.025
![]() |
[24] |
S. Kumar, D. Kumar, S. Abbasbandy, M. M. Rashidide, Analytical solution of fractional Navier-Stokes equation by using modified Laplace decomposition method, Ain Shams Eng. J., 5 (2014), 569–574. https://doi.org/10.1016/j.asej.2013.11.004 doi: 10.1016/j.asej.2013.11.004
![]() |
[25] |
S. Kumar, S. Ghosh, B. Samet, An analysis for heat equations arises in diffusion process using new Yang-Abdel-Aty-Cattani fractional operator, Math. Meth. Appl. Sci., 43 (2020), 6062–6080. https://doi.org/10.1002/mma.6347 doi: 10.1002/mma.6347
![]() |
[26] |
S. Kumar, R. Kumar, R. P. Agarwal, B. Samet, A study of fractional Lotka-Volterra population model using Haar wavelet and Adams-Bashforth-Moulton methods, Math. Meth. Appl. Sci., 43 (2020), 5564–5578. https://doi.org/10.1002/mma.6297 doi: 10.1002/mma.6297
![]() |
[27] |
Y. Liu, J. Wei, Spatiotemporal dynamics of a modified Leslie-Gower model with weak allee effect, Int. J. Bifurcat. Chaos, 30 (2020), 2050169. https://doi.org/10.1142/S0218127420501692 doi: 10.1142/S0218127420501692
![]() |
[28] |
Y. Li, F. Zhang, X. Zhuo, Flip bifurcation of a discrete predator-prey model with modified Leslie-Gower and Holling-type III schemes, Math. Biosci. Eng., 17 (2019), 2003–2015. https://doi.org/10.3934/mbe.2020106 doi: 10.3934/mbe.2020106
![]() |
[29] |
C. V. Pao, Dynamics of nonlinear parabolic systems with time delays, J. Math. Anal, Appl., 198 (1996), 751–779. https://doi.org/10.1006/jmaa.1996.0111 doi: 10.1006/jmaa.1996.0111
![]() |
[30] |
M. M. Rashidi, A. Hosseini, I. Pop, S. Kumar, N. Freidoonimehr, Comparative numerical study of single and two-phase models of nano-fluid heat transfer in wavy channel, Appl. Math. Mech.-Engl. Ed., 35 (2014), 831–848. https://doi.org/10.1007/s10483-014-1839-9 doi: 10.1007/s10483-014-1839-9
![]() |
[31] |
F. Souna, A. Lakmeche, S. Djilali, The effect of the defensive strategy taken by the prey on predator-prey interaction, J. Appl. Math. Comput., 64 (2020), 665–690. https://doi.org/10.1007/s12190-020-01373-0 doi: 10.1007/s12190-020-01373-0
![]() |
[32] |
F. Souna, A. Lakmeche, S. Djilali, Spatiotemporal patterns in a diffusive predator-prey model with protection zone and predator harvesting, Chaos, Solitons Fract., 140 (2020), 110180. https://doi.org/10.1016/j.chaos.2020.110180 doi: 10.1016/j.chaos.2020.110180
![]() |
[33] |
F. Souna, S. Djilali, A. Lakmeche, Spatiotemporal behavior in a predator-prey model with herd behavior and cross-diffusion and fear effect, Eur. Phys. J. Plus, 136 (2021), 474. https://doi.org/10.1140/epjp/s13360-021-01489-7 doi: 10.1140/epjp/s13360-021-01489-7
![]() |
[34] | M. E. Taylor, Partial differential equations III–Nonlinear equations, Applied Mathematical Science, Springer-Verlag, 1996. |
[35] |
S. Wang, Z. Xie, R. Zhong, Y. Wu, Stochastic analysis of a predator-prey model with modified Leslie-Gower and Holling type II schemes, Nonlinear Dyn., 101 (2020), 1245–1262. https://doi.org/10.1007/s11071-020-05803-3 doi: 10.1007/s11071-020-05803-3
![]() |
[36] |
D. Xu, M. Liu, X. Xu, Analysis of a stochastic predator-prey system with modified Leslie-Gower and Holling-type IV schemes, Phys. A, 537 (2020), 122761. https://doi.org/10.1016/j.physa.2019.122761 doi: 10.1016/j.physa.2019.122761
![]() |
[37] |
C. Xu, S. Yuan, T. Zhang, Global dynamics of a predator-prey model with defence mechanism for prey, Appl. Math. Lett., 62 (2016), 42–48. https://doi.org/10.1016/j.aml.2016.06.013 doi: 10.1016/j.aml.2016.06.013
![]() |
1. | Caijuan Jia, Yan Meng, Jiaxin Xiao, Nonhomogeneous periodic patterns in a predator–prey model with time delay and predator-taxis, 2024, 452, 03770427, 116119, 10.1016/j.cam.2024.116119 | |
2. | Y. Sudha, V. N. Deiva Mani, K. Murugesan, On the Solvability of Time-Fractional Spatio-Temporal COVID-19 Model with Non-linear Diffusion, 2024, 48, 2731-8095, 1281, 10.1007/s40995-024-01663-3 | |
3. | Debasish Bhattacharjee, Nabajit Ray, Dipam Das, Hemanta Kumar Sarmah, A discrete-time dynamical model of prey and stage-structured predator with juvenile hunting incorporating negative effects of prey refuge, 2024, 10, 26668181, 100710, 10.1016/j.padiff.2024.100710 | |
4. | M. Y. Hamada, Dynamics of a ricker type predator–prey model, 2024, 73, 0009-725X, 2689, 10.1007/s12215-024-01062-y | |
5. | K. Manikandan, N. Serikbayev, D. Aravinthan, K. Hosseini, Solitary wave solutions of the conformable space–time fractional coupled diffusion equation, 2024, 9, 26668181, 100630, 10.1016/j.padiff.2024.100630 | |
6. | Anil Kumar, Pawan Kumar Shaw, Sunil Kumar, Numerical investigation of pine wilt disease using fractal–fractional operator, 2024, 0973-1458, 10.1007/s12648-024-03298-x | |
7. | Beatris Adriana Escobedo-Trujillo, José Daniel López-Barrientos, Dario Colorado-Garrido, Francisco Alejandro Alaffita-Hernández, Variance optimality in constrained and unconstrained stochastic differential games, 2024, 14, 26667207, 100354, 10.1016/j.rico.2023.100354 | |
8. | Qamar Din, Khansa Jameel, Muhammad Sajjad Shabbir, Discrete-Time Predator–Prey System Incorporating Fear Effect: Stability, Bifurcation, and Chaos Control, 2024, 23, 1575-5460, 10.1007/s12346-024-01145-2 | |
9. | Zhenxiang Hu, Jiao Li, Lin Hu, Linfei Nie, Global dynamics of a nonlocal delayed reaction-diffusion Cholera model with phage-bacteria interaction, 2025, 0924-090X, 10.1007/s11071-025-10938-2 |