Existence of Forced Waves and Their Asymptotic for Leslie-Gower Prey-Predator Model with Nonlocal Effects under Shifting Environment ()
1. Introduction
Nowadays, Global warming is a hot topic and which has greatly changed the living environment of species. Climate change has caused massive changes in species distributions, abundances and diversity of species, and has led to the extinction of some vulnerable species around the globe, see [1] [2] , so we have to take this factor into account when studying population dynamics.
For the phenomenon that global warming causes habitat quality to change, many researchers have produced very scientific results, see [3] - [8] etc. In the past, Berestycki developed the reaction-diffusion equation under shifting environment and studied the existence of the forced waves for the equation. A few years ago, Berestycki and Fang in [9] considered the following Fisher-KPP equation
they established the existence and nonexistence of forced waves for this reaction-diffusion equation. It’s called forced waves because the growth function r is related to x and t. The form
was forced into existence by the environmental changes, hence the name. If the rate of environmental change
, there will be forced waves, and if
, there will be no forced waves.
In ecosystems, the dynamic relationship between predator and prey is ubiquitous and important, and it has been one of the major themes of ecology. It has been studied and still is by many researchers who have developed mathematical models to predict the interactions between prey and predator species. Of course, species interactions with each other are in a variety of ways, including cooperation, symbiosis and competition. Among them, cooperation is one of the important interactions among species and is common in both social animals and in human societies, see [10] . These species interactions are largely critical and required, see [11] [12] , so many researchers built a number of different models based on these different functional responses.
Yang et al. in [13] concerned a Lotka-Volterra cooperation system under climate change, they obtained the existence and asymptotics of forced waves, the model as follows
(1.1)
where
respectively represent the diffusion rates of prey and predator,
,
represent the growth function of prey and predator, c is the positive rate of environment change,
indicates the cooperative behaviors with positive constants
for
. They showed that for any given positive speed of the shifting habitat edge, there exists a nondecreasing forced wave with the speed consistent with the habitat shifting speed. Hu et al. concerned with the forced traveling wave solution for the modified model of (1.1). They proved the existence of traveling wave solution for any positive constant shifting speed by constructing appropriate upper and lower solutions and using the method of monotone iteration, see [14] . The method of monotone iteration used for constructing the upper and lower solutions is an effective method for solving differential equations, and is widely used in many kinds of initial value problems and boundary value problems. This method is that if the problem has a pair of ordered lower solutions and upper solutions, then under certain conditions, the monotone iterative sequence can be constructed through the pair of lower solutions and upper solutions, so that they converge uniformly to the minimum and maximum solutions between the lower solutions and upper solutions of the equation. If the minimum solution is equal to the maximum solution, then the minimum solution (or maximum solution) is the forced traveling wave solution of the equation.
These prey-predator models are typically represented by the predator function as the increase in the number of predators after predation, while the decrease in the number of predators is due to natural death. However, a case shows that the increase of some predators is also not unlimited, and that the decrease in the number of predator populations is negatively correlated with the per capita availability of its preferred food, see [15] . Leslie in [16] first introduced a prey-predator model, combining the logistic predator equation with the carrying capacity proportional to the number of prey, in order to emphasize that predators have an upper limit on the rate of growth as prey. Since then, many researchers have studied the Leslie-Gower prey-predator model.
For the predator-prey relationships, a great deal of researchers have discussed the Leslie-Gower prey-predator model. In [17] , Fang mainly concerned the spatial dynamics of a modified Leslie-Gower prey-predator model in a shifting habitat, the main concerns are the extinction and persistent conditions under the interaction between two species with different diffusion speed comparing with the shifting habitat edge constant c, the model as follows
(1.2)
Changes in the environment lead to changes in the habitat boundaries of predator and prey, which are represented here by
. Because
, the habitat range is reduced. Only when the shifting speed of environment
,
, the population density of the two species will eventually reach an equilibrium state, that is, there is a forced wave. When the rate of species spread is less than the rate of environmental change, the species will go extinct. Recently, we considered the existence of forced waves and their asymptotic for (1.2). The existence of the forced waves indicates that the prey-predator system will eventually reach a state of equilibrium. Lee et al. in [18] noted that the free movement of some species can be over large areas, and nonlocal diffusion can effectively describe this phenomenon. Therefore, many researchers have considered the nonlocal diffusion species models. For example, Cheng and Yuan in [19] mainly studied the information about the existence and stability of the invasion traveling waves for the nonlocal Leslie-Gower predator-prey model. Then motivated by the aforementioned works, we concern a Leslie-Gower predator-prey model with a nonlocal predation under shifting environment as follows
(1.3)
where all parameters are assumed to be positive,
,
is a continuous and non-negative probability density function. u and v respectively stand for the prey populations
and the predator populations,
denotes the per capita capturing rate of the prey by a predator per unit time and
represents Leslie-Gower term.
In this paper, we will concern the forced traveling wave fronts of (1.3) connecting the trivial equilibrium and positive equilibrium. The results imply that for any given positive speed of the shifting habitat edge, there exists a forced wave with the speed in keeping with the habitat shifting speed. To prove this conclusion, we assume
(A1)
is symmetric and compactly supported, and
;
(A2)
is continuous, nondecreasing with
,
;
(A3)
,
and
;
(A4)
is continuously differentiable in
and both
exist,
.
Note that the system of (1.3) admits four equilibria
,
,
and
, where
This paper is organized as follows. In Section 2, by constructing a pair of appropriate upper and lower solutions of the Equation (2.5) and combined with the monotone iteration approaches, we can establish the existence of forced wave. In Section 3, we consider the asymptotical behavior of forced wave in two tails.
2. Existence of Forced Wave
In this section, we always assume that (A1), (A2) and (A3) hold. For simplicity, denote
(2.1)
For any
,
and
, we have
(2.2)
where
,
, which imply that
,
are Lipschitz continuous in
for any
,
. Define
(2.3)
Then
is nondecreasing in
and nonincreasing in
,
is nondecreasing in
,
. Let
be all continuous functions from
to
and
be all continuous and bounded functions from
to
. Denote
. Then it follows from [ [20] , Theorem 2.1] that the following conclusion holds.
Theorem 2.1. Considering the Cauchy problem
(2.4)
where
are defined in (2.3). If
with
,
in
, then (2.4) has a unique classic solution
with
,
for all
and
.
Let
and plugging it into (1.3), we have
(2.5)
Then we will consider the solution of (2.5) which satisfies the following asymptotic boundary condition
(2.6)
Lemma 2.1 1) Let
be the positive root of
with
,
. Then there exist
,
and small
,
such that the functions
satisfy
,
for all
, where
satisfies
and
.
2) Let
so small that
and
. Assume that
is the positive root of
and
is the positive root of
. Then the functions
,
satisfy
,
for all
.
Furthermore, there holds
(2.7)
(2.8)
(2.9)
and
(2.10)
Proof. 1) Let
be the positive root of the equation
, and
be the positive root of the equation
. Due to
be the positive root of
, we obtain
,
. Since
,
, we have
,
for all
.
2) Since
,
, we can obtain that there exists
small enough that
and
with
. By the definition of
and
, we have
for
. So, when
, we have
; when
,
, it can be seen that
for all
. Similarly, we can get that
for all
.
(i) If
, then
,
, and
. By the nondecreasing of
and the definition of
for
, we have
and
(ii) For
, obviously,
, then
,
,
. Due to the nondecreasing of
for
and the chosen of
, it is easy to show that
and
Thus we can obtain that (2.7) and (2.8) hold for
.
(iii) When
, then
. According to the nondecreasing of
and
, we can show
If
, obviously,
and
, then the definition of
and the nondecreasing of
can imply that
That is, (2.9) hold for
.
(iv) For
,
,
and
, then we can obtain
For
,
and
. Then by the definition of
and the nondecreasing of
, we can have
Thus we can deduce that (2.10) is right.£
Define
and
Then we can rewrite the system (2.5) as the following form
(2.11)
Notice that
is a bounded solution of (2.11) if and only if
is a fixed point of the operator
, where
and
(2.12)
Lemma 2.2. For
,
is nondecreasing in U and nonincreasing in V, and
is nondecreasing in
. Moreover, the operator Q maps Γ into Γ.
Proof. For any
with
, by the choice of
for
, we have
and
It can be seen that
is nondecreasing in U and nonincreasing in V, and
is nondecreasing in
. Hence, according to the definition of
for
and the nonnegativity of
defined in (2.12), we can obtain
and
So, the above inequalities show that
(2.13)
for all
. Next we show that Q maps Γ into Γ. For
, then by (2.7), we can obtain
For
, we have
Similarly, when
, then by (2.8), we have
For
, we can obtain
Since
,
are continuously differentiable in
, we have
,
for
. By the continuity of
, we can further have that
,
for all
.
Using a similar argument with (2.9) and (2.10), we have
for
,
for
, and
for
,
for
. Since
,
,
,
, we can prove that
for
,
for
. Therefore, by the continuity of
,
,
for all
. These, together with (2.13), we obtain that
maps Γ into Γ.
Theorem 2.2. Assume that (A1), (A2) and (A3) hold. Then (1.3) has a forced wave
.
Proof. Define
,
,
,
, then we have
Define the following iterations
for
. Next, when
,
,
,
,
apparently hold, according to induction, when
,
,
,
,
hold.
Moreover, since
is nondecreasing in U and nonincreasing in V, and
is nondecreasing in
, we have when
,
apparently,
and
hold, according to induction, when
,
and
hold.
From all of these, we can conclude that
(2.14)
for all
and
. Thus
,
,
,
all exist. By (2.14), it is easy to see that
By the continuity of
and
, we know that
,
,
,
converge point-wise to
,
,
,
, respectively. Since
then by the Lebesgue’s dominated convergence theorem, we have
and
Similar to the proof of Theorem 4.2.7 in [21] , we can obtain that
Thus
is a solution of (2.11) and satisfies
Next, we prove that
satisfies the asymptotical boundary conditions (2.6). Since
We can easily deduce
(2.15)
Since
,
and
,
,
,
are bounded in
, we can obtain that
,
,
,
exist and denote them by
, respectively. Also,
,
and
In view of L’Hôpital’s rule, we can obtain
and
Hence,
,
,
,
. Since
,
, it must be
,
,
,
. By some trivial calculations, we can show that
,
, that are
,
. Combining with (2.15), the asymptotic boundary conditions (2.6) are satisfied.£
3. Asymptotic Behaviors of Forced Waves
To obtain more asymptotic information of the forced waves at
, we denote the solution of (2.5) with (2.6) as
Differentiating (2.5) with respect to
, then we show that
satisfies
(3.1)
Theorem 3.1. Assume that (A1)-(A4) hold and
. Then there exist positive constants
,
,
with
such that the forced wave front
to (1.3) has the following asymptotic properties
as
; and
as
provided
. Here
are given in (3.7).
Proof. For
, the limiting equations for (3.1) is rewritten as the following form
(3.2)
So, the first equation of (3.2) has two independent solutions
Combining (3.1) with (3.2), we have that
admits the following property as
,
Since
, it must be
. Hence, for
,
Similarly, we can deduce that for
,
Using the integration on
and
from
to
, there are two constants
and
such that
as
. For
, the limiting equations for (3.1) are
(3.3)
where
Note that in (3.3), we have used the fact that
with
. Indeed, by applying L’Hôpital’s rule, we have
which implies
. Choosing
,
, then we can rewrite (3.3) as the following first-order differential equations
(3.4)
Then we can get the characteristic equation of (3.4) as
(3.5)
If
, then (3.5) can be simplified as
Define
. Then s satisfies
. Since
We have
,
. Thus the general solution corresponding to (3.4) can be expressed as
(3.6)
where
(3.7)
and
are eigenvectors corresponding to
,
are arbitrary constants with
. Since
as
, we deduce that
and
from (3.6). Thus,
. Thus,
satisfies the following property
as
, where
are constants and
cannot be zero simultaneously,
. Meanwhile,
,
. For the solution
to (3.4), if one of the first and third components of the eigenvectors
is zero, then the linear system (3.4) leads to the other components are also zero. By integrating from
to
, it follows that
as
and
.
Acknowledgments
The authors would like to thank the anonymous referee for their careful reading and valuable comments.