International Journal of Modern Mathematical Sciences, 2014, 11(3): 130-143

International Journal of Modern Mathematical Sciences ISSN:2166-286X

Journal homepage:www.ModernScientificPress.com/Journals/ijmms.aspx Florida, USA
Article

Stability Analysis and Hopf Bifurcation for a Delayed Logistic

Equation with Strong Allee Effect

E.M.Elabbasya, A. A. Elsadanyb,c*, Waleed A.I. Elmorsia

a
Department of Mathematics, Faculty of Science, Mansoura University, Mansoura 35516, Egypt
b
Department of Basic Science, Faculty of Computers and Informatics, Suez Canal University, Ismailia
41522, Egypt
c
Department of Mathematics, Shanghai University, Shanghai 200444, China

* Author to whom corresponding should be addressed; Email: [email protected]

Article history: Received 27 July 2014, Accepted 21 August 2014, Published 26 September 2014.

Abstract: In this paper, we study the stability of a delayed logistic equation with strong Allee
effect. By using the associated characteristic transcendental equation, we show that the
occurrence of Hopf bifurcation at the positive equilibrium. The direction and the stability of
the bifurcating periodic solutions are obtained by the normal form theory and center manifold
theorem. Finally, a numerical example is given to demonstrate the effectiveness of the
theoretical analysis.

Keywords: Delay logistic equation, Stability, Hopf bifurcation analysis, Allee effect.

Mathematical Subject Classification (2000): 34K20, 34K25.

1. Introduction

The logistic equation [1] is a modification of the exponential equation model which takes into
account the limitation of the environmental resources. The logistic equation (sometimes called the
Verhulst model) is a model of population growth first proposed by Pierre-Francois Verhulst (1845, 1847).
Verhulst [2] proposed that a self-limiting process should operate when a population becomes too large in
the form

Copyright © 2014 by Modern Scientific Press Company, Florida, USA

Int. J. Modern Math. Sci. 2014, 11(3): 130-143 131

𝑥(𝑡)
𝑥̇ (𝑡) = 𝑟𝑥(𝑡) (1 − ) (1)
𝑘

Equation (1) called logistic growth in a population Where 𝑟 is the intrinsic growth rate and 𝑘 is
the carrying capacity (the maximum number of individuals that the environment can support).The logistic
equation has variety of applications. It is used in neural networks to clamp signals to within a specified
range, in economics to illustrate the progress of the diffusion of an innovation through its life cycle, in
medicine to model the growth of tumors and in linguistics to model language change [3].
Since, using ordinary differential equation models for population growth often do not tend to
represent the population as accurately as one might initially expect, time delays have been incorporated
into biological and ecological models to represent some phenomena as resource regeneration times,
maturation periods, feeding times, reaction times, etc. Furthermore [4], so many of the processes, both
natural and manmade, in medicine, chemistry, physics, engineering, economics, physiological diseases
etc., involve time delays. In general, delay differential equations exhibit much more complicated
dynamics than ordinary differential equations. For these reasons, mathematicians, biologists, ecologists,
epidemiologists and economists have given much interest to the delayed systems. [3-11, 20]
Hutchinson [10] proposed more realistic equation by incorporating the effect of delay into the logistic
equation
𝑥(𝑡−𝜏)
𝑥̇ (𝑡) = 𝑟𝑥(𝑡) (1 − ) (2)
𝑘

It has been suggested by Hutchinson that the equation (2) can be used to model the dynamics of
a single species population growing towards a saturation level k with a constant reproduction rate r [11].
Delays bring about interesting topological changes in the population size like damped oscillations, limit
cycles and even chaos [12]. The inclusion of delay in such models has illustrated more complicated and
richer dynamics in terms of stability, bifurcation, periodic solutions and so on [13].
Some species often cooperate among themselves in their search for food and to escape from their
predators. For example a number of social species such as ants, termites, bees, etc., have developed
complex cooperative behavior involving division of labor, altruism, etc. Such cooperative processes have
a positive feedback influence since individuals have been provided a greater chance to survive and
reproduce as density increase.
Aggregation and associated cooperative and social characteristics among members of a species
were extensively studied in animal populations by Allee (1931). The phenomenon in which reproduction
rates of individuals decrease when density drops below a certain critical level is now known as the Allee
effect. [10].
Allee effects, defined as positive effects of increasing density on fitness, are a predicted
consequence of multiple mechanisms in small or low density populations, and are expected to affect

Copyright © 2014 by Modern Scientific Press Company, Florida, USA

Int. J. Modern Math. Sci. 2014, 11(3): 130-143 132

many population-level processes, including extinction, population establishment and biological invasion,
meta population fluctuations, predator–prey interactions and parasite transmission [14].
The so-called Allee effect [15] refers to a population that has a maximal per capita growth rate at
intermediate density. This occurs when the per capita growth rate increases as density increases and
decreases after the density passes a certain value.
The equation
𝑥(𝑡) 𝑥(𝑡)
𝑥̇ (𝑡) = 𝑟𝑥(𝑡) (1 − )( − 1) (3)
𝑘 𝑎

is the prototypical model for a strong Allee effect [16-18].

In our paper, we study the delayed logistic equation with strong Allee effect in the form
𝑥(𝑡−𝜏) 𝑥(𝑡)
𝑥̇ (𝑡) = 𝑟𝑥(𝑡 − 𝜏) (1 − )( − 1) (4)
𝑘 𝑎

Where 𝑘 > 0 is the carrying capacity and 𝑎 > 0 is Allee threshold (the minimal size of the population
required to survive).
The purpose of this paper is to discuss the stability and the Hopf bifurcation properties of model
(4). More specifically, we will prove that, as the delay τ increases, the positive equilibrium point loses its
stability and a sequence of Hopf bifurcations occur. Furthermore, using the center manifold theory[19]
and normal form, we derive sufficient conditions for the stability of the bifurcating periodic solutions.

2. Stability of the Equilibrium Points and Existence of Local Hopf Bifurcations

In this section, we shall study the stability of the positive equilibrium points and the existence of
local Hopf bifurcations. The model (4) has three equilibria 𝑥1∗ = 0 , 𝑥2∗ = 𝑘 and 𝑥3∗ = 𝑎.
For convenience, we indicate to the next lemma which consolidates our stability analysis.
Lemma 1. [20]
All roots of the characteristic equation 𝜆 + 𝑐 + 𝑏𝑒 𝜆 = 0 , where 𝑐 and 𝑏 are real, have negative real
parts if and only if:-
1. 𝑐 > −1;
2. 𝑐 + 𝑏 > 0 𝑎𝑛𝑑
3. 𝑏 < √𝑐 2 + 𝜉 2 .
𝜋
Where 𝜉 is the root of 𝜉 = −𝑐 tan 𝜉, 0 < 𝜉 < 𝜋 if 𝑐 ≠ 0 and 𝜉 = if 𝑐 = 0.
2

Theorem 1.
𝜋 𝜋
1) The equilibrium point 𝑥1∗ = 0 of Eq. (4) is stable if 𝑟𝜏 < and is unstable if 𝑟𝜏 > 2 .
2

Copyright © 2014 by Modern Scientific Press Company, Florida, USA

Int. J. Modern Math. Sci. 2014, 11(3): 130-143 133

𝑎𝜋 𝑎𝜋
2) The equilibrium point 𝑥2∗ = 𝑘 of Eq. (4) is stable if 𝑟𝜏 < 2(𝑘−𝑎) and is unstable if 𝑟𝜏 > 2(𝑘−𝑎).

3) The equilibrium point 𝑥3∗ = 𝑎 of Eq. (4) is unstable.

Proof. By linearizing (4) about equilibrium 𝑥 ∗ = 0 using𝑥(𝑡) = 𝑥 ∗ + 𝑝(𝑡), it becomes
𝑝̇ (𝑡) = −𝑟𝑝(𝑡 − 𝜏) (5)
It is easy to show that equation(5) has the characteristic equation in the form
𝜆 + 𝑟𝑒 −𝜆𝜏 = 0 (6)
Suppose that 𝜆 = 𝑖𝜔 be the root of equation (6) then
𝑖𝜔 + 𝑟 𝑒 −𝑖𝜔𝜏 = 0 (7)
This can be written as
𝑟 cos 𝜔𝜏 + 𝑖(𝜔 − 𝑟 sin 𝜔𝜏) = 0 (8)
By separating and equating real parts and imaginary parts, we have
𝑟 cos 𝜔𝜏 = 0 (9. 𝑎)
ω − r sin ωτ = 0 (9. 𝑏)
𝜋
Eq. (9. 𝑎) implies that 𝜔𝜏 = cos−1 (0) = (2𝑛 + 1) 2 , 𝑛 = 0 ,1,2, …

Substituting in equation (9. 𝑏), we find

𝜋 𝜋
− 𝑟 sin = 0
2𝜏 2
𝜋
𝑟𝜏 =
2
Then the necessary and sufficient condition for stability is

𝜋
𝑟𝜏 < (10)
2
Again, by linearizing about x2∗ = k equation(4) will be
𝑟(𝑎 − 𝑘)
𝑝̇ (𝑡) = 𝑝(𝑡 − 𝜏) (11)
𝑎
This equation has the characteristic equation
𝜆 − 𝐴1 𝑒 −𝜆𝜏 = 0 (12)

𝑟(𝑎−𝑘)
Where 𝐴1 = (13)
𝑎

Let 𝑖𝜔 be the root of equation (12); then:

𝑖𝜔 − 𝐴1 (cos 𝜔𝜏 − 𝑖 sin 𝜔𝜏) = 0
−𝐴1 cos 𝜔𝜏 + 𝑖(𝜔 + 𝐴1 sin 𝜔𝜏) = 0 (14)
Then, by separating and equating real parts and imaginary parts
−𝐴1 cos 𝜔𝜏 = 0 (15. 𝑎)

Copyright © 2014 by Modern Scientific Press Company, Florida, USA

Int. J. Modern Math. Sci. 2014, 11(3): 130-143 134

𝜔 + 𝐴1 sin 𝜔𝜏 = 0 (15. 𝑏)
From Eq. (15. 𝑎), we have
𝜋
𝜔𝜏 = cos−1 0 = (2𝑛 + 1) , 𝑛 = 0,1,2, … (16 )
2
At n = 0 (minimum order solution)
𝜋
𝜔𝜏 = (17)
2
From Eq. (17) substitute in Eq. (15. 𝑏)

𝑎𝜋
𝜏=
2𝑟(𝑘 − 𝑎)
So the necessary and sufficient condition for stability is
𝑎𝜋
𝑟𝜏 < (18)
2(𝑘 − 𝑎)
Once again, by linearizing Eq.(4) about x3∗ = a, it will be
𝑝̇ (𝑡) = 𝐴2 𝑝(𝑡) (19)

𝑟(𝑘−𝑎)
Where 𝐴2 = (20)
𝑘

Equation (19) has the characteristic equation

𝑟(𝑘 − 𝑎 )
𝜆 = 𝐴2 = (21)
𝑘
𝑟(𝑘−𝑎)
Since 𝑎 < 𝑘 ⇒ > 0 , then the system is unstable at x3∗ = a . ∎
𝑘

Theorem.2 If 𝜆𝑗 (τ) = αj (τ) + i𝜔𝑗 (τ) denote a root of Eq. (6) near τ = τ𝑗 , such that 𝜔𝑗 (τ𝑗 ) = 𝜔0
and αj (τ𝑗 ) = 0 then
dαj (τ)
| ≠0
dτ τ=τ
𝑗

Proof. By differentiating the characteristic equation (6) with respect 𝜏 we get

𝑑𝜆 𝑟𝜆𝑒 −𝜆𝜏
=
𝑑𝜏 1 − 𝑟𝜏𝑒 −𝜆𝜏
This gives
𝑑𝜆 −1 𝑒 𝜆𝜏 𝜏
( ) = −
𝑑𝜏 𝑟𝜆 𝜆
Then

Copyright © 2014 by Modern Scientific Press Company, Florida, USA

Int. J. Modern Math. Sci. 2014, 11(3): 130-143 135

dαj (τ) 𝑑𝜆 −1 𝑒 𝜆𝜏 𝜏
Sign { | } = sign {𝑅𝑒 (𝑑𝜏 ) | } = sign {𝑅𝑒 ( 𝑟𝜆 − 𝜆)| }
dτ τ=τ𝑗 τ=τ𝑗
τ=τ𝑗

𝑠𝑖𝑛𝜔0 𝜏𝑗 1
= sign { } = sign { 2 } > 0 . ∎
𝑟𝜔0 𝑟
Theorem.3 if 𝜆𝑗 (τ) = αj (τ) + i 𝜔𝑗 (τ) denote a root of (12) near τ = τ𝑗 , such that 𝜔𝑗 (τ𝑗 ) = 𝜔0 and
αj (τ𝑗 ) = 0 then
dαj (τ)
| ≠0
dτ τ=τ
𝑗

Proof. By differentiating the characteristic equation (12) with respect 𝜏 we get

𝑑𝜆 −𝐴1 𝜆𝑒 −𝜆𝜏
=
𝑑𝜏 1 + 𝐴1 𝜏𝑒 −𝜆𝜏
This gives
𝑑𝜆 −1 𝑒 𝜆𝜏 𝜏
( ) = −
𝑑𝜏 𝐴1 𝜆 𝜆
Then
dαj (τ) 𝑑𝜆 −1 𝑒 𝜆𝜏 𝜏
Sign { | } = sign {𝑅𝑒 (𝑑𝜏 ) | } = sign {𝑅𝑒 (𝐴 − 𝜆)| }
dτ τ=τ𝑗 1𝜆 τ=τ𝑗
τ=τ𝑗

𝑠𝑖𝑛𝜔0 𝜏𝑗 1
= sign { } = sign { 𝑟𝐴 } > 0 . ∎
𝐴1 𝜔 0 1

This proof stated the transversality condition for the occurrence of Hopf bifurcations.
Then the proof of the theorem has been gained. ∎
Theorems 2 and 3 stated the last condition for the occurrence of Hopf bifurcations and the results
can be introduced as follow:
Theorem4:
𝜋
1) When the parameter τ passes through the critical value τ1 = 2𝑟 , there are Hopf bifurcations at the

equilibrium 𝑥2∗ = 0 to a periodic orbit.

𝑎𝜋
2) When the parameter τ passes through the critical value τ2 = 2𝑟(𝑘−𝑎), there are Hopf bifurcations at

the equilibrium 𝑥2∗ = 𝑘 to a periodic orbit .

In the previous section, we have obtained the conditions that lead to Hopf bifurcation. In this
section, we will study the direction of Hopf bifurcation and the stability of bifurcating periodic solutions
by applying normal form theory and the center manifold theorem in [19].

3. Stability and Direction of the Hopf Bifurcation

Copyright © 2014 by Modern Scientific Press Company, Florida, USA

Int. J. Modern Math. Sci. 2014, 11(3): 130-143 136

Let 𝑦(𝑡) = 𝑥(𝜏𝑡), then the equation (4) written as

𝑦(𝑡)𝑦(𝑡 − 1) 𝑦(𝑡)𝑦 2 (𝑡 − 1) 𝑦 2 (𝑡 − 1)
𝑦̇ (𝑡) = −𝑟𝜏𝑦(𝑡 − 1) + 𝑟𝜏 ( − − ) (22)
𝑎 𝑎𝑘 𝑘
In ∁ = ∁ ( [−1, 0], ℝ) equation (22) written as
𝑦̇ (𝑡) = 𝐿𝜇 (𝑦𝑡 ) + 𝐹 (𝜇 , 𝑦𝑡 ) (23)
Where
𝐿𝜇 (𝜑) = −𝑟(𝜏𝑘 + 𝜇)𝜑(−1)
𝜑(0)𝜑(−1) 𝜑(0)𝜑 2 (−1) 𝜑 2 (−1)
𝐹 (𝜇 , 𝜑) = 𝑟(𝜏𝑘 + 𝜇) { − − } (24)
𝑎 𝑎𝑘 𝑘
Using the Riesz representation theorem, there exists a function 𝜂 (𝜃, 𝜇)of bounded variation for
𝜃 ∈ [−1,0] such that:-
0
𝐿𝜇 (𝜑) = ∫ 𝑑𝜂 (𝜃, 0)𝜑(𝜃) for 𝜑𝜖𝑐 (25)
−1

We can choose
𝜂 (𝜃, 𝜇 ) = −𝑟(𝜏𝑗 + 𝜇)𝛿(𝜃 + 1) (26)
where 𝛿(𝜃)is the Dirac delta function.
For 𝜑 𝜖 ∁([−1,0], ℝ); define
𝑑𝜑(𝜃)
𝑖𝑓 𝜃 ∈ [−1,0)
𝑑𝜃
𝐴(𝜇)𝜑 = 0 (27)
∫ 𝑑𝜂 (𝜇, 𝑠)𝜑(𝑠)𝑖𝑓 θ = 0
{ −1

And
0 𝑖𝑓 𝜃 ∈ [−1,0)
𝑅(𝜇)𝜑 = { (28)
𝐹(𝜇, 𝜑) 𝑖𝑓 𝜃 = 0

Then the system (23) is equivalent to

𝑦̇ 𝑡 = 𝐴(𝜇)𝑦𝑡 + 𝑅(𝜇)𝑦𝑡 (29)
where
𝑦𝑡 (𝜃) = 𝑦(𝑡 + 𝜃) 𝑓𝑜𝑟 𝜃 ∈ [−1,0]
The equation (29) is more mathematically tractable because this equation involves a single unknown
variable 𝑦𝑡 .
For Ψ𝜖𝑐[0, −1] , the adjoint operator 𝐴∗ of 𝐴 is defined as
𝑑Ψ(𝑠)
𝑖𝑓 𝑠 ∈ [−1,0)
∗ (𝜇)Ψ(𝑠)
𝐴 = { 0 𝑑𝑠 (30)
∫−1 𝑑𝜂 (𝜇, 𝑠)Ψ(𝑠) 𝑖𝑓 s = 0

Copyright © 2014 by Modern Scientific Press Company, Florida, USA

Int. J. Modern Math. Sci. 2014, 11(3): 130-143 137

For𝜑 𝜖 ∁([−1,0]and Ψ 𝜖 ∁[0, −1] , the bilinear inner product defined as

𝜃
0
̅ (0)𝜑(0) − ∫
⟨ψ(𝑠), 𝜑(𝜃)⟩ = ψ ̅ (𝜉 − 𝜃) 𝑑𝜂 (𝜃)𝜑(𝜉)𝑑𝜉
∫ψ (31)
−1
𝜉=0

To determine the Poincare normal form of the operator 𝐴 we need to calculate the eigenvector
𝑞(𝜃) and 𝑞 ∗ (𝜃) of 𝐴 and 𝐴∗ that corresponding to the eigenvalues 𝑖𝜔0 𝜏j and −𝑖𝜔0 𝜏j respectively.
It is easy to be verified that 𝑞(𝜃) = 𝑒 𝑖𝜔0 𝜏j𝜃 and 𝑞 ∗ (𝜃) = 𝐷𝑒 𝑖𝜔0 𝜏j𝜃 .
In order to assure that ⟨𝑞 ∗ (𝑠), 𝑞(𝜃)⟩ = 1, we need to determine the value of 𝐷.
From (31)
0 𝜃
̅ 𝑒 −𝑖𝜔0 𝜏j(𝜉−𝜃) 𝑑𝜂 (𝜃)𝑒 𝑖𝜔0 𝜏j𝜉 𝑑𝜉
⟨𝑞 ∗ (𝑠), 𝑞(𝜃)⟩ = 𝑞̅ ∗ (0)𝑞(0) − ∫ ∫ 𝐷
−1 𝜉=0
0
⟨𝑞 ∗ (𝑠), ̅ [1 − ∫ 𝜃 𝑒 𝑖𝜔0 𝜏j𝜃 𝑑𝜂 (𝜃)] = 1
𝑞(𝜃)⟩ = 𝐷
−1

1
̅=
𝐷 (32)
1 − 𝑟𝜏𝑗 𝑒 −𝑖𝜔0 𝜏j
Using the same notation introduced by Hassard et al [19], we compute the co-ordinates to
describe the center manifold 𝑐0 at= 0 .
For 𝑦𝑡 , a solution of (29) at 𝜇 = 0 , we define:
𝑧(𝑡) = ⟨𝑞 ∗ ,𝑦𝑡 ⟩and 𝑤(𝑡, 𝜃) = 𝑦𝑡 (𝜃) − 2𝑅𝑒{𝑧(𝑡)𝑞(𝑡)} (33)
On the center manifold 𝑐0 ,we have:
𝑤(𝑡, 𝜃) = 𝑤(𝑧(𝑡), 𝑧̅(𝑡), 𝜃) where
𝑧2 𝑧̅ 2
𝑤(𝑡, 𝜃) = 𝑤20 (𝜃) + 𝑤11 (𝜃)𝑧𝑧̅ + 𝑤02 (𝜃) + ⋯ (34)
2 2
Where 𝑧 𝑎𝑛𝑑 𝑧̅ are local co-ordinates for 𝑐0 in ∁ in the direction of 𝑞 ∗ and 𝑞̅ ∗ . Note that , 𝑤 is real if 𝑦𝑡
is real. We shall deal with real solution only.
Now, for solution 𝑦𝑡 𝜖 𝑐0 of (29)
⟨𝑞 ∗ ,𝑦̇ 𝑡 ⟩ = ⟨𝑞 ∗ , 𝐴(𝜇)𝑦𝑡 + 𝑅(𝜇)𝑦𝑡 ⟩, then since 𝜇 = 0
𝑧̇ (𝑡) = ⟨𝑞̇ ∗ ,𝑦𝑡 ⟩ + ⟨𝑞 ∗ , 𝐴(𝜇)𝑦𝑡 + 𝑅(𝜇)𝑦𝑡 ⟩
𝑧̇ (𝑡) = 𝑖𝜔0 𝜏𝑗 𝑧(𝑡) + 𝑞 ∗ (0)𝑓(𝑧(𝑡), 𝑧̅(𝑡)) (35)
Equation (35) can be written in abbreviated form as
𝑧̇ (𝑡) = 𝑖𝜔0 𝜏j 𝑧(𝑡) + 𝑔(𝑧, 𝑧̅) (36)
Where
𝑔(𝑧, 𝑧̅) = 𝑞 ∗ (0)𝑓(𝑧(𝑡), 𝑧̅(𝑡))
𝑧2 𝑧̅ 2
𝑔(𝑧, 𝑧̅) = 𝑔20 (𝜃) + 𝑔11 (𝜃)𝑧𝑧̅ + 𝑔02 (𝜃) + ⋯ (37)
2 2

Copyright © 2014 by Modern Scientific Press Company, Florida, USA

Int. J. Modern Math. Sci. 2014, 11(3): 130-143 138

Since from (33)

𝑤̇ = 𝑦̇ 𝑡 − 𝑧̇ 𝑞 − 𝑧̅̇𝑞̅
Then

𝐴𝑤 − 2𝑅𝑒[𝑞 ∗ (0)𝑓(𝑧(𝑡), 𝑧(𝑡))𝑞] 𝑖𝑓 𝜃 ∈ [−1,0)

𝑤̇ = { (38)
𝐴𝑤 − 2𝑅𝑒[𝑞 ∗ (0)𝑓(𝑧(𝑡), 𝑧(𝑡))𝑞] + 𝑓(𝑧, 𝑧) 𝑖𝑓 𝜃 = 0
This can be written as
𝑤̇ = 𝐴𝑤 + 𝐻(𝑧(𝑡), 𝑧̅(𝑡), 𝜃) (39)
where
𝑧2 𝑧̅ 2
𝐻(𝑧(𝑡), 𝑧̅(𝑡), 𝜃) = 𝐻20 (𝜃) + 𝐻11 (𝜃)𝑧𝑧̅ + 𝐻02 (𝜃) + ⋯ (40)
2 2
Since
𝑔(𝑧, 𝑧̅) = 𝑞 ∗ (0)𝑓(𝑧(𝑡), 𝑧̅(𝑡))
= 𝑞 ∗ (0)𝑓(0, 𝑦𝑡 )
̅ 𝑟𝜏j
𝐷
= { 𝑘𝑦𝑡 (0)𝑦𝑡 (−1) − 𝑦𝑡 (0)𝑦𝑡 2 (−1) − 𝑎𝑦𝑡 2 (−1)} (41)
𝑎𝑘

Using equations(33) in (41) and comparing coefficient with (37), we find

̅ 𝑟𝜏j
𝐷
𝑔20 = ( 𝑘𝑒 −𝑖𝜔0 𝜏j − 𝑒 2𝑖𝜔0 𝜏j − 𝑎𝑒 −2𝑖𝜔0 𝜏j ) (42)
𝑎𝑘
̅ 𝑟𝜏j
𝐷
𝑔11 = ( 𝑘𝑒 𝑖𝜔0 𝜏j + 𝑘 𝑒 −𝑖𝜔0 𝜏j − 2𝑎) (43)
𝑎𝑘
̅ 𝑟𝜏j
𝐷
𝑔02 = ( 𝑘𝑒 𝑖𝜔0 𝜏j − 𝑎𝑒 2𝑖𝜔0 𝜏j ) (44)
𝑎𝑘
̅ 𝑟𝜏j
𝐷 1 1
𝑔21 = {𝑘 (𝑒 −𝑖𝜔0 𝜏j 𝑤11 (0) + 𝑒 𝑖𝜔0 𝜏j 𝑤20 (0) + 𝑤20 (−1) + 𝑤11 (−1))
𝑎𝑘 2 2

− 𝑎 (2𝑒 −𝑖𝜔0 𝜏j 𝑤11 (−1) + 𝑒 𝑖𝜔0 𝜏j 𝑤20 (−1)) − 𝑒 2𝑖𝜔0 𝜏j

− 2} (45)

Since 𝑤̇ (𝑧, 𝑧̅) = 𝑤𝑧 𝑧̇ + 𝑤𝑧̅ 𝑧̅̇ (46)

Using(39) and substitution by the expansions of previous functions and comparing coefficients,
we find that

𝐻20 (𝜃) = (2𝑖𝜔0 𝜏j – 𝐴)𝑤20 (𝜃) (47)

𝐻11 (𝜃) = (2𝑖𝜔0 𝜏j – 𝐴)𝑤11 (𝜃) (48)

Copyright © 2014 by Modern Scientific Press Company, Florida, USA

Int. J. Modern Math. Sci. 2014, 11(3): 130-143 139

𝐻02 (𝜃) = (2𝑖𝜔0 𝜏j – 𝐴)𝑤02 (𝜃) (49)

Since from (38), (39) and (40) we find
𝐻20 (𝜃) = −𝑞(𝜃)𝑔20(𝜃) − 𝑞̅(𝜃)𝑔̅02 (𝜃) (50)
𝐻11 (𝜃) = −𝑞(𝜃)𝑔11 (𝜃) − 𝑞̅(𝜃)𝑔̅11 (𝜃) (51)
𝐻02 (𝜃) = −𝑞(𝜃)𝑔20 (𝜃) − 𝑞̅(𝜃)𝑔̅20 (𝜃) (52)
From (47) and (50), we have
𝐴𝑤20 (𝜃) = 2𝑖𝜔0 𝜏j 𝑤20 − 𝐻20 (𝜃)
𝑤̇20 (𝜃) = 2𝑖𝜔0 𝜏j 𝑤20 + 𝑞(𝜃)𝑔20 (𝜃) + 𝑞̅(è)𝑔̅02 (𝜃) (53)
This equation has the solution
𝑔20 𝑖𝜔0 𝜏 𝜃 𝑔̅02 −𝑖𝜔0 𝜏 𝜃
𝑤20 (𝜃) = − 𝑒 j − 𝑒 j + 𝐸 𝑒 2𝑖𝜔0 𝜏j 𝜃 (54)
1
𝑖𝜔0 𝜏j 3𝑖𝜔0 𝜏j
By the same way
𝑔11 𝑖𝜔0 𝜏 𝜃 𝑔̅11 −𝑖𝜔0 𝜏 𝜃
𝑤11 (𝜃) = 𝑒 j − 𝑒 j +𝐸
2 (55)
𝑖𝜔0 𝜏j 𝑖𝜔0 𝜏j
Where 𝐸1 and 𝐸2 are constants and they are evaluated from the formulas
̅ 𝑟[𝑘𝑒 −𝑖𝜔0 𝜏j − 𝑒 2𝑖𝜔0 𝜏j − 𝑎𝑒 −2𝑖𝜔0 𝜏j ]
𝐷
𝐸1 = (56)
𝑎𝑘 [ 2𝑖𝜔0 + 𝑟𝑒 −2𝑖𝜔0 𝜏j ]
̅ 𝑟[𝑘𝑒 𝑖𝜔0 𝜏j + 𝑘 𝑒 −𝑖𝜔0 𝜏j − 2𝑎]
𝐷
𝐸2 = (57)
𝑎𝑘 [ 2𝑖𝜔0 + 𝑟 ]

We can also compute

𝑖 |𝑔02 |2 𝑔21
𝑐1 (0) = (𝑔20 𝑔11 − 2|𝑔11 |2 − )+
2𝜔0 𝜏j 3 2
−ℜ𝑒{ 𝑐1 (0)} (58)
𝜇2 =
ℜ𝑒{𝜏j 𝛼̀ (𝜏j )}
{ 𝛽2 = 2ℜ𝑒{ 𝑐1 (0)}

Then a general result for the direction and stability of Hopf bifurcation (see [19]), implies that
the direction of the Hopf bifurcation is determined by the sign of 𝛽2, and the stability of the bifurcating
periodic solutions is determined by the sign of 𝜇2 . In this case, if 𝜇2 > 0(< 0), then the Hopf bifurcation
is supercritical (subcritical) and if 𝛽2 < 0(> 0)the bifurcating periodic solutions are orbitally stable
(unstable).

4. Numerical Simulations and Discussion

Copyright © 2014 by Modern Scientific Press Company, Florida, USA

 Int. J. Modern Math. Sci. 2014, 11(3): 130-143 140

In this section, we shall carry out some numerical simulations for supporting our theoretical
analysis.
We choose 𝑘 = 10 , 𝑎 = 8 and 𝑟 = 2 for the model (4).
First, we choose 𝜏 = 0.75, the corresponding phase portrait is shown in Fig. 1(a), and waveform
plot is shown in Fig. 1 (b). By Theorem 1, we can see that its zero solution is asymptotically stable.

6 6

4 4

2 2
x(t-tau)

x(t)
0 0

-2 -2

-4 -4

-6 -6

-10 -8 -6 -4 -2 0 2 4 6 8 10 0 10 20 30 40 50 60 70 80 90 100
x(t) t

Fig. 1: (a) Phase portrait of model (4), (b) when Waveform plot of model (4) τ =0.75.

Second, choosing, 𝜏 = 0.85, we can find that Hopf bifurcation happens at the equilibrium point.
The model (4) changes from an equilibrium state to a cycle. The phase portrait shown in Fig. 2 (a) and
the waveform plot shown in Fig. 2 (b).

6
6

4
4

2
x(t-tau)

0
x(t)

-2
-2
-4
-4
-6
-10 -8 -6 -4 -2 0 2 4 6 8 10 -6
x 0 10 20 30 40 50 60 70 80 90 100
t

Fig. 2: (a) Phase portrait of model (4), (b) when Waveform plot of model (4) τ =0.85.

Finally, choosing τ =1, the corresponding Phase portrait is shown in Fig. 3 (a) and waveform plot
is shown in Fig. 3 (b). It is easy to see that in Fig. 3 (a) and Fig. 3 (b) undergo a Hopf bifurcation.

Copyright © 2014 by Modern Scientific Press Company, Florida, USA

 Int. J. Modern Math. Sci. 2014, 11(3): 130-143 141

6 6

4 4

2 2
x(t-tau)

x(t)
0 0

-2 -2

-4 -4

-6 -6
-10 -8 -6 -4 -2 0 2 4 6 8 10 0 10 20 30 40 50 60 70 80 90 100
x t

Fig. 3: (a) Phase portrait of model (4), (b) when Waveform plot of model (4) τ =1.

We also consider model (4) and find that there are many periodic solutions, as plotted Fig. 4.
Choosing the parameters 𝜏 = 0.75 , 𝑘 = 10 , 𝑎 = 8 , 𝑟 = 2 and 𝜔0 = 0.7854 to find the

values of (58) we got 𝑐1 (0) = 0.0018 + 0.0285𝑖 and 𝜇2 = −0.0096 and 𝛽2 = 0.0036. This implies
that the Hopf bifurcation is subcritical and the bifurcating periodic solutions are orbitally unstable.
6
6

4
4

2 2
x(t-tau)

x(t)

0 0

-2 -2

-4
-4

-6
-6

-10 -8 -6 -4 -2 0 2 4 6 8 10 0 10 20 30 40 50 60 70 80 90 100
x t

Fig. 4: (a) Phase portrait of model (4), (b) when Waveform plot of model (4) τ =1.2.

Choosing the parameters 𝜏 = 1 ,𝑘 = 10 , 𝑎 = 8 , 𝑟 = 2 and 𝜔0 = 1.5708 to find the values

of (58) we got 𝑐1 (0) = −0.0207 + 0.0258𝑖 and 𝜇2 = 0.0828 and 𝛽2 = −0.0414. This implies that
the Hopf bifurcation is supercritical and the bifurcating periodic solutions are orbitally stable.

Copyright © 2014 by Modern Scientific Press Company, Florida, USA

Int. J. Modern Math. Sci. 2014, 11(3): 130-143 142

Choosing the parameters 𝜏 = 0.79 ,𝑘 = 10 , 𝑎 = 8 , 𝑟 = 2 and 𝜔0 = 1.9883 to find the

values of (58) we got 𝑐1 (0) = −0.0158 + 0.0228𝑖 and 𝜇2 = 0.08427 and 𝛽2 = −0.0316. This
implies that the Hopf bifurcation is supercritical and the bifurcating periodic solutions are orbitally stable.
Choosing the parameters 𝜏 = 3.1416 , 𝑘 = 10 , 𝑎 = 8 , 𝑟 = 2 and 𝜔0 = 0.5 to find the

values of (58) we got 𝑐1 (0) = −0.0158 + 0.0270𝑖and 𝜇2 = 0.005029 and 𝛽2 = −0.0316. This
implies that the Hopf bifurcation is supercritical and the bifurcating periodic solutions are orbitally stable.

5. Conclusions

In this paper, We have investigated the stability and Hopf bifurcation of a delayed logistic
equation with strong Allee effect. A stability conditions for the delayed logistic model with strong Allee
effect are obtained. Also, we show that a Hopf Bifurcation will occur once some parameters pass through
critical values; that is, a family of periodic orbits bifurcates from the equilibrium. The direction of Hopf
bifurcation and the stability of the bifurcating periodic orbits are discussed by applying the normal form
theory and the center manifold theorem. Numerical simulations have shown that the analytical results are
correct.

References

[1] http://mathworld.wolfram.com/LogisticEquation.html
[2] J.D. Murray, Mathematical Biology I. An Introduction, Third Edition, Springer, 2002.
[3] M. M. Rao, K. L. Preetish K, Stability and Hopf bifurcation analysis of delay logistic equation,
arXiv:1211.7022v1 [q-bio.PE] 29 Nov 2012 .
[4] Y. kuang, Delay Differential Equations with Applications in population Dynamics, Academic
Press, Inc, 1998.
[5] H. Smith, An Introduction to Delay Differential Equations with Applications to the Life Sciences,
springer, 2010.
[6] K. Gopalsamy, M. R. S. Kulenovic, G. Ladas, Oscillations of a system of delay logistic equations,
Journal of Mathematical Analysis and Applications, 146(1990): 192–202.
[7] I. Kubiaczyk,S. H. Saker, Oscillation and stability in nonlinear delay differential equations of
population dynamics, Mathematical and Computer Modeling, 35(2002): 295–301.
[8] Y. Song, J. Wei, Bifurcation analysis for Chen’s system with delayed feedback and its application
to control of chaos, Chaos, Solitons and Fractals, 22(2004): 75–91.
[9] J. Zhang, Z. Jin , J. Yan, G. Sun , Stability and Hopf bifurcation in a delayed competition system,
Nonlinear Analysis, 70(2009): 658–670.

Copyright © 2014 by Modern Scientific Press Company, Florida, USA

Int. J. Modern Math. Sci. 2014, 11(3): 130-143 143

[10] O. Arino, Delay Differential Equations and Applications, Springer, Berlin, 2006.
[11] K. Gopalsamy, Stability and Oscillations in Delay Differential Equations of Population Dynamics
Kluwer Academic Press, London, 1992.
[12] S.H. Storgatz, Nonlinear Dynamics and Chaos, Perseus Books Publishing, 1994.
[13] M. Ferrara, L. ,Guerrini Center manifold reduction and perturbation method in a delayed model
with a mound-shaped Cobb-Douglas production function, Abstract and Applied Analysis, 2013;
Article ID 738460.
[14] A. M. Kramer, The evidence for Allee effects, Popul. Ecol., 51(2009): 341–354.
[15] E. M. Elabbasy , S. H. Saker ,H. EL-Metwally , Oscillation And Stability Of Nonlinear Discrete
Models Exhibiting The Allee Effect , Mathematica Versita Slovaca, 57(2007): 243–258.
[16] F. Courchamp, L. Berec, J. Gascoigne, Allee effects in ecology and Conservation, oxford University
Press 2008.
[17] M. A. Lewis , P. Kareiva , Allee dynamics and the spread of invading organisms, theor. Popul.
Boil., 43(1993): 141–158.
[18] Sebastian J. Schreiber, Allee effects, extinctions, and chaotic transients in simple population
models, theor. Popul. Boil. 64(2003): 201–209.
[19] B. D. Hassard , N.D. Kazarinoff , Y. H. Wan, Theory and Applications of Hopf Bifurcation,
Cambridge University Press, Cambridge, 1981.
[20] J. Hale, M. V.Lunel, Introduction to Functional Differential Equations, Springer, New York, 1993.

Copyright © 2014 by Modern Scientific Press Company, Florida, USA