Journal of Mathematical Analysis and Applications 246, 5879 2000.

doi:10.1006rjmaa.2000.6741, available online at http:rrwww.idealibrary.com on

Modelling the Interaction of Two Biological Species in

a Polluted Environment
B. Dubey and J. Hussain
Department of Mathematical Sciences, School of Science and Technology,
Tezpur Uni ersity, Tezpur-784 001, Sonitpur, Assam, India
Submitted by J. Eisenfeld
Received December 8, 1998

In this paper a mathematical model is proposed and analysed to study the effect
of an environmental pollutant on two interacting biological species. The interaction
between the two species is considered to be of three types, namely, competition,
cooperation, and preypredator. In each case criteria for local stability, instability,
and global stability of the nonnegative equilibria of the system are obtained. The
effect of diffusion on the equilibrium state of the system is also studied. 2000
Academic Press

1. INTRODUCTION
A large amount of pollutants and contaminants released from various
industries, motor vehicles, and other man-made projects enter into the
environment and affect human population and other biological species
seriously. In recent years some investigations have been carried out to
study the effect of pollution on a single-species population w2, 4, 610, 15x.
In particular, Hallam et al. w8x studied the effect of a toxicant present in
the environment on a single-species population by assuming that its growth
rate density decreases linearly with the concentration of toxicant but the
corresponding carrying capacity does not depend upon the concentration
of toxicant present in the environment. Considering this aspect, Freedman
and Shukla w6x studied the effect of toxicant on a single species and on a
predatorprey system by taking into account the introduction of toxicant
from an external source. Shukla and Dubey w15x studied the simultaneous
effect of two toxicants, one being more toxic than the other, on a biological
58
0022-247Xr00 $35.00
Copyright 2000 by Academic Press
All rights of reproduction in any form reserved.

SPECIES INTERACTION MODELLING

59

species. Dubey w4x proposed a model to study the depletion and conservation of forestry resources in a polluted environment.
We know that species do not exist alone in nature. They interact with
other species in their surroundings for their survival. So it is more
biologically significant to study two-species systems exposed to a pollutant.
In recent decades some investigations have been made to study the system
of two biological species w1, 12, 14, 16x in a polluted environment. In
particular, Ma Zhien and Hallam w14x studied two-dimensional nonautonomous LotkaVolterra models by the average method and obtained
sufficient conditions for persistence and extinction of the populations.
Chattopadhyay w1x studied the effect of toxic substances on a two-species
competitive system. He assumed that each of the competing species
produces a substance toxic to the other, but only when the other is present.
Huaping and Ma Zhien w12x investigated the effects of toxicants on
naturally stable two-species communities and derived persistenceextinction criteria for each population. But in modelling the system they assumed that the individuals of the two species have identical organismal
toxicant concentration, which need not always be true in nature. Recently,
Shukla and Dubey w16x studied the effects of population and pollution on
the depletion and conservation of forestry resources. It may be pointed out
here that the recycling effect of a toxicant and the effect of diffusion on
the stability of the equilibrium state of the system do not appear in the
above literature.
In view of the above, in this paper we propose a mathematical model for
studying the effect of environmental pollution on two interacting biological
species having different organismal pollutant concentrations. Three types
of interaction between the two species have been considered, namely,
competition, cooperation, and preypredator. The effect of diffusion on
the stability of the system is also studied. In the absence of diffusion our
model is more general than that of Huaping and Ma Zhien w12x. In the
presence of diffusion our results agree with those in Hastings w11x, Shukla
and Verma w19x, Shukla and Shukla w18x, Shukla et al. w17x, and Freedman
and Shukla w5x. The stability theory of the ordinary differential equation is
used to analyse the model w13x. In this paper we have also included
numerical examples to illustrate the applicability of the results obtained.

2. THE MODEL
Consider a polluted environment where two biological species are interacting with each other in a closed region D with smooth boundary D.
The variables of the model are x 1 s x 1 x, y, t . and x 2 s x 2 x, y, t ., the

60

DUBEY AND HUSSAIN

densities of the species 1 and 2 respectively; T s T x, y, t ., the concentration of pollutant present in the environment; U1 s U1 x, y, t . and U2 s
U2 x, y, t ., the concentration of pollutant in species 1 and in species 2
respectively at coordinates x, y . g D and time t G 0. In modelling the
system we assume that the carrying capacities of the species are constant.
Then following Huaping and Ma Zhien w12x and Dubey w4x, the
LotkaVolterra model of two species with pollutant effect and diffusion
can be written as

x1
t
x2
t
T
t
U1
t
U2
t

s r 10 x 1 y r 11 x 1U1 y a11 x 12 y a12 x 1 x 2 q D 1 2 x 1

s r 20 x 2 y r 21 x 2 U2 y a21 x 1 x 2 y a22 x 22 q D 2 2 x 2
s Q0 y 0 T q 1 1U1 q 2 2 U2 y 1 x 2 T y 2 x 2 T q D 3 2 T

2.1.

s y 1U1 q 0 0 T q 1 x 1T q 1 x 1
s y 2 U2 q 0X 0 T q 2 x 2 T q 2 x 2

0 F 0 q 0X F 1,

0 F 1 , 2 F 1.

We impose the following initial and boundary conditions on system 2.1.

x 1 x, y, 0 . s x, y . G 0,

x 2 x, y, 0 . s x, y . G 0,

T x, y, 0 . s x, y . G 0,

U1 x, y, 0 . s x, y . G 0,

U2 x, y, 0 . s x, y . G 0,

x1
n

x2
n

T
n

U1
n

U2
n

s 0,

x, y . g D

2.2.

x, y . g D, t ) 0,

where n is the unit outward normal to D.

In model 2.1., 2 s 2r x 2 q 2r y 2 is the Laplacian diffusion operator. Di i s 1, 2, 3. are the diffusion rate coefficients of x 1 x, y, t .,
x 2 x, y, t ., and T x, y, t . respectively in D. ri0 , ri1 , and a i j i, j s 1, 2. in
the first two equations of model 2.1. are constants. ri0 is the intrinsic
growth rate of the species i in the absence of pollutant, and ri1 the
depletion rate coefficient of species i due to organismal pollutant concen-

61

SPECIES INTERACTION MODELLING

tration. a12 and a21 are the interspecific interference coefficients and a11 ,
a22 are intraspecific interference coefficients of species 1 and 2 respectively. Q0 represents the rate of introduction of pollutant into the environment beyond the initial concentration, which is assumed to be positive or
zero. It is assumed that the pollutant in the environment is washed out or
broken down with rate 0 , and fractions 0 and 0X of it may again reenter
into species 1 and 2 respectively with the uptake of pollutant. 1 and 2
are the depletion rate coefficients of the pollutant in the environment due
to its intake by species 1 and 2, respectively. 1 and 2 are natural
depletion rate coefficients of U1 and U2 respectively due to ingestion and
depuration of pollutant, and fractions 1 and 2 of these may again
reenter the environment. 1 and 2 are the net uptake of pollutant from
resource by species 1 and 2 respectively.
It is assumed that the parameters 0 , 1 , and 2 are strictly positive and
1 , 2 , 1 , 2 , 1 , and 2 are nonnegative constants.
The following three cases will be dealt with:
i. Competition r 10 ) 0, r 20 ) 0, a12 ) 0, and a21 ) 0..
ii. Cooperation r 10 ) 0, r 20 ) 0, a12 - 0, and a21 - 0..
iii. Preypredator r 10 ) 0, r 20 - 0, a12 ) 0, and a21 - 0., assuming
x 1 as prey and x 2 as predator.

3. COMPETITION MODEL
We first analyse model 2.1. without diffusion i.e., D 1 s D 2 s D 3 s 0..
3.1. No Diffusion
In this case model 2.1. has four nonnegative equilibria, viz,

E0 0, 0,

Q0

0 1 y 01 y

0X 2

0 Q0

. 1 1 y 0 1 y 0X 2 .
0X Q0
2 1 y 0 1 y 0X 2 .

E
x 1 , 0, T, U1 , U2 ,

/ E 0, x , T, U , U / ,
E x , x , T , U , U / .
2

and

62

DUBEY AND HUSSAIN

The equilibrium E0 exists if

1 y 0 1 y 0X 2 ) 0.

3.1.

We shall show the existence of three other equilibria as follows:

U1 , U2 .. x 1 , T,
U1 , and U2 are the positive
Existence of E
x 1 , 0, T,
solutions of the following algebraic equations
a11 x 1 s r 10 y r 11U1 ,

0 T q 1 x 1T s Q0 q 1 1U1 q 2 2 U2 ,
1U1 s 0 0 T q 1 x 1T q 1 x 1 ,
2 U2 s 0X 0 T .
A little algebraic manipulation yields
a11 x 1 s r 10 y r 11 g x 1 . ,
Q0 q 1 1 x 1

Ts

0 1 y 0 1 y 0X 2 . q 1 1 y 1 . x 1

s f x1 . ,

say,
U1 s

 0 0 f x 1 . q 1 x 1 f x 1 . q 1 x 1 4 s g x 1 . ,

say,
U2 s

0X 0 f x 1 . s h x 1 . ,

say. Taking
F x 1 . s a11 x 1 y r 10 q r 11 g x 1 .
we note that F 0. - 0 if
r 11 0 Q0 - r 10 1 1 y 0 1 y 0X 2 .

3.2.

and F r 10 ra11 . ) 0, showing the existence of

x 1 in the interval 0 -
x1
r 10 ra11. For
x 1 to be unique the following condition must be satisfied at E;
Q0 1 1 y 1 . - 1 1 0 1 y 0 1 y 0X 2 . .

3.3.

Thus from the above analysis we note that the equilibrium E exists
under conditions 3.2. and 3.3..

63

SPECIES INTERACTION MODELLING

U1 , U2 .. As in the existence of E,
it can be seen
Existence of E0,
x 2 , T,
that the equilibrium E exists if the following inequalities hold
r 21 0X Q0 - r 20 2 1 y 0 1 y 0X 2 . ,
Q0 2 1 y 2 . - 2 2 0 1 y 0 1 y

0X 2

3.4.
..

3.5.

Existence of E x 1 , x 2 , T, U1 , U2 .. Here, x 1 , x 2 , T, U1 , and U2 are the

positive solutions of the system of algebraic equations
a11 x 1 q a12 x 2 q r 11 g x 1 , x 2 . s r 10 ,
a21 x 1 q a22 x 2 q r 21 h x 1 , x 2 . s r 20 ,
T s f x1 , x 2 . ,
U1 s g x 1 , x 2 . ,
U2 s h x 1 , x 2 . ,
where
f x1 , x 2 . s
g x1 , x 2 . s
h x1 , x 2 . s

Q0 q 1 1 x 1 q 2 2 x 2

0 1 y 0 1 y 0X 2 . q 1 1 y 1 . x 1 q 2 1 y 2 . x 2
1

1
1

 0 0 f x 1 , x 2 . q 1 x 1 f x 1 , x 2 . q 1 x 1 4 ,
 0X 0 f x 1 , x 2 . q 2 x 2 f x 1 , x 2 . q 2 x 2 4 .

It can be checked easily that E exists if in addition to conditions 3.2.

and 3.4., the conditions
yb q b 2 y 4 ac .

1r2

2a
yB q B 2 y 4 AC .
2A

1r2

ybX q bX 2 y 4 aX cX .

1r2

2 aX
yBX q BX 2 y 4 AX CX .

a12 q r 11 gr x 2 .
a11 q r 11 gr x 1 .
a22 q r 21 hr x 2 .
a21 q r 21 hr x 1 .

2 AX

3.6a.

1r2

3.6b.

) 0,

3.7a.

) 0,

3.7b.

64

DUBEY AND HUSSAIN

hold, where
a s 1 a11 1 1 y 1 . q r 11 1 4 ,
b s a11 0 1 1 y 0 1 y 0X 2 . q r 11 0 1 1 y 0X 2 . q r 11 1 Q0
y 1 1 r 10 1 y 1 . ,
c s r 11 0 0 Q0 y 0 1 r 10 1 y 0 1 y 0X 2 . ,
aX s 1 a21 2 1 y 1 . ,
bX s a21 0 2 1 y 0 1 y 0X 2 . q r 21 0X 0 1 1 y 1 2 r 20 1 y 1 . ,
cX s r 21 0X 0 Q0 y 0 2 r 20 1 y 0 1 y 0X 2 . ,
A s 2  a22 2 1 y 2 . q r 21 2 4 ,
B s a22 0 2 1 y 0 1 y 0X 2 . q r 21 0 2 1 y 0 1 . q r 21 2 Q0
y 2 2 r 20 1 y 2 . ,
C s r 21 0X 0 Q0 y 0 2 r 20 1 y 0 1 y 0X 2 . ,
AX s 2 a12 1 1 y 2 . ,
BX s a12 0 1 1 y 0 1 y 0X 2 . q r 11 0 0 2 2 y 2 1 r 10 1 y 2 . ,
CX s r 11 0 0 Q0 y 0 1 r 10 1 y 0 1 y 0X 2 . .
It may be noted here that E exists even when the inequalities 3.6a. and
3.6b. are reversed.
To study the local stability behavior of the equilibria, we first compute
the variational matrices corresponding to each equilibrium point. From
these matrices we conclude the following:
E0 is a saddle point with an unstable manifold locally in the x 1 y x 2
plane and a stable manifold locally in the T y U1 y U2 space. E and E are
locally unstable in the x 2 and x 1 directions respectively.
In the following theorem we have shown that E is locally asymptotically
stable.
THEOREM 3.1.

Let the inequalities

2
a12 q a21 . - 49 a11 a22

3.8a.

 cX1 1 1 q cX2 0 0 q 1 x 1 . 4 2 - 12 cX1 cX2 1 0 q 1 x 1 q 2 x 2 . 3.8b.

 cX1 2 2 q cX3 0X 0 q 2 x 2 . 4 2 - 12 cX1 cX3 2 0 q 1 x 1 q 2 x 2 . 3.8c.

65

SPECIES INTERACTION MODELLING

hold, where
cX1 s min
cX2 s
cX3 s

1 a11 0 q 1 x 1 q 2 x 2 1 a22 0 q 1 x 1 q 2 x 2
,
,
4 12
4 22
T2
T2

r 11

1T q 1
r 21

2 T q 2

,
.

Then E is locally asymptotically stable.

Proof. Consider the following positive definite function in the linearized form of system 2.1. with no diffusion,

x1 y x1 .

V x 1 , x 2 , T , U1 , U2 . s

2 x1
q

cX2
2

x2 y x2 .
2 x2

U1 y U1 .

cX3
2

cX1
2

T y T .

U2 y U2 .

It can be seen easily that the derivative of V with respect to t along the
solution of model 2.1. with no diffusion is negative definite under conditions 3.8., proving the theorem.
To show that E is globally asymptotically stable, we need the following
lemma, which establishes a region of attraction for system 2.1.. The proof
of this lemma is easy and hence is omitted.
LEMMA 3.1. The set

1 s x 1 , x 2 , T , U1 , U2 . : 0 F x 1 F

r 10
a11

, 0 F x2 F

r 20
a22

0 F T q U1 q U2 F L1
attracts all solutions initiating in the positi e orthant, where
L1 s

Q0 q 1

r 10
a11

q 2

r 20
a22

s min  0 1 y 0 y 0X . , 1 1 y 1 . , 2 1 y 2 . 4 .
In the following theorem the global stability of E is studied.

66

DUBEY AND HUSSAIN

Let the inequalities

THEOREM 3.2.

2
a12 q a21 . - 49 a11 a22

3.9a.

 c1 1 1 q c2 0 0 q 1 x 1 . 4 2 - 12 c1 c2 1 0 q 1 x 1 q 2 x 2 . 3.9b.
 c1 2 2 q c3 0X 0 q 2 x 2 . 4 2 - 12 c1 c3 2 0 q 1 x 1 q 2 x 2 . 3.9c.
hold, where
c1 s min
c2 s
c3 s

1 a11 0 q 1 x 1 q 2 x 2 1 a22 0 q 1 x 1 q 2 x 2
,
,
4 12
4 22
L21
L21

r 11

1 L1 q 1
r 21

2 L1 q 2

,
.

Then E is globally asymptotically stable with respect to all solutions initiating in

the interior of the positi e orthant.
Proof. Consider the positive definite function around E,
W x 1 , x 2 , T , U1 , U2 . s x 1 y x 1 y x 1 ln
q

c1
2

x1

x1

T y T . q

q x 2 y x 2 y x 2 ln

c2
2

U1 y U1 .

c3
2

x2

/
x2

U2 y U2 .

3.10.
Differentiating W with respect to t along the solutions of system 2.1.
without diffusion, we get
dW
dt

s x 1 y x 1 . w r 10 y r 11U1 y a11 x 1 y a12 x 2 x

q x 2 y x 2 . w r 20 y r 21U2 y a21 x 1 y a22 x 2 x
q c1 T y T . w Q0 y 0 T q 1 1U1 q 2 2 U2 y 1 x 1T y 2 x 2 T x
q c2 U1 y U1 . w y 1U1 q 0 0 T q 1 x 1T q 1 x 1 x
q c3 U2 y U2 . w y 2 U2 q 0X 0 T q 2 x 2 T q 2 x 2 x .

3.11.

67

SPECIES INTERACTION MODELLING

After some algebraic manipulations, Eq. 3.11. can be written as

dW
dt

sy
y
y
y
y
y
y

1
2
1
2
1
2
1
2
1
2
1
2
1
2

A11 x 1 y x 1 . q A12 x 1 y x 1 . x 2 y x 2 . y

2
1

A11 x 1 y x 1 . q A13 x 1 y x 1 . T y T . y

2
1

A 22 x 2 y x 2 . q A 23 x 2 y x 2 . T y T . y

A 22 x 2 y x 2 . q A 25 x 2 y x 2 . U2 y U2 . y
2

A 33 T y T . q A 35 T y T . U2 y U2 . y

A 33 T y T .

A11 x 1 y x 1 . q A14 x 1 y x 1 . U1 y U1 . y

A 33 T y T . q A 34 T y T . U1 y U1 . y

A 22 x 2 y x 2 .

1
2
1
2

A 44 U1 y U1 .

A 33 T y T .
1
2

A 55 U2 y U2 .

A 44 U1 y U1 .

A 55 U2 y U2 . ,

where
A11 s 23 a11 ,
A 44 s c 2 1 ,

A 22 s 23 a22 ,

A 55 s c 3 2 ,

A 33 s 12 c1 0 q 1 x 1 q 2 x 2 . ,

A12 s y a12 q a21 . ,

A14 s c2 1T q 1 . y r 11 ,
A 25 s c3 2 T q 2 . y r 21 ,

A13 s yc1 1T ,

A 23 s yc1 2 T ,

A 34 s c1 1 1 q c2 0 0 q 1 x 1 . ,

A 35 s c1 2 2 q c 3 0X 0 q 2 x 2 . .
Sufficient conditions for dWrdt to be negative definite are that the
following inequalities hold
A212 - A11 A 22 ,

3.12a.

A213 - A11 A 33 ,

3.12b.

A214 - A11 A 44 ,

3.12c.

A223 - A 22 A 33 ,

3.12d.

- A 22 A 55 ,

3.12e.

A234 - A 33 A 44 ,

3.12f.

A235 - A 33 A 55 .

3.12g.

A225

68

DUBEY AND HUSSAIN

Under the suitable choice of constants c1 , c 2 , c 3 as in Theorem 3.2, we

note that inequalities 3.12b. 3.12e. are automatically satisfied and 3.9a.
3.12a., 3.9b. 3.12f., and 3.9c. 3.12g.. Thus W is a Liapunov
function with respect to E, whose domain contains the region 1 , proving
the theorem.
Remark 3.1. In the case of instantaneous introduction of a pollutant
i.e., Q0 s 0. into the environment, it can be verified that there are four
U1 , U2 ., E0, x 2 ,
nonnegative equilibria, namely E0 0, 0, 0, 0, 0., E
x 1 , 0, T,
U1 , U2 ., and E x 1 , x 2 , T, U1 , U2 .. E0 obviously exists and the existence of
T,
the remaining three equilibria can be seen in a fashion similar to that
discussed earlier. Further, the stability behavior of the equilibria is similar
to the corresponding equilibria as given in the case of the constant
introduction of pollutant into the environment. It has been noted here that
equilibrium levels of the competing species in the constant introduction of
pollutant into the environment is lower than the case of instantaneous
introduction, keeping other parameters and functions the same in the
model.
3.2. Model with Diffusion
In this section we consider the complete model 2.1. 2.2. and we state
the main results of this section in the form of the following theorem.
THEOREM 3.3. i. If the equilibrium E of the system without diffusion is
globally asymptotically stable, then the corresponding uniform steady state of
the initial-boundary alue problems 2.1. 2.2. is also globally asymptotically
stable.
ii. If the equilibrium E of the system without diffusion is unstable, e en
the uniform steady state of the initial-boundary alue problems 2.1. 2.2. can
be made stable by increasing diffusion coefficients appropriately.
Proof. Let us consider the positive definite function

U x 1 t . , x 2 t . , T t . , U1 t . , U2 t . . s

HHD W x , x
1

2 , T , U1 , U2

. dA,
3.13.

where W is defined by Eq. 3.10..

69

SPECIES INTERACTION MODELLING

We have
dU

dt

W x1

HHD

x1 t

q
s

W x2
x2 t

W U1

U1 t

HHD W dA q HHD

D1

W T

T t

W U2
U2 t

2 x1 q D2

x1

dA

x2

qD 3

2 x2

W
T

2 T dA

s I1 q I2 ,

3.14.

where
I1 s
I2 s

D1

HHD

x1

HHD W dA,

2 x1 q D2

and
2 x 2 q D3

x2

W
T

2 T dA.

We note the following properties of W, namely,

W
x1

s
D

W
x2

s
D

W
T

U1

W
U

s 0,

and for all points of D,

2W
x1 x 2

s
s

2W
x1 T

2W
x 2 U2

2W
x 1 U1

2W
T U1

s
s

2W
x 1 U2
2W
T U2

s
s

2W
x2 T
2W
U1 U2

2W
x 2 U1

s0

and

2W
x 12

) 0,

2W
x 22

) 0,

2W
T 2

) 0,

2W
U12

) 0,

2W
U22

) 0.

We now consider I2 and determine the sign of each term. We utilize the
formula known as Greens first identity in the plane

HHD F

G dA s

H D F n

ds y

HHD F.G . dA,

70

DUBEY AND HUSSAIN

where Gr n is the directional derivative in the direction of the unit

outward normal to D and s is the arc length.
Then with F s Wr x 1 and G s x 1 , we get

HHD x

2 x 1 dA s

W x1

H D x

sy

HHD

ds y

/
x1

HHD

/
x1

.x 1 dA

.x 1 dA,

since x 1r n s 0.
Now,

/
x1

2 W x1

x 12 x

i q

2 W x1
x 12 y

j.

Hence,

HHD x

2 x 1 dA s y

HHD

2W

x1

x 12

x1

/ / /
q

dA F 0. 3.15a.

Similarly,

HHD x

2 x 2 dA F 0

and

HHD

W
T

2 T dA F 0. 3.15b.

i.e., I2 F 0.
Thus we note that if I1 F 0, i.e., if E is globally asymptotically stable in
the absence of diffusion, then the uniform steady state of the initialboundary value problems 2.1. 2.2. also must be globally asymptotically
stable. This proves the first part of the theorem.
) 0, i.e., I1 ) 0, then E will be unstable in
We further note that if W
the absence of diffusion. However, Eqs. 3.14. and 3.15. show that by
making the diffusion coefficients Di sufficiently large, U can be made
negative even if I1 ) 0. This proves the second part of the theorem.
Now we shall prove Theorem 3.3 for a rectangular region. Let us
consider D to be a rectangular region given by
D s  x, y . : 0 F x F a, 0 F y F b 4
In this case I2 can be written as
I2 s yD 1

HHD

2W

x1

x 12

x1

/ / /
q

dA q negative terms.

71

SPECIES INTERACTION MODELLING

However,

2W

x1

x 12

x 12

So,
I2 s yD 1

F yD 1

x1

x1

x1

dA q negative terms

/ HH / /
/ HH / /
x 12

x1

2
a11
x1
2
r 10

x1

dA.

Now,

HHD

x1

x1 y x1 .

dA s

/
x

HHD

x
a

H0 H0

dA

x1 y x1 .
x

dx dy.

Let z s xra, then

HHD

x1

dA s

/
x

1
a

H0 H0

x1 y x1 .
z

dz dy.

Now, utilizing the well-known inequality w3x

1

H0

x1

dx G 2

H0

x 12 dx,

we get

HHD

x1

/
x

dA G
s
s

2
a

2
a

H0 H0 x
b

y x 1 . dz dy

y x 1 . dx dy

H0 H0 x
HHD x

y x 1 . dA.

72

DUBEY AND HUSSAIN

Similarly,

HHD

x1

dA G

/
y

b2

HHD x

y x 1 . dA,

hence,
I2 F yD 1
sy

2
a11
x1
2
r 10

HHD

1
a

2
D 1 a11
x 1 2 a2 q b 2 .
2 2 2
r 10
a b

1
a2

x 1 y x 1 . dA

HHD x

y x 1 . dA,

which shows that I2 F 0.

Thus for a given rectangular region, by increasing diffusion coefficients
sufficiently large, an unstable steady state in the absence of diffusion can
be made stable. This theorem implies that in the presence of diffusion the
competing species converge towards their respective carrying capacities
faster than in the case of no diffusion.

4. COOPERATION MODEL
In this case we have r 10 ) 0, r 20 ) 0, a12 - 0, and a21 - 0. There exist
four nonnegative equilibria, namely,

E0 0, 0,

Q0

0X 2

0 1 y 01 y

0 Q0

. 1 1 y 0 1 y 0X 2 .
0X Q0
2 1 y 0 1 y 0X 2 .

Ec
x 1 c , 0, Tc , U1 c , U2 c ,

E 0,
x
c

2 c , Tc , U1 c , U2 c

/
/,

and
Ec x 1 c , x 2 c , Tc , U1 c , U2 c .

E0 exists if 1 y 0 1 y 0X 2 ) 0. Existence of Ec , Ec , and Ec can be

checked as was done in the competition model in Section 3.

73

SPECIES INTERACTION MODELLING

The local stability behaviors of E0 , Ec , and Ec are similar to the

corresponding equilibria of Section 3.
The following theorem shows the local stability character of Ec , the
proof of which is similar to Theorem 3.1 and hence is omitted.
Let the inequalities

THEOREM 4.1.

2
a12 q a21 . - 49 a11 a22

4.1a.

 kX1 1 1 q kX2 0 0 q 1 x 1 c . 4 2 - 12 kX1 kX2 1 0 q 1 x 1 c q 2 x 2 c . 4.1b.

 kX1 2 2 q kX3 0X 0 q 2 x 2 c . 4 2 - 12 kX1 kX3 2 0 q 1 x 1 c q 2 x 2 c . 4.1c.
hold, where
1 a11 0 q 1 x 1 c q 2 x 2 c 1 a22 0 q 1 x 1 c q 2 x 2 c
,
,
4 12
4 22
Tc2
Tc2

kX1 s min

kX2 s

r 11

kX3 s

1Tc q 1
r 21

2 Tc q 2

,
,

then Ec is locally asymptotically stable.

In order to show the global stability of Ec , we need the following lemma
whose proof is easy and hence is omitted.
LEMMA 4.1. The set
2 s  x 1 , x 2 , T , U1 , U2 . : 0 F x 1 F x 1 - , 0 F x 2 F x 2 - ,
0 F T q U1 q U2 F L2 4
attracts all the solutions initiating in the interior of the positi e orthant, where
L2 s

Q0 q 1 x 1 q 2 x 2 . ,

s min  0 1 y 0 y 0X . , 1 1 y 1 . , 2 1 y 2 . 4 .
The following theorem shows the global stability of Ec whose proof is
similar to that of Theorem 3.2 and hence is omitted.

74

DUBEY AND HUSSAIN

THEOREM 4.2.

Let the inequalities

2
a12 q a21 . - 49 a11 a22

4.2a.

 k 1 1 1 q k 2 0 0 q 1 x 1 c . 4 2 - 12 k 1 k 2 1 0 q 1 x 1 c q 2 x 2 c . 4.2b.
 k 1 2 2 q k 3 0X 0 q 2 x 2 c . 4 2 - 12 k 1 k 3 2 0 q 1 x 1 c q 2 x 2 c . 4.2c.
hold, where

k 1 s min
k2 s
k3 s

1 a11 0 q 1 x 1 c q 2 x 2 c 1 a22 0 q 1 x 1 c q 2 x 2 c
,
,
4 12
4 22
L22
L22

r 11

1 L2 q 1
r 21

2 L2 q 2

then Ec is globally asymptotically stable with respect to all solutions initiating

in the interior of the positi e orthant.
It may be noted here that conditions in Theorem 4.1 are similar to
Theorem 3.1, and conditions in Theorem 4.2 are similar to Theorem 3.2
where the equilibrium E has been replaced by Ec .
Remark 4.1. The effect of diffusion in the case of the cooperation
model can be studied in a way similar to that of the competition model
given in Section 3. It may be noted here that the results of Theorem 3.3
are also valid in the case of cooperation.

5. PREYPREDATOR MODEL
We consider x 1 and x 2 to be prey and predator respectively. Then in
this case we have
r 10 ) 0, r 20 - 0, a12 ) 0, and a21 - 0.
X
X
We take a21 s yb 21 and r 20 s yr 20
, where b 21 ) 0, r 20
) 0.

75

SPECIES INTERACTION MODELLING

In this case there exist three nonnegative equilibria, namely,

Q0

E0 0, 0,

0 1 y 01 y

0X 2

0 Q0

. 1 1 y 0 1 y 0X 2 .
0X Q0
2 1 y 0 1 y 0X 2 .

Ep
x 1 p , 0, Tp , U1 p , U2 p ,

and
Ep x 1 p , x 2 p , Tp , U1 p , U2 p .

E0 exists if 1 y 0 1 y 0X 2 ) 0. The existence of Ep and Ep can be

established similarly to those of the competition model.
The local stability of E0 and Ep can be studied in a way similar to that
used in Section 3 for the competition model.
The following theorem shows that Ep is locally asymptotically stable.
The proof of this theorem is similar to Theorem 3.1 and hence is omitted.
Let the inequalities

THEOREM 5.1.

k q k q x . 5
k q k q x . 5
1 1 1

1 2

X
0

1 1p

2p

- 12 k 1 k 2 1 0 q 1 x 1 p q 2 x 2 p . 5.1a .
- 12 k 1 k 3 2 0 q 1 x 1 p q 2 x 2 p . 5.1b .

hold, where
k 1 s min
k2 s
k3 s

1 a11 0 q 1 x 1 p q 2 x 2 p 1 a22 a12 0 q 1 x 1 p q 2 x 2 p

,
,
4 12
4 22 b 21
Tp2
Tp2

r 11

1Tp q 1
a12

r 21

b 21 2 Tp q 2

then Ep is locally asymptotically stable.

In order to show the global stability of Ep , we need the following lemma
whose proof is easy and hence is omitted.

76

DUBEY AND HUSSAIN

LEMMA 5.1. The set

3 s x 1 , x 2 , T , U1 , U2 . : 0 F x 1 F

r 10
a11

, 0 F x2 F

r 10 b 21
a11 a22

0 F T q U1 q U2 F L3

attracts all solutions initiating in the interior of the positi e orthant, where
L3 s

1 r 10

a11

1 q

b 21
a22

2 ,

s min  0 1 y 0 y 0X . , 1 1 y 1 . , 2 1 y 2 . 4 .
In the following theorem we are able to write down conditions for Ep to
be globally asymptotically stable. The proof of this theorem is similar to
that of Theorem 3.2 and hence is omitted.
Let the inequalities

THEOREM 5.2.

k q k q x . 5
k q k q x . 5
1 1 1

1 2

1 1p

X
0

2p

- 12
k1
k 2 1 0 q 1 x 1 p q 2 x 2 p . 5.2a .
- 12
k1
k 3 2 0 q 1 x 1 p q 2 x 2 p . 5.2b .

hold, where

k 1 s min
k 2 s
k 3 s

1 a11 0 q 1 x 1 p q 2 x 2 p 1 a22 a12 0 q 1 x 1 p q 2 x 2 p

,
,
4 12
4 22 b 21
L23
L23

r 11

1 L3 q 1

a12 r 21
b 21 2 L3 q 2 .

then Ep is globally asymptotically stable with respect to all solutions initiating

in the interior of the positi e orthant.
Remark 5.1. The effect of diffusion in the case of the preypredator
model is found to be similar to the competition model given in Section 3.
In particular, the results of Theorem 3.3 remain valid in this case.

77

SPECIES INTERACTION MODELLING

6. NUMERICAL EXAMPLES
In this section we present numerical examples to explain the applicability of the results discussed above. We choose the following values of the
parameters in model 2.1. without diffusion
r 11 s 0.05,

r 21 s 0.04,

Q0 s 15.0

0 s 6.7,

a11 s 0.22,

1 s 15.5,

a22 s 0.26,

2 s 10.4,

1 s 0.02, 2 s 0.03, 0 s 0.01, 0X s 0.04,

1 s 0.06,

2 s 0.09,

1 s 0.25, and

6.1.

2 s 0.3.

EXAMPLE 1. In this example we consider the case when the two species
compete with each other. In addition to the values of the parameters given
in Eq. 6.1., we choose the following parameters in model 2.1. without
diffusion:
r 10 s 5.0, r 20 s 3.0 a12 s 0.07, and a21 s 0.08.
With the above values of the parameters, it can be checked that the
interior equilibrium E exists and is given by
x 1 s 21.01420, x 2 s 5.03089, T s 1.81106, U1 s 0.49409, U2 s 0.27064.
It can also be checked that conditions 3.8. in Theorem 3.1 are satisfied
which shows that E is locally asymptotically stable.
Further, we note that the conditions 3.9. in Theorem 3.2 are also
satisfied which shows that E is globally asymptotically stable.
EXAMPLE 2. Here we consider the case when the two species cooperate
with each other. In addition to the values of the parameters given in Eq.
6.1., we choose the following values of the parameters in model 2.1.
without diffusion:
r 10 s 5.0, r 20 s 3.0, a12 s y0.07, and a21 s y0.08.
With the above values of the parameters, it can be verified that the
interior equilibrium Ec exists and is given by
x 1 c s 29.05284, x 2 c s 20.34072, Tc s 1.50652,
U1 c s 0.64453, U2 c s 0.89076.
It can also be verified that conditions 4.1. in Theorem 4.1 are satisfied,
showing the local stability character of Ec .

78

DUBEY AND HUSSAIN

Further, it is easy to verify that conditions 4.2. in Theorem 4.2 are

satisfied, showing the global stability character of Ec .
EXAMPLE 3. In this example we consider the case when x 2 preys on x 1.
In addition to the values of the parameters given in Eq. 6.1., we choose
the following values of the parameters in model 2.1. without diffusion:
r 10 s 5.0, r 20 s y0.5, a12 s 0.2, and a21 s y0.1.
With the above values of the parameters, it can be verified that the
interior equilibrium Ep exists, and is given by
x 1 p s 18.09059, x 2 p s 4.99304, Tp s 1.84798, U1 p s 0.42918, U2 p s 0.27150.
It can also be verified that the conditions 5.1. in Theorem 5.1 are
satisfied. This shows that Ep is locally asymptotically stable.
Further, it can also be checked that the conditions 5.2. in Theorem 5.2
are satisfied. This shows that Ep is globally asymptotically stable.
7. SUMMARY
In this paper we have proposed and analysed a mathematical model for
studying the survival of two interacting species in a polluted environment,
the modes of interaction being competition, cooperation, and predation.
The model has been analysed with and without diffusion. When there is no
diffusion it has been shown that in the case of a constant introduction of
pollutant into the environment the competing species settle down to their
respective equilibrium levels, the magnitude of which depends upon the
equilibrium levels of washout and uptake rates of the pollutant. It has also
been noted that if the concentration of pollutant increases unabatedly,
then the survival of the species would be threatened. In the case of an
instantaneous introduction of the pollutant into the environment, it has
been found that the competing species again settle down to their respective equilibrium levels whose magnitude is higher than in the case of a
constant introduction of pollutant into the environment.
The effect of diffusion on the interior equilibrium state of the system
has also been investigated. It has been shown that if the positive equilibrium of the system without diffusion is globally asymptotically stable, then
the corresponding uniform steady state of the system with diffusion is also
globally asymptotically stable. It has further been noted that if the positive
equilibrium of the system with no diffusion is unstable, then the corresponding uniform steady state of the system with diffusion can be made
stable by increasing the diffusion coefficients appropriately. From the

SPECIES INTERACTION MODELLING

79

proof of Theorem 3.3, it should be noted that U contains some extra

negative terms which imply that the global stability is more feasible in the
case of diffusion than in the case of no diffusion. In the cases of
cooperation and preypredator, similar results have been found. In each
case, a numerical example has been given to illustrate the results obtained.
REFERENCES
1. J. Chattopadhyay, Effect of toxic substances on a two-species competitive system, Ecol.
Model. 84 1996., 287289.
2. J. T. De Luna and T. G. Hallam, Effects of toxicants on populations: a qualitative
approach. IV. Resourceconsumertoxicant models, Ecol. Model. 35 1987., 249273.
3. M. M. Denn, Stability of Reaction and Transport Processes, PrenticeHall, Englewood
Cliffs, NJ, 1975.
4. B. Dubey, Modelling the effect of toxicant on forestry resources, Indian J. Pure Appl.
Math. 28 1997., 112.
5. H. I. Freedman and J. B. Shukla, The effect of predator resource on a diffusive predator
prey system, Natur. Resource Model. 3, No. 3 1989., 359383.
6. H. I. Freedman and J. B. Shukla, Models for the effect of toxicant in single-species and
predatorprey systems, J. Math. Biol. 30 1991., 1530.
7. T. G. Hallam, C. E. Clark, and G. S. Jordan, Effects of toxicants on populations: a
qualitative approach. II. First order kinetics, J. Math. Biol. 18 1983., 2537.
8. T. G. Hallam, C. E. Clark, and R. R. Lassiter, Effects of toxicants on populations: A
qualitative approach. I. Equilibrium environmental exposure, Ecol. Model. 18 1983.,
291304.
9. T. G. Hallam and J. T. De Luna, Effects of toxicants on populations: a qualitative
approach. III. Environmental and food chain pathways, J. Theoret. Biol. 109 1984.,
411429.
10. T. G. Hallam and Ma Zhien, Persistence in population models with demographic
fluctuations, J. Math. Biol. 24 1986., 327339.
11. A. Hastings, Global stability in LotkaVolterra systems with diffusion, J. Math. Biol. 6
1978., 163168.
12. L. Huaping and Ma Zhien, The threshold of survival for system of two species in a
polluted environment, J. Math. Biol. 30 1991., 4961.
13. J. La Salle and S. Lefschetz, Stability by Liapunovs Direct Method with Applications,
Academic Press, New YorkrLondon, 1961.
14. Ma Zhien and T. G. Hallam, Effects of parameter fluctuations on community survival,
Math. Biosci. 86 1987., 3549.
15. J. B. Shukla and B. Dubey, Simultaneous effect of two toxicants on biological species: a
mathematical model, J. Biol. Systems 4, No. 1 1996., 109130.
16. J. B. Shukla and B. Dubey, Modelling the depletion and conservation of forestry
resources: effects of population and pollution, J. Math. Biol. 36 1997., 7194.
17. J. B. Shukla, H. I. Freedman, V. N. Pal, O. P. Misra, M. Agarwal, and A. Shukla,
Degradation and subsequent regeneration of a forestry resource: A mathematical model,
Ecol. Model. 44 1989., 219229.
18. V. P. Shukla and J. B. Shukla, Multispecies food webs with diffusion, J. Math. Biol. 13
1982., 339344.
19. J. B. Shukla and S. Verma, Effects of convective and dispersive interactions on the
stability of two species, Bull. Math. Biol. 43, No. 5 1981., 593610.