You are on page 1of 57

Computational Ecology and Software

Vol. 4, No. 2, 1 June 2014

Computational Ecology and Software Vol. 4, No. 2, 1 June 2014 International Academy of Ecol ogy
Computational Ecology and Software Vol. 4, No. 2, 1 June 2014 International Academy of Ecol ogy

International Academy of Ecology and Environmental Sciences

Computational Ecology and Software

ISSN 2220-721X Volume 4, Number 2, 1 June 2014

Editor-in-Chief

WenJun Zhang Sun Yat-sen University, China International Academy of Ecology and Environmental Sciences, Hong Kong E-mail: zhwj@mail.sysu.edu.cn, wjzhang@iaees.org

Editorial Board

Ronaldo Angelini (The Federal University of Rio Grande do Norte, Brazil) Andre Bianconi (Sao Paulo State University (Unesp), Brazil) Bin Chen (Beijing Normal University, China) Daniela Cianelli (University of Naples Parthenope, Italy) Alessandro Ferrarini (University of Parma, Italy) Yanbo Huang (USDA-ARS Crop Production Systems Research Unit, USA) Istvan Karsai (East Tennessee State University, USA) Vladimir Krivtsov (Heriot-Watt University, UK) Lev V. Nedorezov (University of Nova Gorica, Slovenia) Fivos Papadimitriou (Environmental and Land Use Consultancies, Greece) George P. Petropoulos (Institute of Applied and Computational Mathematics, Greece) Vikas Rai (Jazan University, Saudi Arabia) Santanu Ray (Visva Bharati University, India) Kalle Remm (University of Tartu, Estonia) Rick Stafford (University of Bedfordshire, UK) Luciano Telesca (Institute of Methodologies for Environmental Analysis, Italy) Bulent Tutmez (Inonu University, Turkey) Ranjit Kumar Upadhyay (Indian School of Mines, India) Ezio Venturino (Universita’ di Torino, Italy) Michael John Watts (The University of Adelaide, Australia) Peter A. Whigham (University of Otago, New Zealand) ZhiGuo Zhang (Sun Yat-sen University, China)

Editorial Office: ces@iaees.org

Publisher: International Academy of Ecology and Environmental Sciences Address: Flat C, 23/F, Lucky Plaza, 315-321 Lockhart Road, Wanchai, Hong Kong Tel: 00852-6555 7188 Fax: 00852-3177 9906 Website: http://www.iaees.org/ E-mail: office@iaees.org

Computational Ecology and Software, 2014, 4(2): 82-88

Article

Spatial risk assessment of alien plants in China using biodiversity resistance theory

YouHua Chen

Department of Renewable Resources, University of Alberta, Edmonton, T6G 2H1, Canada E-mail: haydi@126.com

Received 24 February 2014; Accepted 27 March 2014; Published online 1 June 2014

; Accepted 27 March 2014 ; Published online 1 June 2014 Abstract In the present study,

Abstract In the present study, the potential occurrence risk of invasive plants across different provinces of China is studied using disease risk mapping techniques (empirical Bayesian smoothing and Poisson-Gamma model). The biodiversity resistance theory which predicts that high-biodiversity areas will have reduced risk of species invasion serves as the base for performing spatial risk assessment of plant invasion across provinces. The results show that, both risk mapping methods identified that north-eastern part of China have the highest relative risk of plant invasion. In contrast, south-western and south-eastern parts of China, which have high woody plant richness, are predicted to possess low relative risks of plant invasion. Through spatial regression analysis (simultaneous autoregression model), nine environmental variables representing energy availability, water availability, seasonality, and habitat heterogeneity are used to explain the relative risk of plant invasion across provinces of China. The fitting results suggest that, PRECrange and TEMrange are the most two important covariates correlated with the occurrence risks of alien plants at provincial level in China. As indicated by Moran’s I index, spatial regression analysis can effectively eliminate the potential biases caused by spatial autocorrelation.

Keywords spatial ecology; risk assessment; invasion biology; plant distribution.

Computational Ecology and Software ISSN 2220 ­721X URL: http://www.iaees.org/publications/journals/ces/online­ version.asp RSS: http://www.iaees.org/publications/journals/ces/rss.xml E­mail: ces@iaees.org Editor ­in­Chief: WenJun Zhang Publisher: International Academy of Ecolo g y and Environmental Sciences

1 Introduction Alien plants may cause severe ecological problems onto local ecosystems because they can better compete for the resources than native species (Zhang and Chen, 2011). Biodiversity hotspots are believed to be more resistant to species invasion: more diverse communities would have less species invasions (Law and Morton, 1996; Stachowicz et al., 2002; Zhang, 2011, 2012a). This biodiversity resistance theory can be an option for performing risk assessment of species invasions. Modeling plant invasion risk in China has been done in some previous studies recently (Bai et al., 2013; Liu et al., 2005; Wu et al., 2006; Feng and Zhu, 2010). However, all these studies only rely on the richness

IAEES

www.iaees.org

Computational Ecology and Software, 2014, 4(2): 82-88

83

information of alien plants, the dependence of alien plant diversity and native plant diversity has never been considered yet. As such, one may utilize the biodiversity resistance theory to quantify relative risk of alien plants by explicitly incorporating the information of native plant diversity. In the present study, we perform the relative risk assessment of alien plants in China by using spatial disease mapping methods (including empirical Bayesian smoothing and Poisson-Gamma model) so as to incorporate richness information of native woody plant species.

2 Materials and Methods

2.1 Invasive plant diversity data

The province-level 127 alien plant richness data are collected from previous studies (Chen, 2013; Wang et al., 2009; Wu et al., 2006; Yan et al., 2012). The resultant data include provincial distributional information and physiological trait patterns of each invasive plant found in China.

2.2 Woody plant diversity data

The population size at risk when performing risk mapping is required as an entry. In the present study, we regard the population size at risk as the number of native woody plant species. As such, similar to the disease transmission model, areas where there are higher population sizes would have less relative risk of disease outbreak. In our modeling of relative risk estimation of plant invasion, we hold the recognition that areas with high biological diversity should have less plant invasion risk (Law and Morton, 1996; Stachowicz et al., 2002).

2.3 Risk mapping using empirical Bayesian smoothing and Poisson-Gamma model

We use both empirical Bayesian smoothing and Poisson-Gamma model for predicting the outbreak risk of invasive plants over the provinces of China. For testing whether the relative risk of plant invasion is significantly higher than 1, we utilize Poisson exact test.

2.4 Identification of important covariates associated with the occurrence risk of alien plants

We use generalized linear models to identify important covariates which are closely related to the occurrence

risk of alien plants at provincial level in China. The fitting model is given by,

log() X

where is the relative risk of alien plants across various provinces of China calculated as above; is the

associated coefficients for the covariates presented in terms of columns in the covariate matrix X.

We standardize all the covariates in X so that the estimated coefficients are standardized partial

coefficients. We then choose those covariates with high values of as the important ones for explaining the

occurrence risk of alien plants at provincial level in China. However, the above model is spatially implicit, for which it can not be able to partial out the influence of spatial autocorrelation and may inflate Type I errors. As such, we implemented the alternative model which explicitly incorporate the spatial influence in the model as,

log() X W log()

where W is the matrix with each element

boundaries), otherwise

(log-transformed relative risk). This is the formula of spatial simultaneous autoregression lag model (SAR)

(Dormann et al., 2007; Dormann, 2007). Implication of SAR model has been done using R (R Development

Core Team, 2011) package “spdep” (Bivand et al., 2013). The following environmental variables at provincial level are used for fitting the above model (Qian, 2013):

mean annual temperature (TEM, °C), mean temperature of the coldest month (TEMmin, °C), mean

w ij

=1 if provinces i and j are adjacent (i.e., they share some

w =0. measures the strength of spatial autocorrelation for the response variable

ij

IAEES

www.iaees.org

84

Computational Ecology and Software, 2014, 4(2): 82-88

50

40

30

20

10

0

50

40

30

20

10

0

temperature of the warmest month (TEMmax, °C), annual precipitation (PREC, mm), annual actual evapotranspiration (AET, mm), annual potential evapotranspiration (PET, mm), the range of mean annual temperature within a province (TEMrange, °C), the range of annual precipitation within a province (PRECrange, mm), and the range of elevation within a province (ELEVrange, m). These nine environmental variables represent energy availability, water availability, seasonality, and habitat heterogeneity (Qian, 2013). All these covariates are standardized before performing SAR modeling.

3 Results As showed in Fig. 1, the woody plant diversity is highest in southern part of China, which was identified as one of the global biodiversity hotspots of the world (Myers et al., 2000; Chen and Bi, 2007). Yunnan has the highest woody plant richness, followed by Guangxi Province.

highest woody plant richness, followed by Guangxi Province. 80 90 100 110 120 130 Fig. 1
highest woody plant richness, followed by Guangxi Province. 80 90 100 110 120 130 Fig. 1
highest woody plant richness, followed by Guangxi Province. 80 90 100 110 120 130 Fig. 1
highest woody plant richness, followed by Guangxi Province. 80 90 100 110 120 130 Fig. 1
highest woody plant richness, followed by Guangxi Province. 80 90 100 110 120 130 Fig. 1
highest woody plant richness, followed by Guangxi Province. 80 90 100 110 120 130 Fig. 1
highest woody plant richness, followed by Guangxi Province. 80 90 100 110 120 130 Fig. 1
highest woody plant richness, followed by Guangxi Province. 80 90 100 110 120 130 Fig. 1
highest woody plant richness, followed by Guangxi Province. 80 90 100 110 120 130 Fig. 1
highest woody plant richness, followed by Guangxi Province. 80 90 100 110 120 130 Fig. 1
highest woody plant richness, followed by Guangxi Province. 80 90 100 110 120 130 Fig. 1

80

90

100

110

120

130

Fig. 1 Observed woody plant diversity at provincial level of China. Colors from light to dark grey indicate diversity from low to high.

light to dark grey indicate diversity from low to high. 80 90 100 110 120 130
light to dark grey indicate diversity from low to high. 80 90 100 110 120 130
light to dark grey indicate diversity from low to high. 80 90 100 110 120 130
light to dark grey indicate diversity from low to high. 80 90 100 110 120 130
light to dark grey indicate diversity from low to high. 80 90 100 110 120 130
light to dark grey indicate diversity from low to high. 80 90 100 110 120 130
light to dark grey indicate diversity from low to high. 80 90 100 110 120 130
light to dark grey indicate diversity from low to high. 80 90 100 110 120 130
light to dark grey indicate diversity from low to high. 80 90 100 110 120 130
light to dark grey indicate diversity from low to high. 80 90 100 110 120 130
light to dark grey indicate diversity from low to high. 80 90 100 110 120 130

80

90

100

110

120

130

Fig. 2 Observed alien plant diversity at provincial level of China. Colors from light to dark grey indicate diversity from low to high.

IAEES

www.iaees.org

Computational Ecology and Software, 2014, 4(2): 82-88

85

50

40

30

20

10

50

40

30

20

10

As shown in Fig. 2, the alien plant diversity is highest in southern and eastern part of China. Interestingly, the biodiversity hotspot, Yunnan Province, also has very high alien plant diversity following Jiangsu Province (Wang et al., 2009).

diversity following Jiangsu Province (Wang et al., 2009). 80 90 100 110 120 130 Fig. 3
diversity following Jiangsu Province (Wang et al., 2009). 80 90 100 110 120 130 Fig. 3
diversity following Jiangsu Province (Wang et al., 2009). 80 90 100 110 120 130 Fig. 3
diversity following Jiangsu Province (Wang et al., 2009). 80 90 100 110 120 130 Fig. 3
diversity following Jiangsu Province (Wang et al., 2009). 80 90 100 110 120 130 Fig. 3
diversity following Jiangsu Province (Wang et al., 2009). 80 90 100 110 120 130 Fig. 3

80

90

100

110

120

130

Fig. 3 Relative risk of plant invasiveness at provincial level of China using empirical Bayesian smoothing method. Provinces which have red boundaries indicate that their relative risks of plant invasion are significantly higher than 1. Colors from light to dark grey indicate invasive risks from low to high.

ht to dark grey indicate invasive risks from low to high. 80 90 100 110 120
ht to dark grey indicate invasive risks from low to high. 80 90 100 110 120
ht to dark grey indicate invasive risks from low to high. 80 90 100 110 120
ht to dark grey indicate invasive risks from low to high. 80 90 100 110 120
ht to dark grey indicate invasive risks from low to high. 80 90 100 110 120
ht to dark grey indicate invasive risks from low to high. 80 90 100 110 120
ht to dark grey indicate invasive risks from low to high. 80 90 100 110 120
ht to dark grey indicate invasive risks from low to high. 80 90 100 110 120
ht to dark grey indicate invasive risks from low to high. 80 90 100 110 120
ht to dark grey indicate invasive risks from low to high. 80 90 100 110 120
ht to dark grey indicate invasive risks from low to high. 80 90 100 110 120

80

90

100

110

120

130

Fig. 4 Relative risk of plant invasiveness at provincial level of China using Poisson-Gamma model. Provinces which have red boundaries indicate that their relative risks of plant invasion are significantly higher than 1. Colors from light to dark grey indicate invasive risks from low to high.

IAEES

www.iaees.org

86

Computational Ecology and Software, 2014, 4(2): 82-88

The relative risk of plant invasion based on empirical Bayesian smoothing and Poisson-Gamma models are highly similar (Figs. 3-4). As seen, northern and western parts of China have high risk of plant invasion (dark grey colors). In contrast, southern part of China is less sensitive to plant invasion (light grey colors). Using Poisson exact test, it is identified that 13 provinces from northern part of China have significant relative risk values higher 1 (those provinces with red boundaries in Figs. 3-4). As indicated by Table 1, the SAR model predicted that PRECrange was the most significant covariate (P<0.05) predicting the relative risk of plant invasion across different provinces of China. The next two important variables are TEMrange and PET respectively, but their coefficients are marginally significant (P<0.1). The coefficient for quantifying the strength of spatial autocorrelation is equal to 0.7903, which is a quite large value given that the interval for is [0, 1].

Before applying SAR model, there is significant spatial autocorrelation for the relative risk of plant invasion with Moran’s I=0.639 (P<0.001) (Table 1 and Figs. 3-4). However, after SAR model has been done, the spatial autocorrelation has been removed to a great extent as indicated by the non-significant Moran’s I value for the residuals of the model (Moran’s I=0.131, P>0.05).

Table 1 Coefficient estimation for the covariates in the SAR model used for predicting relative risk of plant invasion across different provinces of China. Here the relative risk values are from empirical Bayesian smoothing method, the result for Poisson- Gamma method is almost identical and thus not presented here.

Covariates

Estimate

Std. Error

z value

Pr(>|z|)

(Intercept)

0.083

0.049

1.695

0.09

Area

0.119

0.102

1.171

0.242

Elev

0.527

0.883

0.597

0.55

TEM

-0.496

1.219

-0.407

0.684

TEMmin

-0.721

0.755

-0.955

0.339

TEMmax

0.505

0.669

0.755

0.45

PET

0.81

0.486

1.668

0.095

PREC

-0.384

0.375

-1.026

0.305

AET

0.341

0.376

0.907

0.364

ElevRange

-0.21

0.789

-0.267

0.79

TEMrange

-0.267

0.151

-1.768

0.077

PRECrange

-0.214

0.076

-2.827

0.005

Moran’s I

P value

Before SAR

0.639

0.001

After SAR

0.131

0.717

4 Discussion Different from previous studies, which showed that areas with high relative outbreak risk of alien plants are typically those with high richness of alien plants, the present study found that those areas with low native plant diversity would have higher risk of alien plant occurrence. The remarkable difference on the risk assessment is due to the inclusion of biodiversity information from native woody plant species. As mentioned above, all the previous studies (Bai et al., 2013; Liu et al., 2005; Wu et al., 2006; Feng and Zhu, 2010) did not explicitly consider the interaction between native and alien plant diversity. The biodiversity resistance hypothesis (Law

IAEES

www.iaees.org

Computational Ecology and Software, 2014, 4(2): 82-88

87

and Morton, 1996; Stachowicz et al., 2002) provides a way to effectively incorporate diversity information from native plants for quantifying invasive risk of alien plants over provinces. Based on the comparative results (Figs 3-4), both Poisson-Gamma and empirical Bayesian smoothing methods consistently identified that provinces from northern part of China are remarkably higher than those provinces in southern part of China. This is expected, since we applied biodiversity resistance hypothesis when modeling the invasive risk of alien plants. In southern and south-western part of China, the provinces there have very high species richness in comparison to those provinces in northern part of China. In particular, Yunnan province constitutes a well recognized biodiversity hotspot of the world (Chen and Bi, 2007; Myers et al., 2000; Zhang et al., 2010; Hua, 2008; Zhang, 2012b). Thus, even though this province is found to have a very high number of alien plants (Fig. 2), its relative risk of alien plant invasion is still comparatively low (Figs. 3-4). The very reason is that the calculation of relative risk of alien plant invasion requires the population at risk in the denominator, for which we use the native woody plant richness as the surrogate. Yunnan Province has the highest woody plant diversity (Fig. 1), thus its relative risk is predicted to be low accordingly.

Acknowledgements This work is supported by China Scholarship Council.

References Bai F, Chisholm R, Sang W, Dong M. 2013. Spatial risk assessment of alien invasive plants in China. Environmental Science & Technology, 47: 7624-7632 Bivand R, Altman M, Anselin L . 2013. spdep: spatial dependence: weighting schemes, statistics and models. http://CRAN.R-project.org/package=spdep Chen Y. 2013. Distributional patterns of alien plants in China: the relative importance of phylogenetic history and functional attributes. ISRN Ecology, 527052 Chen Y, Bi J. 2007. Biogeography and hotspots of amphibian species of China: Implications to reserve selection and conservation. Current Science, 92: 480-489 Dormann C. 2007. Effects of incorporating spatial autocoorrelation into the analysis of species distribution data. Global Ecology and Biogeography, 16: 129-138 Dormann C, McPherson J, Araujo M, et al. 2007. Methods to account for spatial autocorrelation in the analysis of species distirbutional data: a review. Ecography (Cop), 30: 609-628 Feng J, Zhu Y. 2010. Alien invasive plants in China: risk assessment and spatial patterns. Biodiversity and Conservation, 19: 3489-3497 Hua Z. 2008. Advances in biogeography of the tropical rain forest in southern Yunnan, southwestern China. Tropical Conservation Science, 1: 34-42 Law R, Morton R. 1996. Permanence and the assembly of ecological communities. Ecology, 77: 762-775 Liu J, Liang S, Liu F, et al. 2005. Invasive alien plant species in China: regional distribution patterns. Diversity and Distributions, 11: 341-347 Myers N, Mittermeier R, Mittermeier C, et al. 2000. Biodiversity hotspots for conservation priorities. Nature, 403: 853-858 Qian H. 2013. Environmental determinants of woody plant diversity at a regional scale in China. PLoS One, 8:

e75832

IAEES

www.iaees.org

88

Computational Ecology and Software, 2014, 4(2): 82-88

R Development Core Team. 2011. R: A Language and Environment for Statistical Computing. Vienna, Austria Stachowicz J, Fried H, Osman R, Whitlatch R. 2002. Biodiversity, invasion resistance, and marine ecosystem function: reconciling pattern and process. Ecology, 83: 2575-2590 Wang Z, Chen Y, Chen Y. 2009. Functional grouping and establishment of distribution patterns of invasive plants in China using self-organizing maps and indicator species analysis. Archives of Biological Science, 61: 71-78 Wu X, Luo J, Chen J, Li B. 2006. Spatial patterns of invasive alien plants in China and its relationship with environmental and anthropological factors. Journal of Plant Ecology, 30: 576-584 Yan X, Shou H, Ma J. 2012. The problem and status of the alien invasive plants in China. Plant Diversity and Resources, 34: 287-313 Zhang D, Chen M, Murphy R, et al. 2010. Genealogy and palaeodrainage basins in Yunnan Province:

phylogeography of the Yunnan spiny frog, Nanorana yunnanensis (Dicroglossidae). Molecular Ecology, 19: 3406-3420 Zhang WJ. 2011. Constructing ecological interaction networks by correlation analysis: hints from community sampling. Network Biology, 1(2): 81-98 Zhang WJ. 2012a. Computational Ecology: Graphs, Networks and Agent-based Modeling. World Scientific, Singapore Zhang WJ. 2012b. Did Eucalyptus contribute to environment degradation? Implications from a dispute on causes of severe drought in Yunnan and Guizhou, China. Environmental Skeptics and Critics, 1(2): 34-38 Zhang WJ, Chen B. 2011. Environment patterns and influential factors of biological invasions: a worldwide survey. Proceedings of the International Academy of Ecology and Environmental Sciences, 1(1): 1-14

IAEES

www.iaees.org

Computational Ecology and Software, 2014, 4(2): 89-103

Article

Stability analysis of a discrete ecological model

Q. Din 1 , E. M. Elsayed 2,3

1 Department of Mathematics, Faculty of Basic & Applied Sciences, University of PoonchRawalakot, Pakistan

2 Department of Mathematics, Faculty of Science,King Abdulaziz University, P.O. Box 80203, Jeddah 21589, Saudi Arabia

3 Department of Mathematics, Faculty of Science, Mansoura University, Mansoura 35516, Egypt E-mail: qamar.sms@gmail.com,emmelsayed@yahoo.com

Received 14 January 2014; Accepted 20 February 2014; Published online 1 June 2014

; Accepted 20 February 2014 ; Published online 1 June 2014 Abstract In this paper, we

Abstract

In this paper, we study the qualitative behavior of following discrete-time population model:

, , where parameters , , , , , and initial conditions , , , are positive real numbers. More precisely, we investigate the existence and uniqueness of positive equilibrium point,boundedness character, persistence, local asymptotic stability, global behavior and rate of convergence of unique positive equilibrium point of this model. Some numerical examples are given to verify our theoretical results.

Keywords population models; difference equations; steady-states; boundedness; local and global character.

Computational Ecology and Software ISSN 2220 ­721X URL: http://www.iaees.org/publications/journals/ces/online­ version.asp RSS: http://www.iaees.org/publications/journals/ces/rss.xml E­mail: ces@iaees.org Editor ­in­Chief: WenJun Zhang Publisher: International Academy of Ecolo gy and Environmental Sciences

1 Introduction

In population dynamics, difference and differential equations are used in many models (Nedorezov, 2012; Nedorezov and Sadykov, 2012; Elena et al., 2013). Exponential difference equations have many applications in population dynamics. One can see the references (El-Metwally et al., 2001; Papaschinopoulos et al., 2011; Papaschinopoulos et al., 2012) for some interesting results related to qualitative behavior of population models. From Zhou and Zou (2003) and Liu (2010) it is clear that difference equations are much better as compared to differential equations, when the populations are of non-overlapping generations. In literature, there are many papers in which discrete dynamical systems are used to study the qualitative behavior of population models (Ahmad, 1993; Zhou and Zou, 2003; Tang andZou, 2006; Din, 2013; Din and. Donchev, 2013; Din et al., 2013; Din, 2014). In case of discrete–time models one can compute mathematical results more efficiently. There are many research papers which are related to mathematical models in population dynamics. During the last few decades, many researchers are attracted by such computational models. For detail of biological models see (Edelstein-Keshet, 1988; Brauer and Castillo-Chavez, 2000; Allen, 2007). Nonlinear difference equations are

IAEES

www.iaees.org

90

Computational Ecology and Software, 2014, 4(2): 89-103

very important in application. Naturally, such equations appear as discrete approximation of differential and delay differential equations and these models are related to applied sciences such as ecology,biology, physics, chemistry, physiology, engineering, bio-chemistry and economics. We cannot find the solutions of nonlinear difference equations in all cases. So, one can study the behavior of solutions by local asymptotic stability of equilibrium points. El-Metwally et al. (2001) studied the qualitative behavior of following population model: . Papachinopoulous et al. (2011) investigated the dynamics of following population model: , . Recently, Papachinopoulous et al. (2012) study following two exponential nonlinear difference equations: ,

, and , .

Our aim in this paper is to investigate existence and uniqueness of positive equilibrium point, boundedness character, persistence, local asymptotic stability, global behavior of unique positive equilibrium point of following four-dimensional discrete-time population model:

,

,

1

where parameters , , , , , and initial conditions , , , are positive real numbers. System (1) is an extension of (Papachinopoulouset al., 2011).

2 Boundedness and Persistence

The following Theorem shows that every positive solution of (1) is bounded and persists.

Theorem 2.1 Assume that 4 2 , 4 2 , then every positive solution {( , } of system (1) is bounded and persists.

Proof. Consider the following second-order difference equations:

, . The solutions of (2.1) are given by

1 2 4

(2.1)

2 4 ,

2 4 2 4 ,

where , , , depend on initial values , , , .

Assume that 4 2 , 4 2 . Then it follows that:

and for all 1,2, . Evidently, the sequences

and are bounded for all 1,2, . Then, by comparison we obtain

and

for all 1,2, . Hence we have

and for all 1,2, . This completes the proof.

3 Existence of Invariant Set for Solutions

Theorem 3.1 Let , be a positive solution of system of (1). Then ,

IAEES

www.iaees.org

Computational Ecology and Software, 2014, 4(2): 89-103

91

,

,

is invariant set for system (1).

Proof. Let

,

be a solution of system (1) with initial conditions , ,

and

, , .

Then, it follows from system (1) that

1 1 1 ,

and

1 1 1 .

Hence,

1 from system (1) one has:

and . Suppose that result is true

for 1, i.e.,

1 1 1 ,

and .

Then,

Hence, the proof follows by induction.

1 1 1

for

4 Existence and Uniqueness of Positive Equilibrium and Local Stability

Theorem 4.1 Assume that 4 2 , 4 2 , 0 , 1 and

exp

1 exp

.

Then the system (1) has a unique positive equilibrium point , such that ,

,

.

Proof. Consider the following system of equations:

and

,

Then, one has

and

. Taking , where

and

. Then, it follows that

. Now, 0 if and only if 1 0,

.

., 1 exp

.

Hence,

it

follows

that 0

if

Furthermore, we have

F

,

and

only

if 1 exp

.

IAEES

www.iaees.org

92

Computational Ecology and Software, 2014, 4(2): 89-103

where

.

It

is

easy

to see that 0

, i.e., exp

.

It follows that 0, if and only if exp

Hence, has positive solution in . Moreover, we have that

1

1

1

0.

.

if

and

only

if

Hence, has a unique positive solution in . The proof is therefore completed. Consider a fourth-dimensional discrete dynamical system of the form:

, , ,

, 0,1, ,

(4.1)

, , ,

where : and : are continuously differentiable functions, and , are real intervals.

conditions ,

for 1,0 . Along with system (4.1) we consider a vector map of the form , , , . A constant solution or equilibrium point (fixed-point) of (4.1) is point that satisfies:

, , , ,

Furthermore, a solution , of system

(4.1) is uniquely determined

by initial

, , , .

The point , is also point called a fixed point of the vector map . Definition 4.1 Assume that , be aconstant solution (steady-state) of (4.1).

1. A constant solution , is called stable, if for an arbitrary 0 there exists a positive such that

, , , then one has ,

for any initial condition

, for all 0, where · denotes usual Euclidean norm in .

2.

3.

4.

5.

Assume that , be a fixed-point of map defined by = , , , , where and are continuously differentiable functions about ( , . The linearized system of (4.1)about the fixed-point ( , ) is given by:

A constant solution ( , ) is said to be asymptotically global attractor if it is global attractor and stable.

, , 1, 0 if

A constant solution , is called unstable if it did not satisfy stability condition.

A constant solution , is locally asymptotically stable, if there exists a positive such that

, , and

, , as .

A constant solution( , ) is said to be global attracter if , , as .

where

,

and is a Jacobain matrix of (4.1) about the fixed-point , .

Lemma 4.1 (Sedaghat, 2003) Consider the discrete dynamical system of the form , 0,1, ,

is

(i) If all eigenvalues of Jacobian at

such that

be an equilibrium point of (4.1). Then following statements are true:

lie inside an open unit disc

| | 1,

then the fixed-point

IAEES

www.iaees.org

Computational Ecology and Software, 2014, 4(2): 89-103

93

locally asymptotically stable.

(ii) If one of the eigenvalue of Jacobian matrix

To construct corresponding linearized form of system (1) we consider the following transformation:

has an absolute value greater than one, then

, , , , , , , ,

(4.2)

is unstable.

where , , and . The Jacobian matrix about the fixed point , of (1) under the transformation (4.2) is given by

,

1

0

0

0

1

0

0

0

0

0

.

Theorem 4.2The unique positive equilibrium point , of system (1) is locally asymptotically stable if the following condition is satisfied:

1 1 1

1. (4.3)

Proof. The characteristic polynomial of Jacobain matrix , about , is given by

,

(4.4)

where and . Let and . Assume that condition (4.3) is satisfied and | | 1, then one has | | | | | |

1 1

1

1.

Then, by Rouche’s theorem and have same number of zeroes in an open unit disc | | 1. Hence, all the roots of (4.3) satisfies | | 1, and it follows from Lemma 4.1 that the unique positive equilibrium point , of system (1) is locally asymptotically stable.

5 Global Stability Theorem 5.1 The unique positive equilibrium point , of system (1) is globally asymptotically stable. Proof. Let , be an arbitrary positive solution of system (1) and let 0 Ι lim ,0 Ι lim ,S lim ,S lim . Then, from (1) one has:

Ι Ι Ι Ι Ι Ι S S S Ι S S S Ι

Furthermore,

IAEES

Ι

Ι

1

,

,

1

www.iaees.org

94

Computational Ecology and Software, 2014, 4(2): 89-103

Then, it follows that:

S

 

1

Ι ,

S

 

1

Ι .

 

Ι

1

,

Ι S

S

1

,

Ι S

Ι

1

Ι ,

Ι S

Ι

1 Ι .

Hence one has

Ι and

Ι

S

Ι

Ι . Now consider the following functions:

1

1 , Ι, y ϵ

Moreover, for every , Ι :

1

1 0,

1 0.

1

Hence, Ι and

Ι .

J.

6 Existence of Unbounded Solutions Theorem 6.1 Let , 0,1 and {( , be a positive solution of system (3.1) Then following are true:

If , , .

(ii) If , , .

Proof. Let . Choosing the initial conditions , , ,

for the system (1)

, and , with

, . Then from system (1) one has:

such that

, and

. Suppose that result is true for 1, i.e., and . Then for 1 from system (1) one has:

,

and

 

Hence,

, for

all

.

1,2, . Furthermore,

. Let with initial conditions , . Then,

IAEES

www.iaees.org

Computational Ecology and Software, 2014, 4(2): 89-103

95

2 4 2 4 . Hence by comparison . But , so lim . Moreover, . Hence, lim . The proof of is similar, therefore it is omitted.

7 Rate of Convergence To discuss the rate of convergence of discrete dynamical system, we consider the following system:

,

(7.1)

where is a -dimensional vector, is a constant matrix, and : is a matrix function such that:

0,

(7.2)

as , where . denotes a matrix norm which associated with the vector norm of the form:

, . Lemma 7.1 (Perron’s theorem) (Pituk, 2002) Assume that the condition (7.2) holds true. If be a solution of (7.1) then either 0 for all , or

lim

is defined and it is equal to the absolute value of one of the eigenvalues of matrix . Lemma 7.2 (Pituk, 2002) Assume that condition (7.2) holds. If is a solution of (7.1), then either 0 for all , or

lim

,

is defined and is equal to the absolute value of one the eigenvalues of matrix .

Let , be any solution of the system (1) such

that lim and lim , where

, and , . To find the error terms, one has from the system (1)

and

,

.

Let , and , then one has

 

and

 

where

,

,

IAEES

www.iaees.org

96

Computational Ecology and Software, 2014, 4(2): 89-103

Moreover,

, , , 0, , ,

0, .

, , , 0 , ,

, 0, .

Now the limiting system of error terms can be written as

1

0

0

0

1

0

0

0

0

0

,

which is similar to linearized system of (1) about the equilibrium point , .

Using Lemma 7.1 one has the following result.

Theorem 7.1 Assume that , be a solution of (1) such that lim , lim , where

, be unique equilibrium point of (1), then the error vector

satisfies both of the following asymptotic relations:

lim

lim

, , , , ,

, , , , ,

of every solution of system (1)

where , , , , are the characteristic roots of the Jacobian matrix , .

8 Examples

In this section, we consider the following numerical examples.

Example 8.1 Let 1.1, 0.01, 2.9, 1.2, 0.02, 2.304.Then, system (1) can be written

as

1.1 0.01 2.9 , 1.2 0.02 2.9 (8.1)

with initial conditions 1.3, 1.2, 1.3, 1.5.

In this case the unique positive equilibrium point of the system (8.1)is given by

, 7.93981,1.22552 . Moreover, in Fig. 8.1 the plot of is shown in Fig. 8.1a, the plot of

is

shown in Fig. 8.1b, and an attractor of the system is shown in Fig. 8.1c.

IAEES

www.iaees.org

Computational Ecology and Software, 2014, 4(2): 89-103

97

Computational Ecology and Software, 2014, 4(2): 89-103 97 (8.1a) Plot of for the system (8.1) IAEES

(8.1a) Plot of for the system (8.1)

4(2): 89-103 97 (8.1a) Plot of for the system (8.1) IAEES (8.1b) Plot of for the

IAEES

(8.1b) Plot of for the system (8.1)

www.iaees.org

98

Computational Ecology and Software, 2014, 4(2): 89-103

98 Computational Ecology and Software, 2014, 4(2): 89-103 (8.1c) An attractor of the system (8.1) Fig.

(8.1c) An attractor of the system (8.1) Fig. 8.1 Plots for the system (8.1).

Example 8.2 Let 1.5, 0.026, 3.14, 1.67, 0.0096, 2.6. Then, system (1) can be written as 1.5 0.026 3.14 , 1.67 0.0096 2.6 (8.2) with initial conditions 2.2, 2.1, 2.35, 2.25.

by

, 2.16435,2.41379 . Moreover, in Fig. 8.2 the plot of is shown in Fig. 8.2a, the plot of is shown in Fig. 8.2b, and an attractor of the system is shown in Fig. 8.2c.

In

this

case

the

unique

positive

equilibrium

point

of

the

system

(8.2)

is

given

positive equilibrium point of the system (8.2) is given IAEES (8.2a) Plot of for the system

IAEES

(8.2a) Plot of for the system (8.2)

www.iaees.org

Computational Ecology and Software, 2014, 4(2): 89-103

99

Computational Ecology and Software, 2014, 4(2): 89-103 99 (8.2b) Plot of for the system (8.2) (8.2c)

(8.2b) Plot of for the system (8.2)

4(2): 89-103 99 (8.2b) Plot of for the system (8.2) (8.2c) An attractor of the system

(8.2c) An attractor of the system (8.2) Fig.8.2 Plots for the system (8.2)

Example 8.3 Let 1.5, 0.0001, 3.5, 1.67, 0.00002, 2.77. Then, system (1) can be written as

2.3 0.0001 3.5 1.67 0.00002 2.77 (8.3) with initial conditions 2.25, 2.15, 2.36, 2.27. In this case the unique positive equilibrium point of the system (8.3) is given by , 1.71494,3.33035 .

IAEES

www.iaees.org

100

Computational Ecology and Software, 2014, 4(2): 89-103

Moreover, in Fig. 8.3 the plot of is shown in Fig. 8.3a, the plot of is shown in Fig. 8.3b, and an attractor of the system is shown in Fig. 8.3c.

8.3b, and an attractor of the system is shown in Fig. 8.3c. (8.3a) Plot of for

(8.3a) Plot of for the system (8.3)

shown in Fig. 8.3c. (8.3a) Plot of for the system (8.3) IAEES (8.3b) Plot of for

IAEES

(8.3b) Plot of for the system (8.3)

www.iaees.org

Computational Ecology and Software, 2014, 4(2): 89-103

101

Computational Ecology and Software, 2014, 4(2): 89-103 101 (8.3c) An attractor of the system (8.3) Fig.

(8.3c) An attractor of the system (8.3) Fig. 8.3 Plots for the system (8.3).

Example 8.4 Let 2.3, 0.0009, 4.5, 2.09, 0.0008, 6.35. Then system (1) can be written as:

2.3 0.0009 4.5 , 2.09 0.0008 6.35 (8.4) with initial conditions 2.1, 2.2, 1.1, 1.5. In this case the unique positive equilibrium point of the system (8.4) is given by , 2.3581,5.24482 . Moreover, in Fig. 8.4 the plot of is shown in Fig. 8.4a,the plot of is shown in Fig. 8.4b,and an attractor of the system (8.4) is shown in Fig.

8.4c.

an attractor of the system (8.4) is shown in Fig. 8.4c. IAEES (8.4a) Plot of for

IAEES

(8.4a) Plot of for the system (8.4)

www.iaees.org

102

Computational Ecology and Software, 2014, 4(2): 89-103

102 Computational Ecology and Software, 2014, 4(2): 89-103 (8.4b) Plot of for the system (8.4) (8.4c)

(8.4b) Plot of for the system (8.4)

2014, 4(2): 89-103 (8.4b) Plot of for the system (8.4) (8.4c) An attractor of the system

(8.4c) An attractor of the system (8.4) Fig. 8.4 Plots for the system (8.4).

9 Conclusion This work is related to qualitative behavior of a four-dimensional exponential discrete population model. We prove that system (1) has a unique positive equilibrium point which is globally asymptotically stable. Usually, biologists believe that a globally asymptotically stable equilibrium point is very important in ecological point of view. Method of linearization is used for local asymptotic stability of steady-state of (1). Furthermore, we prove the boundedness and persistence of positive solutions of (1), and an invariant set for its solutions is

IAEES

www.iaees.org

Computational Ecology and Software, 2014, 4(2): 89-103

103

investigated. We prove the rate of convergence of positive solutions of system (1) which converge to its unique positive equilibrium point.

References Ahmad S. 1993. On the nonautonomous Lotka-Volterra competition equation. Proceedings of the American Mathematical Society, 117: 199-204 Allen LJS. 2007. An Introduction to Mathematical Biology. Prentice Hall, USA Brauer F, Castillo-Chavez C. 2000.Mathematical Models in Population Biology and Epidemiology. Springer Din Q. 2013.Dynamics of a discrete Lotka-Volterra model. Advances in DifferenceEquations, 1: 1-13 Din Q. 2014.Global stability of a population model. Chaos, Soliton and Fractals, 59: 119-128. Din Q, Donchev T. 2013.Global character of a host-parasite model. Chaos, Soliton and Fractals, 54: 1-7 Din Q, Khan AQ, Qureshi MN. 2013. Qualitative behavior of a host-pathogen model. Advances in Difference Equations, 1: 263 Edelstein-Keshet L. 1988. Mathematical Models in Biology. McGraw-Hill, USA Elaydi S. 2005. An Introduction to Difference Equations (3rd ed). Springer-Verlag, New York, USA Elena E, Grammauro M, Venturino E. 2013. Predator’s alternative food sources do not support ecoepidemics with two strains-diseased prey. Network Biology, 3(1): 29-44 El-Metwally E, Grove EA, Ladas G, et al. 2001.On the difference equation . Nonlinear Analysis, 47: 4623-4634 Grove EA, Ladas G. 2004. Periodicities in Nonlinear Difference Equations. Chapman and Hall/CRC Press, Boca Raton, USA Liu X. 2010. A note on the existence of periodic solution in discrete predator-prey models. Applied Mathematical Modelling, 34: 2477-2483 Nedorezov LV. 2012. Continuous-discrete model of population dynamics with time lag in a reaction of intra- population self-regulative mechanisms. Network Biology, 2(4):139-147 Nedorezov LV, Sadykov AM. 2012. About a modification of Rogers model of parasite-host system dynamics. Proceedings of the International Academy of Ecology and Environmental Sciences, 2(1): 41-49 Papaschinopoulos G, Radin MA, Schinas CJ. 2011. On the system of two difference equations of exponential form: , . Mathematical and Computer Modelling, 54:

2969-2977

Papaschinopoulos G, Schinas CJ. 2012. On the dynamics of two exponential type systems of difference equations. Computers & Mathematics with Applications, 64(7): 2326-2334 Pituk M. 2002.More on Poincare’s and Perron’s theorems for difference equations. Journal of Difference Equations and Applications, 8: 201-216 Sedaghat H. 2003. Nonlinear difference equations: Theory with applications to social science models. Kluwer Academic Publishers, Dordrecht, Netherlands Tang X, Zou X. 2006.On positive periodic solutions of Lotka-Volterra competition systems withdeviating arguments. Proceedings of the American Mathematical Society, 134: 2967-2974 Zhou Z, Zou X. 2003. Stable periodic solutions in a discrete periodic logistic equation. Applied Mathematics Letters, 16(2): 165-171

IAEES

www.iaees.org

Computational Ecology and Software, 2014, 4(2): 104-115

Article

Qualitative behavior of an anti-competitive system of third-order rational difference equations

Q. Din 1 , M. N. Qureshi 2 , A. Q. Khan 2

1 Department of Mathematics, Faculty of Basic & Applied Sciences, University of PoonchRawalakot, Pakistan

2 Department of Mathematics, University of Azad Jammu & Kashmir, Muzaffarabad, Pakistan E-mail: qamar.sms@gmail.com,nqureshi@ajku.edu.pk,abdulqadeerkhan1@gmail.com

Received 6 December 2013;