Home Global stability and asymptotic profiles of a partially degenerate reaction diffusion Cholera model with asymptomatic individuals
Article Open Access

Global stability and asymptotic profiles of a partially degenerate reaction diffusion Cholera model with asymptomatic individuals

  • Shengfu Wang and Linfei Nie EMAIL logo
Published/Copyright: December 18, 2024

Abstract

Considering the prevalence of asymptomatic individuals during the spread of disease, this article develops a model of degenerate reaction diffusion Cholera with asymptomatic individuals. First, the well-posedness of model is studied, including the global existence of solutions and the existence of attractor. Second, the basic reproduction number 0 is defined to determine whether the disease is vanishing or persistent. In particular, we also analyze the asymptotic behavior of the endemic steady state when the diffusion rate of susceptible or asymptomatic individuals tends to 0 or infinity. Finally, by fitting the theoretical results with some numerical simulations, we find that the spatial distribution of disease and local epidemic risk are less affected by the mobility of susceptible populations, whereas the mobility of asymptomatic or symptomatic populations significantly affects the spatial and temporal distribution of infected populations. In addition, we found that the proportion of asymptomatic individuals to infected individuals is also a key factor in disease epidemics, and how to quickly diagnose asymptomatic individuals for disease control and prevention should be of a particular concern.

MSC 2010: 35B40; 35J57; 35K57; 92D30

1 Introduction

Cholera is an acute diarrheal infection caused by eating or drinking food and water contaminated with Vibrio cholerae. It is typical symptoms are diarrhea and vomiting, dehydration, muscle cramps, etc. [22]. When medical care is inadequate and treatment is not timely, the mortality rate is high. Although health care is improving in many countries around the world, the highly contagious cholera still occurs after natural disasters such as earthquakes and tsunamis, or in areas with relatively poor sanitation. In recent years, cholera outbreaks have remained severe in less developed countries, including Haiti, Yemen, Zimbabwe, and the Congo. Researchers estimate that there are 1,300,000–4,000,000 cases of cholera worldwide each year, and 21,000–143,000 deaths due to infection [38].

As is well known, mathematical models play a critical role in comprehending the dynamics of disease transmission. Over the past decade or so, many scholars have studied the spread of cholera through the development of various mathematical models. Capasso and Paveri-Fontana [6] proposed the first model of indirect cholera transmission in response to the cholera epidemic that occurred in the Mediterranean region in the summer of 1973, and the model included only infected population and pathogen. Codeço [7] in 2001 extended this model by adding equations for susceptible populations and showed the aquatic hosts have an important influence on cholera transmission. Subsequently, many more researchers have extended the model in [7], such as, Tien and Earn [30] extended the SIR model by adding pathogen concentration compartment ( W ), considered the impact of multiple transmission routes, and investigated the global stability of the equilibrium points. Note that, during the spread of the disease, the fact that asymptomatic infected individuals have no clinical symptoms associated with the disease can be just as contagious as symptomatic ones, leading to a longer duration of the disease in some areas. There is strong evidence that asymptomatic carriers of cholera pathogen may be responsible for the long-distance spread of the pathogen elsewhere, and their numbers may be much higher than reported [15]. Ogola et al. [26] analyzed a cholera model with indirect transmission and asymptomatic infected individuals, and their results suggested that asymptomatic individuals are the main cause of cholera epidemics in Senegal and other sub-Saharan African countries. Other related articles can be found in [1,3,14,23,31] and the references therein.

It is worth noting that strengthening personal hygiene, improving environmental sanitation, and using safe drinking water are the most important strategies for controlling and preventing cholera. Therefore, vaccination has become an effective measure against cholera that has received widespread interest and became the subject of many clinical and theoretical studies. Many researchers as well have taken the cholera vaccine into account and have come up with a wealth of theoretical results. In [2,16], it was shown that backward branching can exist in models of all susceptible individuals vaccinated when the basic reproduction number is less than one. Cai et al. [5] modeled SAIVB infectious diseases with vaccine age and showed that not considering vaccine age underestimates the risk of transmission of cholera outbreaks.

Note that most of the aforementioned models are either ordinary differential equations (ODEs) or hybrids with partial differential equations (PDEs), which do not introduce spatial heterogeneity, resulting in an inadequate comprehension of the spatial propagation of cholera epidemics. Due to the differences in temperature, climate and geographic characteristics among different countries and regions, which in turn may contribute to the spatial distribution of the disease. Therefore, it is particularly important to integrate geographic and pathogen characteristics to develop cholera models with spatial heterogeneity. Much work has been done on the spatial dynamics of cholera transmission. Zhang et al. [41] developed a reaction diffusion model incorporating both direct and indirect transmission routes, and investigated the effects of diffusion coefficients and multiple transmission routes on disease propagation. Wang and Feng [34] proposed a model with different diffusion rates and spatial heterogeneity to study threshold dynamics and asymptotic behavior. Wang and Wu [35] investigated a degenerate cholera diffusion model, focusing on the asymptotic behavior of the positive steady state at small and large diffusion rates in susceptible individuals. Other studies on asymptotic behavior are continuing and are detailed in [4,9,36,39].

As mentioned earlier, the existing studies have mainly focused on how spatial heterogeneity and individual mobility contribute to the cholera propagation, prevention, and control. Actually, vaccine effectiveness, the prevalence of asymptomatic individuals, and spatial heterogeneity are all essential features that must be taken into account in the propagation and sustenance of cholera. Throughout this article, we establish a model of cholera with general incidence, spatial transmission, and incomplete immunity, ignoring person-to-person transmission (direct transmission) and assuming no transmission of environmental viruses, concentrating on the effects of these features on the spatial and temporal distribution of the disease and its prevention and control.

The rest of the article is organized as follows. The model is presented in Section 2. In Section 3, the global existence, ultimate boundedness, and the existence of attractors of model solution are studied. In Section 4, some relations between the principal eigenvalues and the basic reproduction number 0 are presented, and formulas for the local reproduction number ˜ 0 are derived. The threshould dynamics of the model are studid by 0 in Section 5. The asymptotic distribution of the endemic steady state, as the diffusion coefficients converge to different scenarios, is examined in Section 6. Finally, a number of numerical simulations are carried out in Section 7 to validate the theoretical results, and a brief conclusion is presented in Section 8.

2 Model formulation

Let S ( x , t ) , V ( x , t ) , I A ( x , t ) , and I S ( x , t ) stand for the densities of the susceptible individuals, vaccinated individuals, asymptomatic individuals, and symptomatic individuals at position x and time t , respectively. The concentration of Vibrio cholerae in the environment at position x and time t is denoted by B ( x , t ) . D i > 0   ( i = 1 , , 4 ) means the dispersal rates of S , V , I A , and I S , respectively. Λ ( x ) denotes the recruitment rate of the susceptible individuals. μ ( x ) denotes the natural mortality rate of the population. r A ( x ) and r S ( x ) denote the recovery rate of asymptomatic individuals and symptomatic individual, respectively. δ ( x ) means the probability that an asymptomatic individual turns into a symptomatic individuals. The functions η ( x ) and σ represent the rate of immune loss and the probability that the vaccine is immune protective, respectively. β ( x ) f ( S , B ) denotes the infectivity of a susceptible individual in contact with a pathogen, where β ( x ) indicates the effective contact rate betweens usceptible individuals and pathogens. We hypothesize that a fraction p   ( 0 < p < 1 ) of the entering susceptible individuals is vaccinated. In addition, when the susceptible population is infected with the disease, a fraction of θ   ( 0 θ 1 ) becomes asymptomatic, while 1 θ develops symptoms. Compared to symptomatic individuals, asymptomatic individuals carry a lower amount of virus infectiously, and therefore, they exhibit lower shedding rates and contribute less to the pathogen density in the environment, denoted by γ A ( x ) and γ S ( x ) for asymptomatic and symptomatic individuals, respectively. α ( x ) stands for the clearance rate of the environment. Motivated by these considerations, we propose the following model

(1) S t = D 1 Δ S + ( 1 p ( x ) ) Λ ( x ) μ ( x ) S β ( x ) f ( S , B ) + η ( x ) V , V t = D 2 Δ V + p ( x ) Λ ( x ) ( μ ( x ) + η ( x ) ) V σ β ( x ) g ( V , B ) , I A t = D 3 Δ I A + θ β ( x ) f ( S , B ) + σ β ( x ) g ( V , B ) ( μ ( x ) + r A ( x ) + δ ( x ) ) I A , I S t = D 4 Δ I S + ( 1 θ ) β ( x ) f ( S , B ) + δ ( x ) I A ( μ ( x ) + r S ( x ) ) I S , B t = γ A ( x ) I A + γ S ( x ) I S α ( x ) B ,

where ( x , t ) D × [ 0 , ) , with the boundary condition

(2) S ν = V ν = I A ν = I S ν = 0 , x D , t > 0

and S ( x , 0 ) = S 0 ( x ) , V ( x , 0 ) = V 0 ( x ) , I A ( x , 0 ) = I A 0 ( x ) , I S ( x , 0 ) = I S 0 ( x ) , B ( x , 0 ) = B 0 ( x ) , and x D , where D is a general open bounded domain in R n with smooth boundary D . ν is the outward normal unit vector on D . The initial value ( S 0 ( x ) , V 0 ( x ) , I A 0 ( x ) , I S 0 ( x ) , B 0 ( x ) ) is a nonegative continuous function. The propagation diagram of model eq. (1) is depicted in Figure 1.

Figure 1 
               Schematic of the propagation of model (1).
Figure 1

Schematic of the propagation of model (1).

In this article, we formulate the following general hypotheses:

  1. Functions f ( S , B ) , g ( V , B ) C 2 ( R + × R + ) , f 1 ( S , B ) , g 1 ( V , B ) , f 2 ( S , B ) , and g 2 ( V , B ) are positive for S , V , B > 0 ; f ( S , B ) = 0 if and only if S B = 0 ; g ( V , B ) = 0 if and only if V B = 0 ; f 22 ( S , B ) 0 , g 22 ( V , B ) 0 , for S , V , B 0 . Here, ( F ) 1 and ( F ) 2 denote the first-order derivatives of the function f ( S , B ) or g ( V , B ) with respect to S or V and B , respectively. ( F ) 22 denotes the second-order derivative of the function f ( S , B ) or g ( V , B ) with respect to B ;

  2. There exists a Hölder continuous function k i : R R + such that f ( y , B ) k 1 y B , g ( y , B ) k 2 y B , y { S , V } for x D , S , V , B 0 .

3 Well-posedness of the problem

Throughout this section, we shall confirm that model (1) has a unique global nonnegative solution and admits a connected global attractor. Let X C ( D ¯ , R 4 ) be equipped with the supreme norm X , and X + C ( D ¯ , R + 4 ) be the positive cone of X. For that, we introduce the notation h * max x D ¯ { h ( ) } , h * min x D ¯ { h ( ) } , where h ( ) = λ ( ) , p ( ) , μ ( ) , β ( ) , η ( ) , r A ( ) , r S ( ) , δ ( ) , γ A ( ) , α ( ) , γ S ( ) . Let T k ( t ) : C ( D ¯ , R ) C ( D ¯ , R )   ( k = 1 , , 4 ) be the C 0 -semigroups of D k Δ π k ( x ) with (2), where π 1 ( x ) = μ ( x ) , π 2 ( x ) = μ ( x ) + η ( x ) , π 3 ( x ) = μ ( x ) + r A ( x ) + δ ( x ) and π 4 ( x ) = μ ( x ) + r S ( x ) . Then, for any initial value ϕ X + , we have

( T k ( t ) ϕ ) ( x ) = D Γ k ( t , x , y ) d y , t > 0 , ϕ C ( D ¯ , R ) , k = 1 , 2 , 3 , 4 ,

where Γ k ( x , t , y ) denotes the Green function associated with D k Δ π k ( x ) subject to the Neumann boundary conditions, respectively. Let ( T 5 ( t ) ϕ ) ( x ) = e α ( x ) t ϕ ( x ) , thus, T ( t ) ( T 1 ( t ) , T 2 ( t ) , T 3 ( t ) , T 4 ( t ) , T 5 ( t ) ) T : X X forms a strongly continuous semigroup.

Furthermore, set Z = ( Z 1 , Z 2 , Z 3 , Z 4 , Z 5 ) T : X + X be given by

Z 1 ( ϕ ) ( ) = ( 1 p ( x ) ) Λ ( x ) β ( x ) f ( ϕ 1 , ϕ 5 ) + η ( x ) ϕ 2 , Z 2 ( ϕ ) ( ) = p ( x ) Λ ( x ) σ β ( x ) g ( ϕ 2 , ϕ 5 ) , Z 3 ( ϕ ) ( ) = θ β ( x ) f ( ϕ 1 , ϕ 5 ) + δ ( x ) β ( x ) g ( ϕ 2 , ϕ 5 ) , Z 4 ( ϕ ) ( ) = ( 1 θ ) β ( x ) f ( ϕ 1 , ϕ 5 ) + σ ϕ 3 , Z 5 ( ϕ ) ( ) = γ A ( x ) ϕ 3 + γ S ( x ) ϕ 4 .

By the aforementioned setting, we can rewrite model (1) as follows:

(3) u ( x , t , ϕ ) = T ( t ) ϕ + 0 t T ( t s ) Z ( u ( x , s , ϕ ) ) d s ,

where u ( x , t , ϕ ) = ( S ( x , t , ϕ ) , V ( x , t , ϕ ) , I A ( x , t , ϕ ) , I S ( x , t , ϕ ) , B ( x , t , ϕ ) ) T .

The local solutions of the model (1) on X + is given first.

Lemma 1

For any ϕ X + , model (1) has a unique mild solution u ( x , t , ϕ ) defined on [ 0 , τ max ) , where τ max and if τ max < , then lim t τ max u ( x , t ) = . Furthermore, u ( x , t , ϕ ) is a classical solution of (1) for all t [ 0 , τ max ) .

Proof

For ϕ = ( ϕ 1 , ϕ 2 , ϕ 3 , ϕ 4 , ϕ 5 ) X + and h 0 , it follows that

ϕ + h Z ( ϕ ) = ϕ 1 + h [ ( 1 p ( x ) ) Λ ( x ) β ( x ) f ( ϕ 1 , ϕ 5 ) + η ( x ) ϕ 2 ] ϕ 2 + h [ p ( x ) Λ ( x ) σ β ( x ) g ( ϕ 2 , ϕ 5 ) ] ϕ 3 + h [ θ β f ( ϕ 1 , ϕ 5 ) + σ β ( x ) g ( ϕ 2 , ϕ 5 ) ] ϕ 4 + h [ ( 1 θ ) β ( x ) f ( ϕ 1 , ϕ 5 ) + σ ϕ 3 ] ϕ 5 + h [ γ A ( x ) ϕ 3 + γ S ( x ) ϕ 4 ] ϕ 1 h β ( x ) f ( ϕ 1 , ϕ 5 ) ϕ 2 h σ β ( x ) g ( ϕ 2 , ϕ 5 ) ϕ 3 ϕ 4 ϕ 5 ,

which means

lim h 0 + 1 h dist ( ϕ + h Z ( ϕ ) , X + ) = 0 , ϕ X + .

It yields from the general results in [21, Corollary 4] that (1) admits positive solution u ( x , t ) in t [ 0 , τ max ) , which is unique. The proof is completed.□

Lemma 2

For any ϕ X + , model (1) has a unique nonnegative solution on D × [ 0 , ) , and satisfies u ( x , 0 , ϕ ) = ϕ . Furthermore, the semiflow Φ ( t ) induced by the solution of (1) is ultimately bounded.

Proof

From Lemma 1, we learn that model (1) has a unique solution u ( x , t ) defined for x D ¯ and t [ 0 , τ max ) . Then, if proves the global existence of the solution, i.e., the existence interval of the local solution can be extended to infinity. Suppose τ max < , then from [21, Theorem 2], we have u ( x , t ) X as t τ max . It follows easily from the second equation of (1) that

V t D 2 Δ V + p ( x ) Λ ( x ) ( μ ( x ) + η ( x ) ) V , t [ 0 , τ max ) , x D , V ν = 0 , t [ 0 , τ max ) , x D .

Combining [17, Lemma 1] and the comparison theorem leads to

(4) limsup t τ max V ( x , t ) W ¯ 2 ( x ) uniformly for x D ¯ ,

where W ¯ 2 ( x ) satisfies

V ¯ t = D 2 Δ V ¯ + p ( x ) Λ ( x ) ( μ ( x ) + η ( x ) ) V ¯ , t [ 0 , τ max ) , x D , V ¯ ν = 0 , t [ 0 , τ max ) , x D .

Thus, there exists positive constant M 1 > 0 , depending on initial value, and t 1 = t 1 ( ϕ ) > 0 such that V ( x , t ) M 1 , t t 1 . Similarly, from the S equation of model (1), it can be seen that

S ¯ t = D 1 Δ S ¯ + ( 1 p ( x ) ) Λ ( x ) μ ( x ) S ¯ β ( x ) f ( S ¯ , B ) + η ( x ) W ¯ 2 ( x ) , t [ t 1 , τ max ) x D , S ¯ ν = 0 , t [ t 1 , τ max ) , x D ,

which implies that

(5) limsup t τ max S ( x , t ) W ¯ 1 ( x ) uniformly for x D ¯ ,

where W ¯ 1 ( x ) meets (5). Hence, there exists a positive constant M 0 , depending on initial value, and t 2 = t 2 ( ϕ ) > 0 such that

S ( x , t ) M 0 , t t 2 .

Choosing sufficiently large N 3 , N 4 such that T 3 ( t ) N 3 e κ 3 t , T 4 ( t ) N 4 e κ 4 t , where κ 3 > 0 and κ 4 > 0 represent the principal eigenvalues of D 3 Δ π 3 ( x ) and D 4 Δ π 4 ( x ) , respectively. Let N ˜ 3 = N 3 β * ( θ k 1 M 0 + σ k 2 M 1 ) and N ˜ 4 = ( 1 θ ) k 2 β * M 0 N 4 , then

I A ( x , t ) = T 3 ( t ) I A 0 ( x ) + 0 t T 3 ( t s ) [ θ β ( x ) f ( S ( x , s ) , B ( x , s ) ) + σ β ( x ) g ( V ( x , s ) , B ( x , s ) ) ] d s N 3 e κ 3 t I A 0 ( x ) + N ˜ 3 0 t e κ 3 ( t s ) B ( x , s ) d s , I S ( x , t ) = T 4 ( t ) I S 0 ( x ) + 0 t T 4 ( t s ) [ ( 1 θ ) β ( x ) f ( S ( x , s ) , B ( x , s ) ) + δ ( x ) I A ( x , s ) ] d s N 4 e κ 4 t I S 0 ( x ) + N ˜ 4 0 t e κ 4 ( t s ) B ( x , s ) d s + N 4 δ * 0 t e κ 4 ( t s ) I A ( x , s ) d s ,

for 0 t < τ max . According to the fifth equation of (1), one has

B ( x , t ) e α m t B 0 + γ A * 0 t e α m ( t s ) I A ( x , s ) d s + γ S * 0 t e α m ( t s ) I S ( x , s ) d s ,

where α m = min { κ 3 2 , κ 4 2 , α * } and 0 t < τ max . Combining the aforementioned results, one has

I S ( x , t ) N 4 e κ 4 t I S 0 ( x ) + N ˜ 4 0 t e κ 4 ( t s ) B ( x , s ) d s + δ * N 4 0 t e κ 4 ( t s ) N 3 I A 0 ( x ) + N ˜ 3 0 s e κ 3 ( s r ) B ( x , r ) d r d s N 4 e κ 4 t I S 0 ( x ) + N ˜ 4 0 t e κ 4 ( t s ) B ( x , s ) d s + δ * N 3 N 4 I A 0 ( x ) κ 4 + δ * N ˜ 3 N 4 0 t e α m ( t s ) r t e α m ( s r ) B ( x , s ) d s d r N 4 e κ 4 t I S 0 ( x ) + N ˜ 4 0 t e κ 4 ( t s ) B ( x , s ) d s + δ * N 3 N 4 I A 0 ( x ) κ 4 + δ * N ˜ 3 N 4 e 2 α m t ( t r ) 0 t e 2 α m s B ( x , s ) d s ,

for 0 t < τ max . Similarly, one has

I A ( x , t ) N 3 e κ 3 t I A 0 ( x ) + N ˜ 3 0 t e κ 3 ( t s ) e α m s B 0 + γ A * 0 s e α m ( s r ) I A ( x , r ) d r + γ S * 0 s e α m ( s r ) I S ( x , r ) d r d s C 1 + C 3 e α m t 0 t e α m s I A ( x , s ) d s + C 2 e α m t 0 t e α m s I S ( x , s ) d s ,

where C 1 = N 1 I A 0 + N ˜ 1 B 0 ( κ 3 α m ) and C 2 = N ˜ 3 γ A * ( κ 3 α m ) , C 3 = γ S * C 2 and 0 t < τ max . By Gronwall’s inequality, one has

(6) I A ( x , t ) C 1 + C 3 e α m t 0 t e α m s I S ( x , s ) d s e C 2 t , 0 t < τ max .

Then,

I S ( x , t ) N 3 I S 0 ( x ) + δ * N 3 N 4 I A 0 ( x ) κ 4 + N ˜ 4 0 t e κ 4 ( t s ) × e α m s B 0 ( x ) + γ A * 0 s e α m ( s r ) I A ( x , r ) d r + γ S * 0 s e α m ( s r ) I S ( x , r ) d r d s + δ * N ˜ 3 N 4 e 2 α m t ( t r ) 0 t e α m s e α m s B 0 ( x ) + γ A * 0 s e α m ( s r ) I A ( x , r ) d r + γ S * 0 s e α m ( s r ) I S ( x , r ) d r d s C 4 + C 5 e α m t 0 t e α m I S ( x , s ) d s ,

where C 4 = N 4 I S 0 ( x ) + δ * N 3 N 4 I A 0 ( x ) κ 4 + N ˜ 4 B 0 ( x ) ( κ 4 α m ) + δ * N ˜ 3 N 4 ( t r ) B 0 ( x ) α m + N ˜ 4 C 2 e ( C 2 + α m ) t ( C 2 + α m ) , C 5 = ( γ A * + γ S * ) [ N ˜ 4 C 3 e C 2 t C 2 + δ * N ˜ 3 N 4 ( t r ) ] and 0 t < τ max . Using Gronwall’s inequality again yields

(7) I S ( x , t ) C 4 e C 5 t , 0 t < τ max .

The arrangement (6) can be derived as follows:

(8) I A ( x , t ) C 1 e C 2 t + C 3 C 4 e ( C 2 + C 5 ) t ( α m + C 5 ) , 0 t < τ max .

Combining (7) and (8), we obtain

B ( x , t ) B 0 ( x ) + γ A * C 1 e C 2 t ( α m + C 2 ) + γ A * C 3 C 4 e ( C 2 + C 5 ) t ( ( α m + C 2 + C 5 ) ( α m + C 5 ) ) + γ S * C 4 e C 5 t ( α m + C 5 ) , 0 t < τ max .

These, together with Lemma 1, we realize that model (1) has a unique nonnegative global solution, i.e., τ max = . In the next step we prove the boundedness of the solutions of model (1). First, we claim that ( I A ( x , t ) , I S ( x , t ) , B ( x , t ) ) meets the following L 1 -bounded estimate,

(9) limsup t ( I A ( x , t ) L 1 + I S ( x , t ) L 1 + B ( x , t ) L 1 ) M 2 .

On the basis of the model (1), we have

d d t ( S ( x , t ) L 1 + V ( x , t ) L 1 + I A ( x , t ) L 1 + I S ( x , t ) L 1 + B ( x , t ) L 1 ) D Λ ( x ) d x μ ( x ) D ( S ( x , t ) + V ( x , t ) + I A ( x , t ) + I S ( x , t ) ) d x Λ * D μ * ( S ( x , t ) L 1 + V ( x , t ) L 1 + I A ( x , t ) L 1 + I S ( x , t ) L 1 ) .

Hence, we obtain

(10) limsup t ( S ( x , t ) L 1 + V ( x , t ) L 1 + I A ( x , t ) L 1 + I S ( x , t ) L 1 ) M 2 ,

where M 2 = Λ * D μ * . Similarly, B equation satisfies

t D B ( x , t ) d x ( γ A * + γ S * ) M 2 α * D B ( x , t ) d x ,

it follows that limsup t B ( x , t ) L 1 M 3 , where M 3 = ( γ A * + γ S * ) M 2 α * .

In the next step, we show that I A , I S and B fulfill L 2 k -bounded estimate for k 0 ,

(11) limsup t ( I A ( x , t ) 2 k 2 k + I S ( x , t ) 2 k 2 k + B ( x , t ) 2 k 2 k ) M 2 k , for some M 2 k > 0 .

It is apparent that (11) holds when k = 0 . Suppose that for k 1 , (11) also holds, i.e., there is constant M 2 k 1 > 0 such that

limsup t ( I A ( x , t ) 2 k 1 2 k 1 + I S ( x , t ) 2 k 1 2 k 1 + B ( x , t ) 2 k 1 2 k 1 ) M 2 k 1 .

Multiplying the third equation of (1) by I A 2 k 1 and integrating over D , one has

(12) 1 2 k t D I A 2 k d x D 3 D I A 2 k 1 Δ I A d x + D θ β ( x ) I A 2 k 1 f ( S , B ) d x + D σ β ( x ) I A 2 k 1 g ( V , B ) d x D ( μ ( x ) + r A ( x ) + δ ( x ) ) I A 2 k d x .

Since

D 3 D I A 2 k 1 Δ I A d x D 3 D I A I A 2 k 1 d x = ( 2 k 1 ) D 3 D ( I A I A ) I A 2 k 2 d x = 2 k 1 2 2 k 2 D 3 D I A 2 k 1 2 d x .

Thus, (12) becomes

1 2 k t D I A 2 k d x 2 k 1 2 2 k 2 D 3 D I 2 k 1 2 d x + D θ β ( x ) I A 2 k 1 f ( S , B ) d x + D σ β ( x ) I A 2 k 1 g ( V , B ) d x D ( μ ( x ) + r A ( x ) + δ ( x ) ) I A 2 k d x .

According to the boundedness of S and V , one can see that there exists t 3 > 0 such that

D θ β ( x ) I A 2 k 1 f ( S , B ) d x θ β * ( M 0 + 1 ) D I A 2 k 1 B d x , for t t 3 , D σ β ( x ) I A 2 k 1 g ( V , B ) d x σ β * ( M 1 + 1 ) D I A 2 k 1 B d x , for t t 3 .

By using Young’s inequality, a b ε a p + ε q p b q , where a , b , ε > 0 , p , q > 1 and 1 p + 1 q = 1 . One can estimate D B I A 2 k 1 d x by setting ε 0 = α * ( 4 β * max { θ ( M 0 + 1 ) , σ ( M 1 + 1 } ) ) , p = 2 k and q = 2 k ( 2 k 1 ) as follows:

D B I A 2 k 1 d x ε 0 D B 2 k d x + C ε 0 D I A 2 k d x , where C ε 0 = ε 0 1 2 k 1 .

Hence, (12) can be estimated by

(13) 1 2 k t D I A 2 k d x G k D I A 2 k 1 2 d x + α * 4 D B 2 k d x + C k 1 D I A 2 k d x ,

where G k = ( 2 k 1 ) 2 2 k 2 , C k 1 = β * C ε 0 ( θ ( M 0 + 1 ) + σ ( M 1 + 1 ) ) .

Similarly, we multiply the fourth equation of (1) by I S 2 k 1 and integrating over D , one has

1 2 k t D I S 2 k d x D 4 D I S 2 k 1 Δ I S d x + D ( 1 θ ) β ( x ) I S 2 k 1 f ( S , B ) d x + D δ ( x ) I A I S 2 k 1 d x D ( μ ( x ) + r S ( x ) ) I S 2 k d x .

Since

D 4 D I S 2 k 1 Δ I S d x D 4 D I S I S 2 k 1 d x = ( 2 k 1 ) D 4 D ( I S I S ) I S 2 k 2 d x = L k D I S 2 k 1 2 d x ,

where L k = ( 2 k 1 ) D 3 2 2 k 2 and

D ( 1 θ ) β ( x ) I S 2 k 1 f ( S , B ) d x ( 1 θ ) β * ( M 0 + 1 ) D B I S 2 k 1 d x , D δ ( x ) I A I S 2 k 1 d x δ * D I A I S 2 k 1 d x .

By again using Young’s inequality by setting ε 1 = α * ( 4 ( 1 θ ) β * ( M 0 + 1 ) ) , p = 2 k and q = 2 k ( 2 k 1 ) , we obtain

D B I S 2 k 1 d x ε 1 D B 2 k d x + C ε 1 D I S 2 k d x , where C ε 1 = ε 1 1 2 k 1 .

Similarly, we also let ε 2 = 1 δ * , p = 2 k and q = 2 k ( 2 k 1 ) , it follows that

D I A I S 2 k 1 d x ε 2 D I A 2 k d x + C ε 2 D I S 2 k d x , where C ε 2 = ε 2 1 2 k 1 .

Hence, we can obtain

(14) 1 2 k t D I S 2 k d x L k D I S 2 k 1 2 d x + α * 4 D B 2 k d x + D I A 2 k d x + C k 2 D I S 2 k d x ,

where C k 2 = ( 1 θ ) β * ( M 0 + 1 ) C ε 1 + δ * C ε 2 .

Multiplying the fifth equation of (1) by B 2 k 1 and integrating over D , one has

(15) 1 2 k t D B 2 k d x D ( γ S ( x ) I S + γ A ( x ) I A ) B 2 k 1 d x D α ( x ) B 2 k d x γ S * D I S B 2 k 1 d x + γ A * D I A B 2 k 1 d x α * D B 2 k d x .

By again applying Young’s inequality by choosing ε 3 = α * 8 γ S * , ε 4 = α * 8 γ A * , p = 2 k ( 2 k 1 ) , q = 2 k , we have

γ S * D I S B 2 k 1 d x α * 8 D B 2 k d x + C ε 3 D I S 2 k d x , for C ε 3 = ε 3 1 2 k 1 , r A * D I A B 2 k 1 d x α * 8 D B 2 k d x + C ε 4 D I A 2 k d x , for C ε 4 = ε 4 1 2 k 1 .

Thus, (15) becomes

(16) 1 2 k t D B 2 k d x 3 4 α * D B 2 k d x + C ε 3 D I S 2 k d x + C ε 4 D I A 2 k d x .

By combining (13), (14), and (16), we have

1 2 k t D ( I A 2 k + I S 2 k + B 2 k ) d x G k D I A 2 k 1 2 d x L k D I S 2 k 1 2 d x + Q k D I A 2 k d x + W k D I S 2 k d x α * 4 D B 2 k d x ,

where Q k = ( C k 1 + 1 + C ε 4 ) and W k = C k 2 + C ε 3 .

By applying interpolation inequality

ξ 2 2 ε ξ 2 2 + C ε ξ 1 2 , where ξ W 1 , 2 ( D ) .

Let ε 5 = G k 2 Q k , ξ = E 2 k 1 , ε 6 = L k 2 W k and ξ = I 2 k 1 , then

G k D I A 2 k 1 2 d x 2 Q k D I A 2 k d x + 2 Q k C ε 5 D I A 2 k 1 d x 2 , L k D I S 2 k 1 2 d x 2 W k D I S 2 k d x + 2 W k C ε 6 D I S 2 k 1 d x 2 .

Hence, we can obtain

1 2 k t D ( I A 2 k + I S 2 k + B 2 k ) m * D ( I A 2 k + I S 2 k + B 2 k ) d x + 2 Q k C ε 5 D I A 2 k 1 d x 2 + 2 W k C ε 6 D I S 2 k 1 d x 2 ,

where m * = min { Q k , W k , α * 4 } . Then from (11), we can see that limsup t ( I A ( x , t ) 2 k 1 2 k 1 ) M 2 k 1 and limsup t ( I S ( x , t ) 2 k 1 2 k 1 ) M 2 k 1 . Hence,

limsup t ( I A ( t , x ) 2 k + I S ( t , x ) 2 k + B ( t , x ) 2 k ) M 2 k ,

where U = 2 Q k C ε 5 + 2 W k C ε 6 and M 2 k = U M 2 k 1 2 m * . Thus, by the continuous embedding, L q ( D ) L p ( D ) for q p 1 yields

limsup t ( I A ( t , x ) L p + I S ( t , x ) L p + B ( t , x ) L p ) M p , where M p > 0 .

Denote by Y a   ( 0 a 1 ) the fractional power space. Similar to [39, Lemma 2.4], one obtains Y a C ( D ¯ ) by selecting p > n 2 and a n 2 p . Hence, we can obtain

limsup t I A ( x , t ) M , limsup t I S ( x , t ) M , limsup t B ( t , x ) γ A * + γ S * α * M .

where M > 0 . Thus, this completes the proof.□

As the fifth equation of model (1) has no diffusion, the weak compactness of the solution semiflow Φ ( t ) is hard to obtain, and we substitute the weak compactness with the asymptotic smoothness of the solution semiflow. First, we define the Kuratowski measure of noncompactness, τ ( ) ,

τ ( W ) inf { r : W has a finite cover of diameter < r } ,

for any bounded set W .

Lemma 3

For any t 0 , Φ ( t ) admits a global compact attractor A in X + .

Proof

Let ( I A , I S , B ) = γ A ( x ) I A + γ S ( x ) I S α ( x ) B . Taking the partial derivative of ( I A , I S , B ) with respect to B yields

( I A , I S , B ) B = α ( x ) α * , ( I A , I S , B ) X + .

Let u 0 = ( S 0 ( x ) , V 0 ( x ) , I A 0 ( x ) , I S 0 ( x ) , B 0 ( x ) ) X + and u ( x , t ) = ( S ( x , t , u 0 ) , V ( x , t , u 0 ) , I A ( x , t , u 0 ) , I S ( x , t , u 0 ) , B ( x , t , u 0 ) ) be the solution of model (1). It is known by a similar argument as in [39, Lemma 2.5] that for any t 0 , the following sets

S 1 = 0 t e α ( x ) ( t s ) ( γ A ( x ) I A ( x , s , u 0 ) + γ S ( x ) I S ( x , s , u 0 ) ) d s : u 0 W

is precompact. Hence, we rewrite Φ ( t ) = Φ 1 ( t ) + Φ 2 ( t ) , t 0 , where

Φ 1 ( t ) u 0 = { S ( x , t , u 0 ) , V ( x , t , u 0 ) , I A ( x , t , u 0 ) , I S ( x , t , u 0 ) , S 1 } and Φ 2 ( t ) u 0 = { 0 , 0 , 0 , 0 , e α ( x ) t B 0 ( x ) } .

Obviously, τ ( Φ 1 ( t ) W ) = 0 . We then estimate Φ 2 ( t ) as follows:

Φ 2 ( t ) = sup u 0 X Φ 2 ( t ) u 0 X u 0 X e α * t sup u 0 X u 0 X u 0 X = e α * t .

It then follows that for t 0

τ ( Φ ( t ) W ) Φ 2 ( t ) τ ( B ) e α * t τ ( B ) .

Therefore, Φ ( t ) is a τ -contraction with the contraction function e α * t . By [13, Lemma 2.3.4], Φ ( t ) is asymptotically smooth. On the basis of Lemma 3 and [13, Theorem 2.4.6], one can obtain the existence of compact attarctor for model (1) in X + .□

4 Basic reproduction number and steady states

Model (1) admits a disease free steady-state 0 = ( S 0 ( x ) , V 0 ( x ) , 0 , 0 , 0 ) . The linearized subsystem of (1) at 0 is

(17) I A t = D 3 Δ I A + θ β ( x ) f 2 ( S 0 ( x ) , 0 ) B + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) B ( μ ( x ) + r A ( x ) + δ ( x ) ) I A , t > 0 , x D , I S t = D 4 Δ I S + ( 1 θ ) β ( x ) f 2 ( S 0 ( x ) , 0 ) B + δ ( x ) I A ( μ ( x ) + r S ( x ) ) I S , t > 0 , x D , B t = γ S ( x ) I S + γ A ( x ) I A α ( x ) B , t > 0 , x D , I A ν = I S ν = 0 , t > 0 , x D .

Define by T ( t ) the solution of (17), that is, T ( t ) ϕ = ( I A ( , t , ϕ ) , I S ( , t , ϕ ) , B ( , t , ϕ ) ) , where ϕ C ( D ¯ , R 3 ) . Since (17) is cooperative, T ( t ) is a positive C 0 -semigroup with generator + V , where

( x ) = 0 0 θ β ( x ) f 2 ( S 0 ( x ) , 0 ) + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) 0 0 ( 1 θ ) β ( x ) f ( S 0 ( x ) , 0 ) 0 0 0 ,

and

V ( x ) = D 3 Δ ( μ ( x ) + r A ( x ) + δ ( x ) ) 0 0 δ ( x ) D 4 Δ ( μ ( x ) + r S ( x ) ) 0 γ A ( x ) γ S ( x ) α ( x ) .

We denote T ˜ ( t ) be the positive semigroup generated V ( x ) . By [29, Theorem 3.12], the next generator operator is defined by

L ( ϕ ) ( x ) = 0 ( x ) T ˜ ( t ) ϕ ( x ) d t = ( x ) 0 T ˜ ( t ) ϕ ( x ) d t , ϕ C ( D ¯ , R 3 ) , x D ¯ .

According to [32], we can define the spectral radius of L as the basic reproduction number as follows:

0 r ( L ) = r ( V 1 ) .

where ( ) 1 denotes the inverse of ( ) . By computing the expression of V 1 ( x ) , one finds

V 1 ( x ) = ( D 3 Δ π 3 ( x ) ) 1 0 0 Q 21 ( x ) ( D 4 π 4 ( x ) ) 1 0 Q 31 ( x ) γ S ( x ) ( α ( x ) ) 1 ( D 4 Δ π 4 ( x ) ) 1 ( α ( x ) ) 1 ,

where

Q 21 ( x ) = δ ( x ) ( D 3 Δ π 3 ( x ) ) 1 ( D 4 Δ π 4 ( x ) ) 1 , ( x ) = θ β ( x ) f 2 ( S 0 ( x ) , 0 ) + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) , Q 31 ( x ) = ( α ( x ) ) 1 ( γ A ( x ) ( D 3 Δ π 3 ( x ) ) 1 γ S ( x ) Q 21 ( x ) ) , J ( x ) = ( 1 θ ) β ( x ) f 2 ( S 0 ( x ) , 0 ) .

Hence, combining the expressions of ( x ) and V 1 ( x ) yields

( x ) V 1 ( x ) = ( x ) Q 31 ( x ) γ S ( x ) ( α ( x ) ) 1 ( D 4 Δ π 4 ( x ) ) 1 ( x ) ( α ( x ) ) 1 ( x ) J ( x ) Q 31 ( x ) γ S ( x ) ( α ( x ) ) 1 ( D 4 Δ π 4 ( x ) ) 1 J ( x ) ( α ( x ) ) 1 J ( x ) 0 0 0 .

Therefore, we have

0 r ( L ) = r ( ( ( x ) Q 31 ( x ) + γ S ( x ) ( α ( x ) ) 1 ( D 4 Δ π 4 ( x ) ) 1 J ( x ) ) ) .

This indicates that 0 is the principal eigenvalue of the below eigenvalue problem

( ( x ) Q 31 ( x ) + γ S ( x ) ( α ( x ) ) 1 ( D 4 Δ π 4 ( x ) ) 1 J ( x ) ) ϕ = λ ϕ , x D , ϕ ( x ) ν = 0 , x D .

So there exists a strictly positive eigenfunction ϕ * such that

( ( x ) Q 31 ( x ) + γ S ( x ) ( α ( x ) ) 1 ( D 4 Δ π 4 ( x ) ) 1 J ( x ) ) ϕ * = 0 ϕ * , x D , ϕ * ( x ) ν = 0 , x D ,

which is equivalent to

(18) ( θ β ( x ) f 2 ( S 0 ( x ) , 0 ) + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) ( α ( x ) ) 1 ( γ A ( x ) ( D 3 Δ π 3 ( x ) ) 1 γ S ( x ) δ ( x ) ( D 3 Δ π 3 ( x ) ) 1 ( D 4 Δ π 4 ( x ) ) 1 ) + γ S ( x ) ( α ( x ) ) 1 × ( D 4 Δ π 4 ( x ) ) 1 ( 1 θ ) β ( x ) f 2 ( S 0 ( x ) , 0 ) ) ϕ * = 0 ϕ * , x D , ϕ * ( x ) ν = 0 , x D .

Let’s further consider the following eigenvalue problem

(19) ( θ β ( x ) f 2 ( S 0 ( x ) , 0 ) + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) × ( α ( x ) ) 1 γ A ( x ) ( D 3 Δ π 3 ( x ) ) 1 ) ψ = λ ˜ ψ , x D , ψ ( x ) ν = 0 , x D ,

where λ ˜ is the eigenvalue and associated with the positive eigenfunction ψ . It is known from [39] that the problem (19) has a principal eigenvalue λ ˜ * and corresponds to a positive eigenfunction ψ * satisfying

(20) ( θ β ( x ) f 2 ( S 0 ( x ) , 0 ) + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) × ( α ( x ) ) 1 γ A ( x ) ( D 3 Δ π 3 ( x ) ) 1 ) ψ * = λ ˜ * ψ * , x D , ψ * ( x ) ν = 0 , x D .

Multiplying the first equation of the problem (20) by ψ * and integrating over D yields

(21) D γ A ( x ) ( α ( x ) ) 1 ( θ β ( x ) f 2 ( S 0 ( x ) , 0 ) + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) ) ψ * 2 d x = λ ˜ * D ( D 3 ψ * 2 + π 3 ( x ) ψ * 2 ) d x .

This shows that

(22) λ ˜ * = D γ A ( x ) ( θ β ( x ) f 2 ( S 0 ( x ) , 0 ) + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) ) ψ * 2 d x D α ( x ) ( D 3 ψ * 2 + π 3 ( x ) ψ * 2 ) d x .

In addition, we consider the following eigenvalue problem

(23) γ S ( x ) ( α ( x ) ) 1 ( D 4 Δ π 4 ( x ) ) 1 ( ( θ β ( x ) f 2 ( S 0 ( x ) , 0 ) + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) ) × δ ( x ) ( D 3 Δ π 3 ( x ) ) 1 ( 1 θ ) β ( x ) f 2 ( S 0 ( x ) , 0 ) ) φ = λ ¯ φ , x D , φ ( x ) ν = 0 , x D ,

which is equivalent to

γ S ( x ) ( α ( x ) ) 1 ( ( θ β ( x ) f 2 ( S 0 ( x ) , 0 ) + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) ) δ ( x ) ( D 3 Δ π 3 ( x ) ) 1 ( 1 θ ) β ( x ) f 2 ( S 0 ( x ) , 0 ) ) φ = λ ¯ ( D 4 Δ π 4 ( x ) ) φ , x D , φ ( x ) ν = 0 , x D .

Similar to the method of (21), one derives

(24) λ ¯ * = D γ S ( x ) ( 1 θ ) β ( x ) f 2 ( S 0 ( x ) , 0 ) φ * 2 d x D α ( x ) ( D 4 φ * 2 + π 4 ( x ) φ * 2 ) d x D γ S ( x ) δ ( x ) ( D 3 Δ π 3 ( x ) ) 1 ( θ β ( x ) f 2 ( S 0 ( x ) , 0 ) + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) ) φ * 2 d x D α ( x ) ( D 4 φ * 2 + π 4 ( x ) φ * 2 ) d x .

Next, consider the following eigenvalue problem

( γ S ( x ) δ ( x ) ( D 3 Δ π 3 ( x ) ) 1 ( θ β ( x ) f 2 ( S 0 ( x ) , 0 ) + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) ) ) χ = λ ˆ χ , x D , χ ( x ) ν = 0 , x D ,

where λ ˆ is the eigenvalue and χ is the corresponding eigenfunction. Similarly, one can also conclude that

(25) λ ˆ * = D γ S ( x ) δ ( x ) ( θ β ( x ) f 2 ( S 0 ( x ) , 0 ) + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) ) χ * 2 d x D ( D 3 χ * 2 + π 3 ( x ) χ * 2 ) d x .

From (22)–(25) and multiplying the first equation of problem (18) by ψ * φ * χ * , one obtains

0 = D γ S ( x ) ( 1 θ ) β ( x ) f 2 ( S 0 ( x ) , 0 ) ψ * 2 ϕ * φ * χ * d x D α ( x ) ( D 4 ψ * 2 + π 4 ( x ) ψ * 2 ) d x + D ( θ β ( x ) f 2 ( S 0 ( x ) , 0 ) + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) ) d x D α ( x ) ( D 3 φ * 2 + π 3 ( x ) φ * 2 ) d x × D γ A ( x ) φ * 2 ϕ * ψ * χ * d x + D γ S ( x ) δ ( x ) χ * 2 φ * 2 ϕ * ψ * d x D ( D 4 χ * 2 + π 4 ( x ) χ * 2 ) d x , D ϕ * ψ * φ * χ * d x = 1 .

Therefore, we finally obtain

0 = sup ϕ , ψ , φ , χ H 1 ( D ) , ϕ , ψ , φ , χ 0 D γ S ( x ) ( 1 θ ) β ( x ) f 2 ( S 0 ( x ) , 0 ) ψ 2 ϕ φ χ d x D α ( x ) ( D 4 ψ 2 + π 4 ( x ) ψ 2 ) d x + D ( θ β ( x ) f 2 ( S 0 ( x ) , 0 ) + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) ) φ 2 d x D α ( x ) ( D 3 φ 2 + π 3 ( x ) φ 2 ) d x × D γ A ( x ) ϕ ψ χ d x + D γ S ( x ) δ ( x ) χ 2 ϕ ψ d x D ( D 4 χ 2 + π 4 ( x ) χ 2 ) d x .

Note that in case the diffusion coefficients are all 0, i.e., D i = 0   ( i = 1 , , 4 ) , the local reproduction number ˜ 0 ( x ) is denoted in the following form

˜ 0 ( x ) = γ S ( x ) ( 1 θ ) f 2 ( S 0 ( x ) , 0 ) α ( x ) ( μ ( x ) + r S ( x ) ) + θ β ( x ) f 2 ( S 0 ( x ) , 0 ) + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) α ( x ) ( μ ( x ) + r A ( x ) + δ ( x ) ) γ A ( x ) + γ S ( x ) δ ( x ) μ ( x ) + r S ( x ) = 1 ( x ) + 2 ( x ) + 3 ( x ) ,

where

1 ( x ) = γ S ( x ) β ( x ) ( 1 θ ) f 2 ( S 0 ( x ) , 0 ) α ( x ) ( μ ( x ) + r S ( x ) ) , 2 ( x ) = γ A ( x ) β ( x ) ( θ f 2 ( S 0 ( x ) , 0 ) + σ g 2 ( V 0 ( x ) , 0 ) ) α ( x ) ( μ ( x ) + r A ( x ) + δ ( x ) ) , 3 ( x ) = γ S ( x ) δ ( x ) β ( x ) ( θ f 2 ( S 0 ( x ) , 0 ) + σ g 2 ( V 0 ( x ) , 0 ) ) α ( x ) ( μ ( x ) + r S ( x ) ) ( μ ( x ) + r A ( x ) + δ ( x ) ) .

For the relation between 0 and ˜ 0 ( x ) , it holds by the following conclusions.

Remark 1

The following facts were established.

  1. If j ( x )   ( j = 1 , 2 , 3 ) are constants, then 0 = ˜ 0 ( x ) .

  2. If min 1 j 3 { j ( x ) } > 1 , then ˜ 0 ( x ) > 1 holds.

By substituting I A ( x , t ) = e λ t ψ 3 ( x ) , I S ( x , t ) = e λ t ψ 4 ( x ) , and B ( x , t ) = e λ t ψ 5 ( x ) into model (17), we obtain the eigenvalue problem as follows:

(26) λ ψ 3 ( x ) = D 3 Δ ψ 3 ( x ) + θ β ( x ) f 2 ( S 0 ( x ) , 0 ) ψ 5 ( x ) + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) ψ 5 ( x ) ( μ ( x ) + r A ( x ) + δ ( x ) ) ψ 3 ( x ) , x D , λ ψ 4 ( x ) = D 4 Δ ψ 4 ( x ) + ( 1 θ ) β ( x ) f 2 ( S 0 ( x ) , 0 ) ψ 5 ( x ) + δ ( x ) ψ 3 ( x ) ( μ ( x ) + r S ( x ) ) ψ 4 ( x ) , x D , λ ψ 5 ( x ) = γ A ( x ) ψ 3 ( x ) + γ S ( x ) ψ 4 ( x ) α ( x ) ψ 5 ( x ) , x D , ψ 3 ( x ) ν = ψ 4 ( x ) ν = 0 , x D .

According to [29] and [32, Lemma 2.2], it is straightforward to derive the following result.

Lemma 4

0 1 has the same sign s ( ) , where s ( ) = sup { λ , λ σ ( L ) } is the spectral bounded of .

Lemma 5

If 0 1   ( s ( ) 0 ) , then s ( ) is the principal eigenvalue of the problem (26) associated with a strongly positive eigenfunction.

Proof

From (17), one can derive

(27) I A ( x , t , ϕ ) = T 3 ( t ) ϕ ( t ) + 0 t T 3 ( t s ) P 1 ( B ( x , s , ϕ ) ) d s , I S ( x , t , ϕ ) = T 4 ( t ) ϕ ( t ) + 0 t T 4 ( t s ) P 2 ( I A ( x , s , ϕ ) , B ( x , s , ϕ ) ) d s , B ( x , t , ϕ ) = T 5 ( t ) ϕ ( t ) + 0 t T 5 ( t s ) ( γ A ( x ) I A ( x , s , ϕ ) + γ S ( x ) I S ( x , s , ϕ ) ) d s ,

where P 1 ( B ) = θ β 1 ( x ) f 2 ( S 0 ( x ) , 0 ) B + σ β 2 ( x ) g 2 ( V 0 ( x ) , 0 ) B and P 2 ( I A , B ) = ( 1 θ ) β 1 ( x ) f 2 ( S 0 ( x ) , 0 ) B + δ ( x ) I A .

Let T ˜ ( t ) as T ˜ ( t ) = T ˜ 3 ( t ) + T ˜ 4 ( t ) , where T ˜ 3 ( t ) ϕ = ( 0 , 0 , Γ 5 ( t ) ϕ 5 ) and

T ˜ 4 ( t ) ϕ = ( I A ( x , t , ϕ ) , I S ( x , t , ϕ ) , 0 t T 5 ( t s ) ( γ A ( x ) I A ( x , s , ϕ ) + γ S ( x ) I S ( x , s , ϕ ) ) d s ) .

According to [39, Lemma 2.5], one has T ˜ 4 ( t ) is compact. From Lemma 3, T ˜ ( t ) is the τ -contraction on, i.e., the essential growth bound ω ess ( T ˜ ( t ) ) = lim t ln ϑ ( T ( t ) ) t , ϑ is the measure of non-compactness ω ess ( T ˜ ( t ) ) α * and the essential spectral radius r ess ( T ˜ ( t ) ) e α * t < 1 , t > 0 . It is well known [11] that

ω ( T ˜ ( t ) ) = max { s ( ) , ω ess ( T ˜ ( t ) ) } ,

where ω ( T ˜ ( t ) ) is the exponential growth bound of T ˜ ( t ) defined as ω ( T ˜ ( t ) ) = lim t ln T ˜ ( t ) t . Under the assumption that s ( ) 0 , the spectral radius of T ˜ satisfies

r ( T ˜ ( t ) ) = e s ( ) t 1 , t > 0 .

As a result, r ess ( T ˜ ( t ) ) < r ( T ˜ ( t ) ) , t > 0 . By the aid of the generalized Krein-Rutman theorem (see [25, lemma 2.2], we can conclude that s ( ) is the principal eigenvalue of (26), which is associated with a strong positive eigenfunction.□

5 Stability of steady states

This section focuses on the analysis of the threshold dynamics of model (1) on 0 . We start by justifying the stability of 0 when 0 < 1 .

Theorem 1

If 0 < 1   ( s ( ) < 0 ) , then 0 is globally asymptotically stable.

Proof

Similar to [32, Theorem 3.1], it is clear that 0 is locally asymptotically stable. Therefore, we just have to prove the global attraction of 0 . Following Lemma 4 reveals that when 0 < 1 , s ( ( S 0 ( x ) , V 0 ( x ) ) ) < 0 . That means there exists a sufficiently small positive number ε 0 such that s ( ( S 0 ( x ) + ε 0 , V 0 ( x ) ) + ε 0 ) < 0 . According to Theorem 3, S ( x , t ) and V ( x , t ) are ultimately bounded, i.e., there exists t 1 > 0 such that 0 S ( x , t ) S 0 ( x ) + ε 0 and 0 V ( x , t ) V 0 ( x ) + ε 0 , for t t 1 . From the comparison principle [21] yields ( I A ( x , t ) , I S ( x , t ) , B ( x , t ) ) ( I ˆ A ( x , t ) , I ˆ S ( x , t ) , B ˆ ( x , t ) ) on D ¯ × [ 0 , ) , where ( I ˆ A ( x , t ) , I ˆ S ( x , t ) , B ˆ ( x , t ) ) satisfies

(28) I ˆ A ( x , t ) t = D 3 Δ I ˆ A ( x , t ) + θ β ( x ) f 2 ( S 0 + ε 0 , 0 ) B ˆ ( x , t ) + σ β ( x ) g 2 ( V 0 ( x ) + ε 0 , 0 ) B ˆ ( x , t ) ( μ ( x ) + r A ( x ) + δ ( x ) ) I ˆ A ( x , t ) , x D , t t 1 , I ˆ S ( x , t ) t = D 4 Δ I ˆ S ( x , t ) + ( 1 θ ) β ( x ) f 2 ( S 0 ( x ) + ε 0 , 0 ) B ˆ ( x , t ) + δ ( x ) I ˆ A ( x , t ) ( μ ( x ) + r S ( x ) ) I ˆ S ( x , t ) , x D , t t 1 , B ˆ ( x , t ) t = γ A ( x ) I ˆ A ( x , t ) + γ S ( x ) I ˆ S ( x , t ) α ( x ) B ˆ ( x , t ) , x D , t t 1 , I ˆ A ( x , t ) ν = I ˆ S ( x , t ) ν = 0 , x D ¯ , t t 1 , I ˆ A ( x , t ) = I A ( x , t 1 ) , I ˆ S ( x , t ) = I S ( x , t 1 ) , x D .

Choose a constant ι > 0 such that ( I A ( x , t 1 , ϕ ) , I S ( x , t 1 , ϕ ) , B ( x , t 1 , ϕ ) ) ι ( ψ 3 ε 0 ( x ) , ψ 4 ε 0 ( x ) , ψ 5 ε 0 ( x ) ) for x D ¯ , where ( ψ 3 ε 0 ( x ) , ψ 4 ε 0 ( x ) , ψ 5 ε 0 ( x ) ) is the corresponding eigenvector of s ( ( S 0 ( x ) + ε 0 , V 0 ( x ) ) + ε 0 ) and ι e s ( ε 0 ) ( t t 1 ) ( ψ 3 ε 0 ( x ) , ψ 4 ε 0 ( x ) , ψ 5 ε 0 ( x ) ) is a solution of (28). By employing the comparison principle, one has

( I A ( x , t 1 , ϕ ) , I S ( x , t 1 , ϕ ) , B ( x , t 1 , ϕ ) ) ι e s ( ε 0 ) ( t t 1 ) ( ψ 3 ε 0 ( x ) , ψ 4 ε 0 ( x ) , ψ 5 ε 0 ( x ) ) , t t 1 .

Thus, ( I A ( x , t ) , I S ( x , t ) , B ( x , t ) ) ( 0 , 0 , 0 ) as t uniformly for x D ¯ . Moreover, using the first and second equations of (1) yields S ( x , t ) S 0 ( x ) , V ( x , t ) V 0 ( x ) as t uniformly for x D ¯ . This completes the proof.□

Next, we address the uniform persistence of the disease. Before the proof, we present a few lemmas that will be utilized afterward.

Lemma 6

Let ( S ( x , t ) , V ( x , t ) , I A ( x , t ) , I S ( x , t ) , B ( x , t ) ) be the solution of (1) satisfying ϕ = ( S 0 ( x ) , V 0 ( x ) , I A 0 ( x ) , I S 0 ( x ) , B 0 ( x ) ) X + .

  1. For any ϕ X + , then S ( x , t ) > 0 and V ( x , t ) > 0 , and there exists a positive constant τ 0 , independent of ϕ , such that

    liminf t S ( x , t ) > τ 0 , liminf t V ( x , t ) > τ 0 uniformly f o r x D ¯ .

  2. For any ϕ X + , if I A 0 ( x ) 0 or I S 0 ( x ) 0 or B 0 ( x ) 0 , then p ( x , t , ϕ ) > 0 , for x D ¯ , t > 0 , where p = I A , I S , B .

Proof

( i ) From Lemma 3, there is a constant M 4 > 0 such that B ( x , t ) M 4 , x D ¯ , t > 0 . It follows that the assumption ( H 2 ) yields f ( S , B ) k 1 M 4 S and g ( V , B ) k 2 M 4 V . In view of V equation of (1), we have

V t D 2 Δ V ( μ ( x ) + η ( x ) + σ k 2 β ( x ) M 4 ) V .

Hence, V ( x , t ) V ˜ ( x , t ) , where V ˜ ( x , t ) is a solution to the following problem:

(29) V ˜ t = D 2 Δ V ˜ + p ( x ) Λ ( x ) ( μ ( x ) + η ( x ) + σ k 2 β ( x ) M 4 ) V ˜ , x D , t > 0 , V ˜ ν = 0 , x D , t > 0 , V ( x , 0 ) = V ˜ ( x , 0 ) , x D ¯ .

By the the maximum principle, V ˜ ( x , t ) > 0 , V ( x , t ) V ˜ ( x , t ) > 0 for any x D ¯ and t > 0 . Similarly, we can derive S ( x , t ) S ˜ ( x , t ) > 0 , where S ˜ ( x , t ) is a solution to the following problem:

(30) S ˜ t = D 1 Δ S ˜ + ( 1 p ( x ) ) Λ ( x ) ( μ ( x ) + k 1 β ( x ) M 4 ) S ˜ + η V ˜ , x D , t > 0 , S ˜ ν = 0 , x D , S ( x , 0 ) = S ˜ ( x , 0 ) , x D ¯ .

Note that problem (29) and (30) possesses a unique steady state, defined by S ˜ 0 ( x ) and V ˜ 0 ( x ) , from the proof of [17, Lemma 1]. Hence,

liminf t S ( x , t ) inf x D ¯ S ˜ 0 ( x ) τ 1 , liminf t V ( x , t ) inf x D ¯ V ˜ 0 ( x ) τ 2 .

Choose τ min { τ 1 , τ 2 } . This concludes ( i ) the proof.

( i i ) For I A 0 ( x ) 0 , from the third equation of model (1), one has

I A t D 3 Δ I A ( μ ( x ) + r A ( x ) + δ ( x ) ) I A , x D , t > 0 , I A ν = 0 , x D , t > 0 .

By [33, Lemma 1.26], it follows that I A ( x , t ) > 0 for all ( x , t ) D ¯ × ( 0 , ) . Further, considering the third equation of model (1) yields

I S t D 4 Δ I S ( μ ( x ) + r S ( x ) ) I S , x D , t > 0 , I S ν = 0 , x D , t > 0 .

Similarly, we can also derive I S ( x , t ) > 0 for ( x , t ) D ¯ × ( 0 , ) . Moreover, for fixed x D ¯ and t 0 > 0 , B equation of (1) yields

B ( x , t ) = e α ( x ) t B ( x , t 0 ) + t 0 t e α ( x ) ( t s ) [ γ A ( x ) I A ( x , s , ϕ ) + γ S ( x ) I S ( x , s , ϕ ) ] d s .

This is, B ( x , t ) > 0 for ( x , t ) D ¯ × [ t 0 , ) . For the cases I S 0 ( x ) 0 or B 0 ( x ) 0 , the proof procedure is similar to I A 0 ( x ) 0 , which we omit here. Thus, Lemma 6 ( i i ) hold.□

Lemma 7

If there exists τ p > 0 such that

liminf t p ( x , t , ϕ ) τ p , for p = I A or I S or B , uniformly f o r x D ,

then there exists σ p > 0 such that

(31) liminf t p ˜ ( x , t , ϕ ) τ p , for p ˜ = S , V , I A , I S , B , uniformly f o r x D .

Proof

In the case of p = I A , it is assumed that there exists t 1 * > 0 such that I A ( x , t ) 1 2 τ p for any x D and t [ t 1 * , ) . Therefore, I S equation of (1) satisfies

(32) I S t D 4 Δ I S + 1 2 τ p δ ( x ) ( μ ( x ) + r S ( x ) ) I S , x D , t t 1 * , I S ν = 0 , x D , t t 1 * .

By the comparison theorem, we learn that (32) there exists positive functions I ˜ S * ( x ) such that

liminf t I S ( x , t ) I ˜ S * ( x ) .

This means that there exists t 2 * > 0 such that I S ( x , t ) I ˜ S * ( x ) 2 , for any x D , t t 2 * . According to the B equation of (1), one has

(33) B t 1 2 τ p γ A ( x ) + 1 2 γ S ( x ) I ˜ S * ( x ) α ( x ) B , x D , t > t 2 * , B ν = 0 , x D , t > t 2 * .

According to the comparison theorem, we learn that (32) there exists positive function B ˜ * ( x ) such that liminf t B ( x , t ) B ˜ * ( x ) . Therefore, the conclusion (31) holds when p = I A ( x , t ) .

In the other p = B case, from Lemma 6 ( i i ) and (33), we learn that there exists t 2 * > 0 such that

S ( x , t ) 1 2 τ p , V ( x , t ) 1 2 τ p , B ( x , t ) 1 2 τ p , x D , t t 2 * .

As a result, the I A equation of (1) fulfills

I A t D 3 Δ I A + θ β * τ p 2 f 2 τ p 2 , τ p 2 + σ β * τ p 2 g 2 τ p 2 , τ p 2 ( μ ( x ) + r A ( x ) + δ ( x ) ) I A , x D , t t 2 * , I A n = 0 , x D , t t 2 * .

By the comparison principle, we obtain liminf t I A ( x , t ) I ˜ A * ( x ) , where I ˜ A * ( x ) is a positive number. Thus, there exists t 3 * > 0 such that I A ( x , t ) I ˜ A * ( x ) 2 , for any x D and t > t 3 * . By using the I S equation from model (1), we can also derive

I S t D 4 Δ I S + 1 2 δ ( x ) I ˜ A * ( x ) ( μ ( x ) + r S ( x ) ) I S , x D , t t 3 * , I S n = 0 , x D , t t 3 * ,

By using the comparison principle again, one has liminf t I S ( x , t ) I ˜ S * ( x ) . In a similar way to the demonstration above, the case p = I S can be verified. This proves (31).□

In the following, we present the uniform persistence of model (1).

Theorem 2

If 0 > 1 , then there exists σ * > 0 such that for ϕ X + with ϕ 3 0 or ϕ 4 0 or ϕ 5 0 , the solution p ˜ = ( S ( x , t , ϕ ) , V ( x , t , ϕ ) , I A ( x , t , ϕ ) , I S ( x , t , ϕ ) , B ( x , t , ϕ ) ) of (1) satisfies

liminf t p ˜ ( x , t , ϕ ) σ * uniformly f o r x D ¯ .

In addition, the model possesses at least one positive steady state.

Proof

According to Lemmas 6 and 7 and the process of [28, Theorem 3], define the set

W 0 = { ϕ X + : ϕ 3 ( x ) 0 } , W 0 = X + \ W 0 = { ϕ X + : ϕ 3 ( x ) 0 } , = { ϕ W 0 : Φ ( t ) ϕ W 0 , t 0 } ,

where Φ ( t ) : X + X + is the semiflow generated by the solution of (1). For any ϕ W 0 , it follows directly from Lemma 7 that I A ( x , t ) > 0 . This indicates that Φ ( t ) W 0 W 0 . Further, we will prove this theorem by the following claims.

  • Claim I: For any ϕ , the ω limit set ω ( ϕ ) is the singleton { 0 } .

For any ϕ , one has Φ ( t ) ϕ 0 , i.e., I A ( x , t , ϕ ) 0 , x D , t 0 . Thus, by I A equation of model (1), one obtains

θ β ( x ) f ( S , B ) + σ β ( x ) g ( V , B ) 0 .

Combining the conditions ( H 1 ) yields B ( x , t ) 0 , x D , t > 0 . Taking the aforementioned leads to the fourth equation of model (1) yields I S ( x , t ) 0 , x D , t > 0 . Further, the first second equations of (1) yields

liminf t S ( x , t , ϕ ) = S 0 ( x ) , liminf t V ( x , t , ϕ ) = V 0 ( x ) uniformly for x D ¯ .

This proves Claim I.

Claim II: limsup t Φ ( t ) ϕ 0 τ * , ϕ W 0 .

By arguing by contradiction, we suppose that there exists ϕ W 0 such that

limsup t Φ ( t ) ϕ 0 < τ * .

So there exists t 1 > 0 such that for any x D ¯ and t t 1 ,

S ( x , t , ϕ ) > S 0 ( x ) τ * , V ( x , t , ϕ ) > V 0 ( x ) τ * , I A ( x , t , ϕ ) < τ * , I S ( x , t , ϕ ) < τ * , B ( x , t , ϕ ) < τ * .

Thus, one has

I A t D 3 Δ I A + θ β ( x ) f 2 ( S 0 ( x ) τ * , τ * ) B + σ β g 2 ( V 0 ( x ) τ * , τ * ) B ( μ ( x ) + r A ( x ) + δ ( x ) ) I A , x D , t t 1 I S t D 4 Δ I S + ( 1 θ ) β ( x ) f 2 ( S 0 ( x ) τ * , τ * ) B + δ ( x ) I A ( μ ( x ) + r S ( x ) ) I S , x D , t t 1 , B t γ A ( x ) I A + γ S ( x ) I S α ( x ) B , x D , t t 1 , I A ν = I S ν = 0 , x D , t t 1 .

Assuming that the linear system

u 3 t = D 3 Δ u 3 + θ β ( x ) f 2 ( S 0 ( x ) τ * , τ * ) u 5 + σ β g 2 ( V 0 ( x ) τ * , τ * ) u 5 ( μ ( x ) + r A ( x ) + δ ( x ) ) u 3 , x D , t t 1 u 4 t = D 4 Δ u 4 + ( 1 θ ) β ( x ) f 2 ( S 0 ( x ) τ * , τ * ) u 5 + δ ( x ) u 3 ( μ ( x ) + r S ( x ) ) u 4 , x D , t t 1 , u 5 t = γ A ( x ) u 3 + γ S ( x ) u 4 α ( x ) u 5 , x D , t t 1 , I A ν = I S ν = 0 , x D , t t 1 .

admits a solution ( u 3 ( x , t ) , u 4 ( x , t ) , u 5 ( x , t ) ) = e λ ( S 0 ( x ) τ * , V 0 ( x ) τ * ) ( t t 1 ) , where ψ ˆ ( x ) = ( ψ ˆ 3 ( x ) , ψ ˆ 4 ( x ) , ψ ˆ 5 ( x ) ) corresponding to λ 0 ( S 0 ( x ) τ * , V 0 ( x ) τ * ) the corresponding eigenvector. Choosing ξ > 0 such that ξ ( ψ ˆ 3 ( x ) , ψ ˆ 4 ( x ) , ψ ˆ 5 ( x ) ) ( I A ( x , t 1 ) , I S ( x , t 1 ) , B ( x , t 1 ) ) and by the comparison principle, we can obtain

( I S ( x , t , ϕ ) , I A ( x , t , ϕ ) , B ( x , t , ϕ ) ) ξ e λ ( S 0 ( x ) σ ˆ * , V 0 ( x ) σ ˆ * ) ( t t 1 ) ψ ˆ ( x ) , for x D , t t 1 .

By λ ( S 0 ( x ) τ * , V 0 ( x ) τ * ) > 0 , lim t ( I A ( x , t , ϕ ) , I S ( x , t , ϕ ) , B ( x , t , ϕ ) ) = ( , , ) ,which contradicts Lemma 2.

Next, we define the continuous function ϱ : X + [ 0 , ) fulfills

ϱ ( ϕ ) ( x ) min x D { ϕ 3 ( x ) } , ϕ X + .

Evidently, ϱ 1 ( 0 , ) W 0 and ϱ possesses the property that either ϱ ( ϕ ) = 0 and ϕ 0 or ϱ ( ϕ ) > 0 , then ϱ ( Φ ( t ) ϕ ) > 0 . Hence, ϱ is a generalized distance function for the semiflow Φ ( t ) : X + X + (see [28]). After the above analysis, we can derive that any forward orbit Φ ( t ) in converges to 0 , and W s ( 0 ) 0 = , where W s ( 0 ) is a stable subset of 0 . Moreover, 0 is an isolated invariant set in X + , and no set of { 0 } forms a circle in 0 . Based on [28, Theorem 3], there exists ι > 0 such that liminf t ϱ ( Φ ( t ) ϕ ) > ι , for ϕ 0 , which means liminf t I A ( x , t , ϕ ) ι for ϕ 0 . It follows from Lemma 7 that model (1) is uniformly persistent and by [20, Theorem 4.7] and Theorem 2, there exists at least one endemic steady state for model (1). This completes the proof.□

Theorem 3

In case 0 = 1 , then 0 is globally asymptotically stable.

Proof

As a start, let us study the local stability of 0 . Suppose that there exists l > 0 such that u 0 0 l , where u 0 = ( S 0 ( x ) , V 0 ( x ) , I A 0 ( x ) , I S 0 ( x ) , B 0 ( x ) ) X + . Define

ω 1 ( x , t ) = S ( x , t ) S 0 ( x ) 1 , ω 2 ( x , t ) = V ( x , t ) V 0 ( x ) 1 , b 1 ( t ) = max x D ¯ { ω 1 ( x , t ) , 0 } , b 2 ( t ) = max x D ¯ { ω 2 ( x , t ) , 0 } .

By D 1 Δ S 0 ( x ) + ( 1 p ( x ) ) Λ ( x ) μ ( x ) S 0 ( x ) + η ( x ) V 0 ( x ) = 0 , D 2 Δ V 0 ( x ) + p ( x ) Λ ( x ) ( μ ( x ) + η ( x ) ) V 0 ( x ) = 0 , one has

ω 1 t D 1 Δ ω 1 2 D 1 S 0 ( x ) ω 1 S 0 ( x ) + ( 1 p ( x ) ) Λ ( x ) + η ( x ) V 0 ( x ) S 0 ( x ) ω 1 = β ( x ) f ( S , B ) S 0 ( x ) + η ( x ) ( V V 0 ( x ) ) S 0 ( x ) , ω 2 t D 2 Δ ω 2 2 D 2 V 0 ( x ) ω 2 V 0 ( x ) + p ( x ) Λ ( x ) V 0 ( x ) ω 2 = σ β ( x ) g ( V , B ) V 0 ( x ) .

Denote T ˜ 1 ( t ) and T ˜ 2 ( t ) be the positive semigroups generated by the below operators, respectively

D 1 Δ + 2 D 1 S 0 ( x ) S 0 ( x ) ( 1 p ( x ) ) Λ ( x ) + η ( x ) V 0 ( x ) S 0 ( x ) , D 2 Δ + 2 D 2 V 0 ( x ) V 0 ( x ) p ( x ) Λ ( x ) V 0 ( x ) .

Then for some M i ( i = 1 , 2 ) , there has ρ ˜ i > 0 such that T ˜ i M i e ρ ˜ i t . Hence,

ω 1 ( x , t ) = T ˜ 1 ( t ) ω 10 ( x ) 0 t T ˜ 1 ( t s ) β ( x ) f ( S ( x , s ) , B ( x , s ) ) S 0 ( x ) η ( x ) ( V ( x , s ) V 0 ( x ) ) S 0 ( x ) d s , ω 2 ( x , t ) = T ˜ 2 ( t ) ω 20 ( x ) 0 t T ˜ 2 ( t s ) σ β ( x ) g ( V ( x , s ) , B ( x , s ) ) V 0 ( x ) d s ,

where ω 10 ( x ) = S 0 ( x ) S 0 ( x ) 1 and ω 20 ( x ) = V 0 ( x ) V 0 ( x ) 1 . Then, it follows from the positivity of T ˜ i ( t )   ( i = 1 , 2 ) that

b 2 ( t ) = max x D ¯ { T ˜ 2 ( t ) ω 20 ( x ) , 0 } T ˜ 2 ( t ) ω 20 ( x ) l M 2 e ρ ˜ 2 t V m H 2 e ρ ˜ 2 t , b 1 ( t ) = max x D ¯ T ˜ 1 ( t ) ω 10 ( x ) 0 t T ˜ 1 ( t s ) β ( x ) f ( S ( x , s ) , B ( x , s ) ) S 0 ( x ) η ( x ) ( V ( x , s ) V 0 ( x ) ) S 0 ( x ) d s , 0 T ˜ 1 ( t ) ω 10 ( x ) + 0 t T ˜ 1 ( t s ) η ( x ) ( V ( x , s ) V 0 ( x ) ) S 0 ( x ) d s M 1 S m l + η * H 2 V m e ( ρ ˜ 1 ρ ˜ 2 ) t ρ ˜ 1 ρ ˜ 2 1 e ρ ˜ 1 t H 1 e ρ ˜ 1 t ,

where S m = min x D ¯ { S 0 ( x ) } and V m = min x D ¯ { V 0 ( x ) } . Note that ( I A , I S , B ) satisfies

(34) I A t = D 3 Δ I A + θ β ( x ) f 2 ( S 0 ( x ) , 0 ) B + σ β ( x ) g 2 ( V 0 ( x ) , 0 ) B ( μ ( x ) + r A ( x ) + δ ( x ) ) I A + G 1 ( x , t ) , I S t = D 4 Δ I S + ( 1 θ ) β ( x ) f 2 ( S 0 ( x ) , 0 ) B + δ ( x ) I A ( μ ( x ) + r S ( x ) ) I S + G 2 ( x , t ) , B t = γ A ( x ) I A + γ S ( x ) I S α ( x ) B ,

for x D , t > 0 and I A ν = I S ν = 0 , x D , where

G 1 ( x , t ) = θ β ( x ) [ f ( S , B ) f 2 ( S 0 ( x ) , 0 ) B ] + σ β ( x ) [ g ( V , B ) g 2 ( V 0 ( x ) , 0 ) ] , G 2 ( x , t ) = ( 1 θ ) β ( x ) [ f ( S , B ) f 2 ( S 0 ( x ) , 0 ) B ] .

Then we can derive

I A I S B T ( t ) I A 0 I S 0 B 0 + 0 t T ( t s ) θ β ( x ) ( f 2 ( S ( x , s ) , 0 ) f 2 ( S 0 ( x ) , 0 ) ) B ( x , s ) + Q ( x ) ( 1 θ ) β ( x ) ( f 2 ( S ( x , s ) , 0 ) f 2 ( S 0 ( x ) , 0 ) ) B ( x , s ) 0 T ( t ) I A 0 I S 0 B 0 + 0 t T ( t s ) θ β ( x ) L 1 S ( x , s ) S 0 ( x ) + θ β ( x ) L 2 V ( x , s ) V 0 ( x ) B ( 1 θ ) β ( x ) L 1 S ( x , s ) S 0 ( x ) B 0 ,

where Q ( x ) = σ β ( x ) ( g 2 ( V ( x , s ) , 0 ) + g 2 ( V 0 ( x ) , 0 ) ) B ( x , s ) , L 1 and L 2 are Lipschitz constants for f 2 and g 2 , respectively. By taking Lemma 4 and 0 = 1 , one derives s ( ) = 0 ; thus, there exists M ¯ > 0 such that T ( t ) M ¯ , t 0 . From b 1 ( s ) H 1 e ρ ˜ 1 s and b 2 ( s ) H 2 e ρ ˜ 2 s , one has

(35) max { I A ( x , t ) , I S ( x , t ) , B ( x , t ) } M ¯ l + M ¯ θ β * L 1 S 0 ( x ) H 1 0 t e ρ ˜ 1 s B ( x , s ) d s + M ¯ σ β * L 2 V 0 ( x ) H 2 0 t e ρ ˜ 2 s B ( x , s ) d s M ¯ l + ( N 1 + N 2 ) 0 t e ρ ˜ s B ( x , s ) d s ,

where N 1 = M ¯ θ β * L 1 S 0 ( x ) H 1 , N 2 = M ¯ σ β * L 2 V 0 ( x ) H 2 and ρ ˜ = min { ρ ˜ 1 , ρ ˜ 2 } . Further, by using Gronwall’s inequality, one can derive

(36) B ( x , t ) M ¯ l e 0 t ( N 1 + N 2 ) e ρ ˜ s d s M ¯ l e N 1 + N 2 ρ ˜ .

By combining (35) and (36), we can also derive

(37) I A ( x , t ) M ¯ l + ( N 1 + N 2 ) M ¯ l e ( N 1 + N 2 ) ρ ˜ 0 t e ρ ˜ s d s M ¯ l 1 + ( N 1 + N 2 ) ρ ˜ e ( N 1 + N 2 ) ρ ˜ , I S ( x , t ) M ¯ l + ( N 1 + N 2 ) M ¯ l e ( N 1 + N 2 ) ρ ˜ 0 t e ρ ˜ s d s M ¯ l 1 + ( N 1 + N 2 ) ρ ˜ e ( N 1 + N 2 ) ρ ˜ .

As a result, by (36), ( S ( x , t ) , V ( x , t ) ) ( S ˆ ( x , t ) , V ˆ ( x , t ) ) , where ( S ˆ ( x , t ) , V ˆ ( x , t ) ) satisfies

(38) S ˆ t = D 1 Δ S ˆ + ( 1 p ( x ) ) Λ ( x ) μ ( x ) S ˆ β ( x ) M ¯ l e ( N 1 + N 2 ) ρ ˜ S ˆ + η ( x ) V ˆ , x D , t > 0 , V ˆ t = D 2 Δ V ˆ + p ( x ) Λ ( x ) ( μ ( x ) + η ( x ) ) V ˆ σ β ( x ) M ¯ l e ( N 1 + N 2 ) ρ ˜ V ˆ , x D , t > 0 , S ˆ ν = V ˆ ν = 0 , x D , t > 0 , S ˆ ( x , 0 ) = S 0 ( x ) , V ˆ ( x , 0 ) = V 0 ( x ) , x D .

Let ( S ˆ l ( x , t ) , V ˆ l ( x , t ) ) is the positive steady state of (38) and S ¯ ( x , t ) = S ˆ ( x , t ) S ˆ l ( x ) , V ¯ ( x , t ) = V ˆ ( x , t ) V ˆ l ( x ) . Then ( S ¯ ( x , t ) , V ¯ ( x , t ) ) satisfies

(39) S ¯ t = D 1 Δ S ¯ ( μ ( x ) + β ( x ) M ¯ l e ( N 1 + N 2 ) ρ ˜ ) S ¯ + η ( x ) V ¯ , x D , t > 0 , V ¯ t = D 2 Δ V ¯ ( ( μ ( x ) + η ( x ) ) σ β ( x ) M ¯ l e ( N 1 + N 2 ) ρ ˜ ) V ¯ , x D , t > 0 , S ¯ ν = V ¯ ν = 0 , x D , t > 0 , S ¯ ( x , 0 ) = S ( x , 0 ) S ˆ l ( x ) , V ¯ ( x , 0 ) = V ( x , 0 ) V ˆ l ( x ) , x D .

Let T ˜ i ( t )   ( i = 1 , 2 ) be the semigroup generated by D 1 Δ μ ( x ) and D 2 Δ ( μ ( x ) + η ( x ) ) , respectively, and satisfying the boundary conditions. We can choose a M i > 0 such that T ˜ 1 ( t ) M 1 e μ * t , T ˜ 2 ( t ) M 2 e ( μ * + η * ) t . By (39), one has

S ¯ ( x , t ) = T ˜ 1 ( t ) ( S 0 ( x ) S ˆ l ( x ) ) 0 t T ˜ 1 ( t s ) ( β ( x ) M ¯ l e ( N 1 + N 2 ) ρ ˜ S ¯ ( x , s ) η ( x ) V ¯ ( x , s ) ) d s , V ¯ ( x , t ) = T ˜ 2 ( t ) ( V 0 ( x ) V ˆ l ( x ) ) 0 t T ˜ 2 ( t s ) σ β ( x ) M ¯ l e ( N 1 + N 2 ) ρ ˜ V ¯ ( x , s ) d s .

Moreover, S ¯ ( x , t ) and V ¯ ( x , t ) can be estimated as follows:

S ¯ ( x , t ) M 1 S 0 ( x ) S ˆ l ( x ) e μ * t + 0 t M 1 e μ * ( t s ) ( β * M ¯ l e ( N 1 + N 2 ) ρ ˆ S ¯ ( x , s ) + η * V ¯ ( x , s ) ) d s , V ¯ ( x , t ) M 2 V 0 ( x ) V ˆ l ( x ) e ( μ * + η * ) t + 0 t M 2 e ( μ * + η * ) ( t s ) σ β * M ¯ l e ( N 1 + N 2 ) ρ ˜ V ¯ ( x , s ) d s .

Let K 1 = M 1 M ¯ β * l e ( N 1 + N 2 ) ρ ˜ and K 2 = M 2 M ¯ l σ β * e ( N 1 + N 2 ) ρ ˜ . After that, it is possible to use Gronwall’s inequality to derive the preceding equation as follows:

S ˆ ( x , t ) S ˆ l ( x ) = S ¯ ( x , t ) M 1 S 0 ( x ) S ˆ l ( x ) + Q m V 0 ( x ) V ˆ l ( x ) 0 t e ( K 2 η * ) s d s e K 1 t μ * t , V ˆ ( x , t ) V ˆ l ( x ) = V ¯ ( x , t ) M 2 V 0 ( x ) V ˆ l ( x ) e K 2 t ( μ * + η * ) t ,

where Q m = η * M 1 M 2 . Selecting adequately small l > 0 such that K 1 < μ * 2 , K 2 < η * 2 , one can launch

(40) S ˆ ( x , t ) S ˆ l ( x ) = S ¯ ( x , t ) ( M 1 S 0 ( x ) S ˆ l ( x ) + η * η * M 1 M 2 V 0 ( x ) V ˆ l ( x ) 2 ) e μ * t 2 , V ˆ ( x , t ) V ˆ l ( x ) = V ¯ ( x , t ) M 2 V 0 ( x ) V l ( x ) e ( μ * + η * ) t 2 .

Hence,

S ( x , t ) S 0 ( x ) S ˆ ( x , t ) S 0 ( x ) S ˆ ( x , t ) S ˆ l ( x ) + S ˆ l ( x ) S 0 ( x ) ( M 1 S 0 ( x ) S ˆ l ( x ) + η * η * M 1 M 2 V 0 ( x ) V ˆ l ( x ) ) e μ * t 2 + S ˆ l ( x ) S 0 ( x ) M 1 ( S 0 ( x ) S 0 ( x ) + S 0 ( x ) S ˆ l ( x ) ) S ˆ l ( x ) S 0 ( x ) Q M V 0 ( x ) V ˆ l ( x ) M 1 l ( M 1 + 1 ) S 0 ( x ) S ˆ l ( x ) - Q M V 0 ( x ) V ˆ l ( x ) , V ( x , t ) V 0 ( x ) V ˆ ( x , t ) V 0 ( x ) V ˆ ( x , t ) V ˆ l ( x ) + V ˆ l ( x ) V 0 ( x ) M 2 V 0 ( x ) V ˆ l ( x ) e ( μ * + η * ) t 2 + V ˆ l ( x ) V 0 ( x ) M 2 ( V 0 ( x ) V 0 ( x ) + V 0 ( x ) V ˆ l ( x ) ) V ˆ l ( x ) V 0 ( x ) M 2 l ( M 2 + 1 ) V 0 ( x ) V ˆ l ( x ) ,

where Q M = η * η * M 1 M 2 . Since b 1 ( t ) H 1 e ρ ˆ 1 t and b 2 ( t ) H 2 e ρ ˆ 2 t , one has

(41) S ( x , t ) S 0 ( x ) = S 0 ( x ) S ( x , t ) S 0 ( x ) 1 S 0 ( x ) H 1 e ρ ˆ 1 t S 0 ( x ) H 1 , V ( x , t ) V 0 ( x ) = V 0 ( x ) V ( x , t ) V 0 ( x ) 1 V 0 ( x ) H 2 e ρ ˆ 2 t V 0 ( x ) H 2 .

Integrating (40) and (41) yields

(42) S ( x , t ) S 0 ( x ) max { M 1 l + ( M 1 + 1 ) S 0 ( x ) S ˆ l ( x ) + Q M V ˆ 0 V ˆ l ( x ) , H 1 S 0 ( x ) } , V ( x , t ) V 0 ( x ) max { M 2 l + ( M 2 + 1 ) V 0 ( x ) V ˆ l ( x ) , H 2 V 0 ( x ) } .

Finally, combining (36)–(37), (42) and lim l 0 S l ( x ) = S 0 ( x ) , lim l 0 V l ( x ) = V 0 ( x ) , choose sufficiently small l such that for any t > 0 ,

S ( x , t ) S 0 ( x ) , V ( x , t ) V 0 ( x ) , I A ( x , t ) , I S ( x , t ) , B ( x , t ) ε .

This proves the local stability of 0 .

Further, let us prove the global attractiveness of 0 . Let Φ ( t ) and X + be as defined in Lemma 2. According to Lemma 3, Φ ( t ) has a global attractor A . Combined with Lemmas 4 and 5, problem (26) has a positive eigenvector ( ψ 3 , ψ 4 , ψ 5 ) correlated to the eigenvalue s ( ) . Define

X 1 = { ( S , V , I A , I S , B ) X + : I A = I S = B = 0 } .

Claim I: For all ϕ A , the ω limit set ω ( ϕ ) X 1 .

From (4) and (5), it follows that S ( x , 0 ) S 0 ( x ) and V ( x , 0 ) V 0 ( x ) . In case ϕ 3 = ϕ 4 = ϕ 5 = 0 , it is easy to conclude that X 1 is invariant for Φ ( t ) . It follows that ϕ 3 0 or ϕ 4 0 or ϕ 5 0 . By Lemma 7, one has I A ( x , t ) , I S ( x , t ) , B ( x , t ) > 0 , x D ¯ , t > 0 . Therefore, ( S ( x , t ) , V ( x , t ) ) fulfills

S t < ( 1 p ( x ) ) Λ ( x ) μ ( x ) S + η ( x ) V , x D , t > 0 , V t < p ( x ) Λ ( x ) ( μ ( x ) + η ( x ) ) V , x D , t > 0 , S ν = V ν = 0 , x D , S ( x , 0 ) S 0 ( x ) , V ( x , 0 ) V 0 ( x ) , x D .

Using the principle of comparison, it follows that S ( x , t ) < S 0 ( x ) , V ( x , t ) < V 0 ( x ) , ( x , t ) D ¯ × [ 0 , ) .

Inspired by [8], we define

c ( t , ϕ ) = inf { c ˘ R : I A c ˘ ψ 3 , I S c ˘ ψ 4 and B c ˘ ψ 5 } .

Clearly, c ( t , ϕ ) > 0 for any t > 0 . It follows immediately that we prove that c ( t , ϕ ) is a strictly monotonically decreasing function. For this reason, fix t ¯ 0 > 0 , and let I ¯ A ( x , t ) = c ( t ¯ 0 , ϕ ) ψ 3 , I ¯ S ( x , t ) = c ( t ¯ 0 , ϕ ) ψ 4 and B ( x , t ) = c ( t ¯ 0 , ϕ ) ψ 5 , t t ¯ 0 . Combined with S ( x , t ) < S 0 ( x ) , V ( x , t ) < V 0 ( x ) , we have

(43) I ¯ A t > D 3 Δ I A + θ ( x ) β ( x ) f ( S , B ¯ ) + σ β ( x ) g ( V , B ¯ ) ( μ ( x ) + r A ( x ) + δ ( x ) ) I ¯ A , x D , t > t 0 , I ¯ S t > D 4 Δ I S + ( 1 θ ) β ( x ) f ( S , B ¯ ) + δ ( x ) I ¯ A ( μ ( x ) + r S ( x ) ) I ¯ S , x D , t > t 0 , B ¯ t = γ A ( x ) I ¯ A + γ S ( x ) I ¯ S α ( x ) B ¯ , x D , t > t 0 , I ¯ A ( x , t ¯ 0 ) I A ( x , t ¯ 0 ) , I ¯ S ( x , t ¯ 0 ) I S ( x , t ¯ 0 ) , B ¯ ( x , t ¯ 0 ) B ( x , t ¯ 0 ) , x D ,

where I ¯ ν = I ¯ S ν = 0 , x D , t > t 0 . Via the comparison principle, we can acquire ( I ¯ A ( x , t ) , I ¯ S ( x , t ) , B ¯ ( x , t ) ) ( I A ( x , t ) , I S ( x , t ) , B ( x , t ) ) for any x D ¯ and t > t ¯ 0 . From the first second equations of (43) and comparison principle, it follows that c ( t ¯ 0 , ϕ ) ψ 3 = I ¯ A ( x , t ) > I ( x , t ) , c ( t ¯ 0 , ϕ ) ψ 4 = I ¯ S ( x , t ) > I S ( x , t ) and c ( t ¯ 0 , ϕ ) ψ 5 = B ¯ ( x , t ) > B ( x , t ) , for any x D ¯ and t > t ¯ 0 . Because t ¯ 0 is arbitrary, c ( t , ϕ ) is strictly decreasing function.

Definition by c * = lim t c ( t , ϕ ) , we verify that c * = 0 . As a matter of fact, by setting F = ( S , V , I A , I S , B ) ω ( ϕ ) , there exists { t k } satisfying Φ ( t k ) ϕ F . Since

lim t k Φ ( t + t k ) ϕ = Φ ( t ) lim t Φ ( t k ) ϕ = Φ ( t ) F ,

we immediately obtain that c ( t , F ) = c * , t 0 . If I A 0 or I S 0 or B 0 , by repeating the above processes, we know that c ( t , F ) is a strictly decreasing function. This results in the contradiction that c ( t , F ) = c * . Therefore, I A = I S = B = 0 .

Claim II. A = { 0 } .

Claim I guarantees that { 0 } is globally attractive in X 1 . Moreover, { 0 } constitutes the unique invariant subset in X 1 . Since the ω limit set ω ( ϕ ) is a compact invariant and ω ( ϕ ) X 1 , we arrive at ω ( ϕ ) = { 0 } , ϕ A . Lemma 3 shows that 0 is a compact invariant in X + . Combining the global attractivity of { 0 } as well as [39, Lemma 3.11] indicates that A = { 0 } .

With the aforementioned analysis, we can deduce that 0 is globally asymptotically stable.□

6 Asymptotic profiles of the endemic steady state

In this section, we study the asymptotic behavior of the steady state to evaluate the impact of diffusion rate on the propagation of the infection. The efficacy of vaccines is increasing with the development of medical technology, we consider ignoring the immune loss rate, i.e., taking η ( x ) = 0 in model (1). For the purpose of better investigating the asymptotic behavior of model (1) endemic steady state, we select f ( S , B ) = S B ( 1 + c 1 B ) , g ( V , B ) = V B ( 1 + c 2 B ) , where c 1 and c 2 are the saturated coefficients. In light of Theorem 2, we learn that if 0 > 1 , then model (1) has at least one steady state ( S , V , I A , I S , B ) , which satisfies

(44) D 1 Δ S = ( 1 p ( x ) ) Λ ( x ) μ ( x ) S β ( x ) S B 1 + c 1 B , x D D 2 Δ V = p ( x ) Λ ( x ) ( μ ( x ) + η ( x ) ) V σ β ( x ) V B 1 + c 2 B , x D , D 3 Δ I A = θ β ( x ) S B 1 + c 1 B + σ β ( x ) V B 1 + c 2 B ( μ ( x ) + r A ( x ) + δ ( x ) ) I A , x D , D 4 Δ I S = ( 1 θ ) β ( x ) S B 1 + c 1 B + δ ( x ) I A ( μ ( x ) + r S ( x ) ) I S , x D , 0 = γ S ( x ) I S + γ A ( x ) I A α ( x ) B , x D , S ν = V ν = I A ν = I S ν = 0 , x D .

It is easy to come up with

B = 1 α ( x ) ( γ A ( x ) I A + γ S ( x ) I S ) ,

then ( S , V , I A , I S ) satisfies

(45) D 1 Δ S = ( 1 p ( x ) ) Λ ( x ) μ ( x ) S β ( x ) S B 1 1 + c 1 B 1 , x D , D 2 Δ V = p ( x ) Λ ( x ) μ ( x ) V σ β ( x ) V B 1 1 + c 2 B 1 , x D , D 3 Δ I A = θ β ( x ) S B 1 1 + c 1 B 1 + σ β ( x ) V B 1 1 + c 2 B 1 ( μ ( x ) + r A ( x ) + δ ( x ) ) I A , x D , D 4 Δ I S = ( 1 θ ) β ( x ) S B 1 1 + c 1 B 1 + δ ( x ) I A ( μ ( x ) + r S ( x ) ) I S , x D , S ν = V ν = I A n = I S ν = 0 , x D ,

here, B 1 is used to represent ( γ A ( x ) I A + γ S ( x ) I S ) α ( x ) .

Lemma 8

Let ( S , V , I A , I S ) be any solution of (45). For any D 1 , D 2 , D 3 , D 4 > 0 , one has the following conclusions that hold true.

  1. The L 1 -bounded of ( S , V , I A , I S )

    D S ( x ) d x Λ * D μ * , D V ( x ) d x p * Λ * D μ * , D I A ( x ) d x ( θ + σ p * ) β * Λ * D μ * ( μ * + r A * + δ * ) , D I S ( x ) d x δ * ( θ + σ p * ) β * Λ * D μ * ( μ * + r S * ) ( μ * + r A * + δ * ) + Λ * D ( 1 θ ) β * μ * ( μ * + r S * ) .

  2. The lower bounds of S and V

    S ( x ) c 1 ( 1 p * ) Λ * c 1 μ * + β * , V ( x ) c 2 p * Λ * c 2 μ * + σ β * , x D ¯ .

Proof

Arranging the first four equations of (45) and integrating over D yields

D Λ ( x ) d x = D ( μ ( x ) ( S ( x ) + V ( x ) + I A ( x ) + I S ( x ) ) + r A ( x ) I A ( x ) + r S ( x ) I S ( x ) ) d x .

Therefore, one obtains D S ( x ) d x Λ * D μ * . Integrating the second to fourth equations in (45) over D , we obtain

μ * D V ( x ) d x D μ ( x ) V ( x ) d x D μ ( x ) V ( x ) + σ β ( x ) V ( x ) B 1 ( x ) 1 + c 2 B 1 ( x ) d x = D p ( x ) Λ ( x ) d x p * Λ * D .

Similarly,

( μ * + r A * + δ * ) D I A ( x ) d x D ( μ ( x ) + r A ( x ) + δ ( x ) ) I A ( x ) d x = D θ β ( x ) S ( x ) B 1 ( x ) 1 + c 1 B 1 ( x ) d x + D σ β ( x ) V ( x ) B 1 ( x ) 1 + c 2 B 1 ( x ) d x θ β * D S ( x ) d x + σ β * D V ( x ) d x ,

and

( μ * + r S * ) D I S ( x ) d x D ( μ ( x ) + r S ( x ) ) I S ( x ) d x = D δ ( x ) I A ( x ) + ( 1 θ ) β ( x ) S ( x ) B 1 ( x ) 1 + c 1 B 1 ( x ) d x δ * D I A ( x ) d x + ( 1 θ ) β * D S ( x ) d x .

Thus, it follows that L 1 bound of V , I A and I S , i.e., D V ( x ) d x p * Λ * D μ * , D I A ( x ) d x ( θ + σ p * ) β * Λ * D ( μ * ( μ * + r A * + δ * ) ) and D I S ( x ) d x δ * ( θ + σ p * ) β * Λ * D / ( μ * ( μ * + r S * ) ( μ * + r A * + δ * ) ) + Λ * ( 1 θ ) p * D ( μ * ( μ * + r S * ) ) .

( i i ) Set x 0 D ¯ such that S ( x 0 ) = min x D ¯ { S ( x ) } and V ( x 1 ) = min x D ¯ { V ( x ) } . By the first equation of (45), combined with [18, Proposition 2.2], one has

( 1 p * ) Λ * ( 1 p ( x 0 ) ) Λ ( x 0 ) μ ( x 0 ) S ( x 0 ) + β ( x 0 ) S ( x 0 ) B 1 ( x 0 ) 1 + c 1 B 1 ( x 0 ) μ * + β * c 1 S ( x 0 ) .

Thus, we obtain S ( x ) c 1 ( 1 p * ) Λ * ( c 1 μ * + β * ) . From the second equation of (45), using the same method as above, we obtain

p * Λ * p ( x 1 ) Λ ( x 1 ) μ ( x 1 ) V ( x 1 ) + σ β ( x 1 ) V ( x 1 ) B 1 ( x 1 ) 1 + c 2 B 1 ( x 1 ) μ * + σ β * c 2 V ( x 1 ) .

Therefore, it is feasible to launch V ( x ) c 2 p * Λ * ( c 2 μ * + σ β * ) .□

6.1 The case of D 1 0

Throughout the present section, we discuss the asymptotic behavior of * as D 1 .

Theorem 4

Suppose that 0 > 1 . Fix D 2 , D 3 , D 4 > 0 , and let D 1 0   (up to a subsequences), then ( S ( x ) , V ( x ) , I S ( x ) , I A ( x ) ) satisfies

( S ( x ) , V ( x ) , I S ( x ) , I A ( x ) ) ( Φ S ( x ) , Φ V ( x ) , Φ I A ( x ) , Φ I S ( x ) ) uniformly o n x D ¯ ,

where

Φ S ( x ) = ( 1 p ( x ) ) Λ ( x ) ( 1 + c 1 Φ B 1 ( x ) ) μ ( x ) ( 1 + c 1 Φ B 1 ( x ) ) + β ( x ) Φ B 1 ( x ) ,

and ( Φ V ( x ) , Φ I A ( x ) , Φ I S ( x ) ) is a solution of the following problem

(46) D 2 Δ Φ V ( x ) = p ( x ) Λ ( x ) μ ( x ) Φ V ( x ) σ β ( x ) Φ V ( x ) Φ B 1 ( x ) 1 + c 2 Φ B 1 ( x ) , x D , D 3 Δ Φ I A ( x ) = θ β ( x ) Φ S ( x ) Φ B 1 ( x ) 1 + c 1 Φ B 1 ( x ) + σ β ( x ) Φ V ( x ) Φ B 1 ( x ) 1 + c 2 Φ B 1 ( x ) ( μ ( x ) + r A ( x ) + δ ( x ) ) Φ I A ( x ) , x D , D 4 Δ Φ I S ( x ) = ( 1 θ ) β ( x ) Φ S ( x ) Φ B 1 ( x ) 1 + c 1 Φ B 1 ( x ) + δ ( x ) Φ I A ( x ) ( μ ( x ) + r S ( x ) ) Φ I S ( x ) , x D , Φ V ( x ) ν = Φ I A ( x ) ν = Φ I S ( x ) ν = 0 , x D .

Proof

In this proof, we suppose that C is a positive constant and does not depend on D 1 , but allows to vary from place to place. In the interest of clarity, let’s divide the proof into three steps.

Step 1: The L p -bound of ( S , V , I A , I S ) for any p 1 .

It follows from Lemma 8 and the ellipse L 1 -estimated [24, Lemma 2.2] that I A ( ) C , I S ( ) C for all 1 q N ( N 1 ) (or if N = 1 , q [ 1 , ) ). The Sobolev embedding theorem W 1 , q ( D ) L p 1 ( D ) [12]. For any q [ 1 , N ( N 1 ) ) yields

I A ( ) L p 1 ( D ) C , I S ( ) L p 1 ( D ) C , p 1 ( 1 , q N ( N q ) ) .

It clearly follows from the fact that q can be close to N ( N 1 ) that

I A ( ) L p 1 ( D ) C , I S ( ) L p 1 ( D ) C , p 1 ( 1 , q N ( N 2 ) ) .

Evidently, if N 2 , then p 1 ( 1 , ) holds. Multiplying the S equation of (45) by S k   ( k > 0 ) and integrating over D , we obtain

k D 1 D S k 1 S 2 d x = D ( 1 p ( x ) ) Λ ( x ) S k d x D μ ( x ) S k + 1 d x D β S k + 1 B 1 1 + c 1 B 1 d x .

It is clear that

(47) μ * D S k + 1 d x ( 1 p * ) Λ * D S k d x .

Let k 1 = 1 q 1 and q 1 = 1 + 1 ( p 1 1 ) = p 1 ( p 1 1 ) (note that 1 p 1 + 1 q 1 = 1 ), we conclude that

μ * D S k 1 + 1 d x ( 1 p * ) Λ * D S k 1 d x ( 1 p * ) D 1 p 1 D S d x 1 q 1 C .

Then, we derive

S ( ) L k 1 + 1 ( D ) C .

Combining the Hölder inequality and (47), the following result can be obtained

μ * D S k 2 + 1 d x ( 1 p * ) Λ * D S k 2 d x ( 1 p * ) Λ * D 1 p 1 D S k 2 q 1 1 q 1 ,

where k 2 = ( k 1 + 1 ) q 1 = 1 q 1 + 1 q 1 2 . Hence, we conclude that S ( ) L k 2 + 1 ( D ) C . The iterative process can be repeated to obtain

k = 1 q 1 + 1 q 1 2 + 1 q 1 3 + = p 1 1 ,

then one can conclude that

S ( ) L k + 1 ( D ) = S ( ) L p 1 ( D ) C .

Similarly, we can also derive V ( ) L k + 1 ( D ) = V ( ) L p 1 ( D ) C . Further, combining the aforementioned results and the elliptic L p -theory for I A equation and I S equation of (45), one has

I A ( ) W 2 , p 1 ( D ) C , I S ( ) W 2 , p 1 ( D ) C .

Combining Sobolev’s embedding theorem W 2 , p 1 ( D ) L p 2 ( D ) for 1 < p 2 < N p 1 ( N 2 p 1 ) and p 1 ( 1 , N ( N 2 ) ) , one has

I A ( ) L p 2 ( D ) C , I A ( ) L p 2 ( D ) C , p 2 ( 1 , N ( N 4 ) or p 2 ( 1 , ) ) if N 4 .

By a similar argument as the L p 1 -bound of S and V , one can show S ( ) L p 2 ( D ) C , V ( ) L p 2 ( D ) C . Therefore, repeating the aforementioned argument concludes

S ( ) L p ( D ) C , V ( ) L p ( D ) C , I A ( ) L p ( D ) C , I S ( ) L p ( D ) C , 1 p < .

Step 2: Convergence of I A and I S .

Note that I A can be solved from

(48) Δ I A + μ ( x ) + r A ( x ) + δ ( x ) D 3 θ β ( x ) S B 1 D 3 ( 1 + c 1 B 1 ) I A σ β ( x ) V B 1 D 3 ( 1 + c 2 B 1 ) I A I A = 0 , x D , I A ν = 0 , x D .

Fix D 3 > 0 , it follows from the Lemma 8 ( i i ) that

μ ( x ) + r A ( x ) + δ ( x ) D 3 θ β ( x ) S B 1 D 3 ( 1 + c 1 B 1 ) I A σ β ( x ) V B 1 D 3 ( 1 + c 2 B 1 ) I A L p ( D ) μ ( x ) + r A ( x ) + δ ( x ) D 3 + θ β ( x ) S c 1 D 3 I A + σ β ( x ) V c 2 D 3 I A L p ( D ) C .

Choosing p sufficiently large, by using [27, Lemma 2.2], one has

(49) max x D ¯ { I A ( x ) } C min x D ¯ { I A ( x ) } ,

where C > 0 does not depend on D 1 . In the light of Lemma 8( i ) and (49), one can conclude that

(50) I A ( x ) C min x D ¯ { I A ( x ) } C D D I A d x C .

Rewrite the I A equation in (45) as follows:

(51) D 3 Δ I A + ( μ ( x ) + r A ( x ) + δ ( x ) ) I A = θ β ( x ) S B 1 1 + c 1 B 1 + σ β ( x ) V B 1 1 + c 2 B 1 , x D , I A ν = 0 , x D .

By (50), there holds

θ β ( x ) S B 1 1 + c 1 B 1 + σ β ( x ) V B 1 1 + c 2 B 1 L p ( D ) C , p 1 .

According to the standard L p estimate for elliptic equations (see, e.g., [12]), then

I A ( ) W 2 , p ( D ) C , p > 1 .

For sufficiently large p , the Sobolev embedding theorem [12] means that

I A ( ) C 1 + α ( D ¯ ) C , for 0 < α < 1 .

Similarly, we can also derive I A ( ) C 1 + α ( D ¯ ) C , for 0 < α < 1 . Consequently one can identify a subsequence D 1 (which is labeled by D S , n ), meeting D S , n 0 as n , and a corresponding positive solution ( S n ( x ) , V n ( x ) , I A n ( x ) , I S n ( x ) ) ( S ( x , D S , n ) , V ( x , D S , n ) , I A ( x , D S , n ) , I S ( x , D S , n ) ) of (45) for D 1 = D S , n satisfies

(52) ( I A n ( x ) , I S n ( x ) ) ( Φ I A ( x ) , Φ I S ( x ) ) uniformly on D ¯ , as n ,

where ( Φ I A ( x ) , Φ I S ( x ) ) C 1 ( D ) × C 1 ( D ) and Φ I A ( x ) , Φ I S ( x ) 0 on D ¯ . From (50), it is apparent that

either Φ I A ( x ) 0 on D ¯ or Φ I A ( x ) > 0 on D ¯ .

Assuming that Φ I A ( x ) 0 on D ¯ , that is,

I A n ( x ) 0 uniformly on D ¯ , as n .

Bringing in the third equation of (45) yields

(53) θ β ( x ) γ S ( x ) S n ( x ) I S n ( x ) α ( x ) + c 1 γ S ( x ) I S n ( x ) + σ β ( x ) γ S ( x ) V n ( x ) I S n ( x ) α ( x ) + c 2 γ S ( x ) I S n ( x ) = 0 , as n .

Combining this with the Lemma 8( i i ) yields

I S n ( x ) 0 uniformly on D ¯ , as n .

Then for any small ε , there exists n 1 > 0 , one has

0 I A n ( x ) ε , 0 I S n ( x ) ε , x D ¯ , for n n 1 .

Along with the first equation of (45), this implies that S n ( x ) is satisfied for all large n

D S , n Δ S n ( 1 p ( x ) ) Λ ( x ) μ ( x ) S n , x D , S n ν = 0 , x D

and

D S , n Δ S n ( 1 p ( x ) ) Λ ( x ) ε β * S n μ ( x ) S n , x D , S n ν = 0 , x D .

Taking an arbitrarily large n , there are two auxiliary systems to consider as follows:

(54) D S , n Δ S n = ( 1 p ( x ) ) Λ ( x ) μ ( x ) S n , x D , S n n = 0 , x D

and

(55) D S , n Δ S n = ( 1 p ( x ) ) Λ ( x ) ε β * S n μ ( x ) S n , x D , S n ν = 0 , x D .

Clearly, (54) and (55) contain unique positive solutions, denoted by u n and v n , respectively. On the basis of the sub-superior solution argument in [40], we can conclude that

v n S n u n on D ¯ for all large n .

Further by a similar proof as in [10, Lemma 2.4], it follows that

u n ( 1 p ( x ) Λ ( x ) ) μ ( x ) uniformly on D ¯ , as n

and

v n ( 1 p ( x ) Λ ( x ) ) μ ( x ) + ε β * uniformly on D ¯ , as n .

Thus, as n , one has

( 1 p ( x ) Λ ( x ) ) μ ( x ) + ε β * liminf n S n limsup n S n ( 1 p ( x ) Λ ( x ) ) μ ( x ) .

By the arbitrariness of ε yields

S n ( 1 p ( x ) ) Λ ( x ) μ ( x ) uniformly on D ¯ , as n .

Similarly, one can conclude that

V ( x ) p ( x ) Λ ( x ) μ ( x ) uniformly on D ¯ , as n .

Consider the I A equation of (45) satisfies

(56) D 3 Δ I A n = θ β ( x ) S n B 1 n 1 + c 1 B 1 n + σ β ( x ) V n B 1 n 1 + c 2 B 1 n ( μ ( x ) + r A ( x ) + δ ( x ) ) I A n , x D , I A n ν = 0 , x D .

Define I ¯ A n = I A n I A n L ( D ) , it is obvious that I ¯ A n L ( D ) = 1 for all n 1 , and I ¯ A n solves

(57) D 3 Δ I ¯ A n = θ γ A ( x ) β ( x ) S n α ( x ) ( 1 + c 1 B 1 n ) + σ γ A ( x ) β ( x ) V n α ( x ) ( 1 + c 2 B 1 n ) ( μ ( x ) + r A ( x ) + δ ( x ) ) I ¯ A n + θ γ S ( x ) β ( x ) S n I S n α ( x ) ( 1 + c 1 B 1 n ) I A n L ( D ) + σ γ S ( x ) β ( x ) V n I S n α ( x ) ( 1 + c 2 B 1 n ) I A n L ( D ) , x D , I ¯ A n ν = 0 , x D .

Starting from the standard compactness argument for elliptic equations, as n , one has I ¯ A n I ¯ A in C 1 ( D ) , as n , where I ¯ A 0 belongs to C 1 ( D ) and satisfies I ¯ A L ( D ) = 1 . Thus, as n , (57) becomes

D 3 Δ I ¯ A = β ( x ) γ A ( x ) Λ ( x ) α ( x ) μ ( x ) ( θ ( 1 p ( x ) ) + σ p ( x ) ) ( μ ( x ) + r A ( x ) + δ ( x ) ) I ¯ A , x D , I ¯ A ν = 0 , x D .

By the Harnack-type inequality, one has I ¯ A > 0 . Further, we consider the I S equation for (45)

D 4 I S n = ( 1 θ ) β ( x ) S n B 1 n 1 + c 1 B 1 n ( μ ( x ) + r S ( x ) ) I S n + δ ( x ) θ β ( x ) S n B 1 n 1 + c 1 B 1 n + σ β ( x ) V n B 1 n 1 + c 2 B 1 n × ( D 3 Δ + ( μ ( x ) + r A ( x ) + δ ( x ) ) ) 1 , x D , I S n ν = 0 , x D .

Define I ¯ S n = I S n I S n L ( D ) , taking a similar approach yields I ¯ S n I ¯ S > 0 in C 1 ( D ) , as n . where, I S satisfies

D 4 Δ I ¯ S = ( 1 θ ) β ( x ) γ S ( x ) ( 1 p ( x ) ) Λ ( x ) α ( x ) μ ( x ) ( μ ( x ) + r S ( x ) ) + Λ ( x ) γ S ( x ) β ( x ) α ( x ) μ ( x ) × δ ( x ) ( θ ( 1 p ( x ) ) + σ p ( x ) ) D 3 Δ + ( μ ( x ) + r A ( x ) + δ ( x ) ) I ¯ S , x D , I ¯ S ν = 0 , x D .

Moreover, due to the uniqueness of the principal eigenvalue, 0 is the principal eigenvalue of (19) and (23). Combined with (18) and Lemma 4, this contradicts 0 > 1 . Thus, Φ I A > 0 and Φ I S > 0 on D ¯ , which means that I A n Φ I A > 0 and I S n Φ I S > 0 uniformly on D ¯ , as n .

Step 3: Convergence of S .

From (45), we know that

D 1 Δ S n = ( 1 p ( x ) ) Λ ( x ) μ ( x ) S n β ( x ) S n I n 1 + c 1 B 1 n , x D .

By (52), for any small ε > 0 , it follows that we have a sufficiently large n such that

0 < Φ I A ( x ) ε I A n ( x ) Φ I A ( x ) + ε , 0 < Φ I S ( x ) ε I S n ( x ) Φ I S ( x ) + ε , x D ¯ .

Thus,

(58) 0 < 1 α ( x ) [ ( Φ I A ( x ) ε ) + ( Φ I S ( x ) ε ) ] = Φ B 1 ( x ) 2 ε 1 α ( x ) ( I A n ( x ) + I S n ( x ) ) 1 α ( x ) [ ( Φ I A ( x ) + ε ) + ( Φ I S ( x ) + ε ) ] = Φ B 1 ( x ) + 2 ε .

So, for sufficiently large n , one has

( 1 p ( x ) ) Λ ( x ) μ ( x ) S n β ( x ) S n B 1 n 1 + c 1 B 1 n ( 1 p ( x ) ) Λ ( x ) μ ( x ) S n β ( x ) S n ( Φ B 1 n 2 ε ) 1 + c 1 ( Φ B 1 n + 2 ε ) = ( g ε ( x , Φ I A , Φ I S ) S n ) φ ε ( x , Φ I A , Φ I S ) 1 + c 1 ( Φ B 1 + 2 ε ) ,

where

g ε ( x , Φ I A , Φ I S ) = ( 1 p ( x ) ) Λ ( x ) ( 1 + c 1 ( Φ B 1 + 2 ε ) ) μ ( x ) ( 1 + c 1 ( Φ B 1 + 2 ε ) ) + β ( x ) ( Φ B 1 2 ε ) , φ ε ( x , Φ I A , Φ I S ) = μ ( x ) ( 1 + c 1 ( Φ B 1 + 2 ε ) ) + β ( x ) ( Φ B 1 2 ε ) .

For adequately large n , let us study the below auxiliary problem

(59) D n Δ z = ( g ε ( x , Φ I A , Φ I S ) z ) φ ε ( x , Φ I A , Φ I S ) 1 + c 1 ( Φ B 1 + 2 ε ) , x D ,

with z ν = 0 , x D . Note that ( S n , C ) is a a couple of sub-supersolutions of (59), where C > 0 is a sufficiently large positive constant C meeting S n C . Therefore, (59) admits at least a positive solution z ¯ n satisfying S n z ¯ n C on D ¯ . Then, similar arguments applied to [10, Lemma 2.4] can be shown that any solution z ¯ n of (59) satisfies

z ¯ n g ε ( x , Φ I A , Φ I S ) uniformly on D ¯ .

By S n z ¯ n C , it indicate that

(60) limsup n S n g ε ( x , Φ I A , Φ I S ) uniformly on D ¯ .

Furthermore, by (58), one has

( 1 p ( x ) ) Λ ( x ) μ ( x ) S n β ( x ) S n B 1 n 1 + c 1 B 1 n ( 1 p ( x ) ) Λ ( x ) μ ( x ) S n β ( x ) S n ( Φ B 1 n + 2 ε ) 1 + c 1 ( Φ B 1 n 2 ε ) = ( g ε ( x , Φ I A , Φ I S ) S n ) φ ε ( x , Φ I A , Φ I S ) 1 + c 1 ( Φ B 1 2 ε ) ,

where

g ε ( x , Φ I A , Φ I S ) = ( 1 p ( x ) ) Λ ( x ) ( 1 + c 1 ( Φ B 1 2 ε ) ) μ ( x ) ( 1 + c 1 ( Φ B 1 2 ε ) ) + β ( x ) ( Φ B 1 + 2 ε ) , φ ε ( x , Φ I A , Φ I S ) = μ ( x ) ( 1 + c 1 ( Φ B 1 2 ε ) ) + β ( x ) ( Φ B 1 2 ε ) .

Arguing similarly as before, this leads to liminf t S n g ε ( x , Φ I A , Φ I S ) . We further obtain

lim ε 0 g ε ( x , Φ I A ( x ) , Φ I S ( x ) ) = lim ε 0 g ε ( x , Φ I A ( x ) , Φ I S ( x ) ) = ( 1 p ( x ) ) Λ ( x ) ( 1 + c 1 Φ B 1 ( x ) ) μ ( x ) ( 1 + c 1 Φ B 1 ( x ) ) + β ( x ) Φ B 1 ( x ) = G ( x , Φ I A ( x ) , Φ I S ( x ) ) .

Combining the above results, it follows that

S n ( x ) G ( x , Φ I A ( x ) , Φ I S ( x ) ) uniformly on D ¯ , as n .

Therefore, it can be conveniently derived that Φ V ( x ) , Φ I A ( x ) and Φ I S ( x ) meet (46). The proof is complete.□

6.2 The case of D 3 0

We investigate the asymptotic behavior of * with respect to D 3 0 in this subsection.

Theorem 5

Suppose that 0 > 1 . Fix D 1 , D 2 , and D 4 are positive constants and D 3 0   (up to a subsequnece), then every positive ( S ( x ) , V ( x ) , I A ( x ) , I S ( x ) ) meets

( S ( x ) , V ( x ) , I A ( x ) , I S ( x ) ) ( Ψ S ( x ) , Ψ V ( x ) , Ψ I A ( x ) , Ψ I S ( x ) ) uniformly o n D ¯ ,

where

(61) I A ( x ) = β Φ B 1 ( θ Φ S ( 1 + c 1 Φ B 1 ) + σ Φ V ( 1 + c 1 Φ B 1 ) ) ( 1 + c 2 Φ B 1 ) ( 1 + c 1 B 1 ) ,

and ( Ψ S ( x ) , Ψ V ( x ) , Ψ I S ( x ) ) satisfies

(62) D 1 Δ Ψ S = ( 1 p ( x ) ) Λ ( x ) μ ( x ) Ψ S β ( x ) Ψ S Ψ B 1 1 + c 1 Ψ B 1 , x D , D 2 Δ Ψ V = p ( x ) Λ ( x ) μ ( x ) Ψ V σ β ( x ) Ψ V Φ B 1 1 + c 2 Ψ B 1 , x D , D 4 Δ Ψ I S = ( 1 θ ) β ( x ) Ψ S Ψ B 1 1 + c 1 Ψ B 1 + δ ( x ) Ψ I A ( μ ( x ) + r S ( x ) ) Ψ I S , x D , Ψ S ν = Ψ V ν = Ψ I S ν = 0 , x D .

Proof

We will deal with proof by means of belowing three claims.

Step 1: The endemic steady state ( S , V , I A , I S ) is L p -bound for p 1 .

Arguing in a similar way to Step 1 of the proof of Theorem 4, the elliptic L 1 -theory of he first to second equations of (45) and the Sobolev embedding theorem lead to

(63) S ( ) L p 1 ( D ) C , V ( ) L p 1 ( D ) C , p 1 ( 1 , N ( N 2 ) ) ( or p 1 ( 0 , ) if N 2 ) ,

where the positive constant C is independent of small D 3 . For k > 0 , integrating over D by multiplying the third equation in (45) by I A k yields

( μ * + r A * + δ * ) D I A k + 1 d x D ( μ ( x ) + r A ( x ) + δ ( x ) ) I A k + 1 d x θ β * c 1 D S I A k d x + σ β * c 2 D V I A k d x .

Using the approach of Step 1 in Theorem 4 yields the L p -bound of I A . Similarly, multiplying I S equation for (45) by I S k and repeating the above process yields L p -bound of I S . Thus, we can conclude that

(64) S ( ) L p ( D ) C , V ( ) L p ( D ) C , I A ( ) L p ( D ) C , I S ( ) L p ( D ) C , 1 q < .

Step 2: Convergence of S , V and I S .

By (63), the elliptic L p -theory and Sobolev embedding theorem assure that

S ( ) C 1 + α ( D ¯ ) C , V ( ) C 1 + α ( D ¯ ) C , I A ( ) C 1 + α ( D ¯ ) C , I S ( ) C 1 + α ( D ¯ ) C , α ( 0 , 1 ) .

Then D 3 0 contains a subsequence that converges to zero as j , denoted D 3 = D 3 , j 0 , such that the corresponding positive solution ( S j ( x ) , V j ( x ) , I A , j ( x ) , I S , j ( x ) ) ( Ψ S ( x ) , Ψ V ( x ) , Ψ I A ( x ) , Ψ I S ( x ) ) = ( S ( x , D 3 , j ) , V ( x , D 3 , j ) , I A ( x , D 3 , j ) , I S ( x , D 3 , j ) ) of (45) for D 3 = D 3 , j fulfills

( S j ( x ) , V j ( x ) , I S , j ( x ) ) ( Ψ S ( x ) , Ψ V ( x ) , Ψ I S ( x ) ) in C 1 ( D ¯ ) × C 1 ( D ) × C 1 ( D ) as j ,

where Ψ I S ( x ) 0 and Ψ S ( x ) , Ψ V ( x ) > 0 due to Lemma 8( i i ).

Step 3: Convergence of I A .

It is noteworthy that I A , j satisfies

(65) D 3 Δ I A , j = θ β ( x ) S j B 1 , j 1 + c 1 B 1 , j + σ β ( x ) V j B 1 , j 1 + c 2 B 1 , j ( μ ( x ) + r ( x ) + δ ( x ) ) I A , j , x D , I A , j ν = 0 , x D .

From (64) and (65), It is possible to employ the sub-supersolution comparison argument as in Claim III of Theorem 4 (see [19, Theorem 1.1]) to conclude that I A , j ( x ) Ψ I A ( x ) in C 1 ( D ¯ ) as j , where Ψ I A ( x ) fulfills (61). Moreover, the formulation of Ψ I A ( x ) , it is evidence that ( Ψ S ( x ) , Ψ V ( x ) , Ψ I A ( x ) , Ψ I S ( x ) ) is the unique positive solution of (62). This completes the proof.□

6.3 The case of D 1 or D 3

In the following, we analyze the asymptotic behavior of * when D 1 or D 3 . The results are presented below.

Theorem 6

Suppose that 0 > 1 . Fix D 2 , D 3 , D 4 > 0 and let D 1 , then every solution ( S ( x ) , V ( x ) , I A ( x ) , I S ( x ) ) of (45) meets

( S ( x ) , V ( x ) , I A ( x ) , I S ( x ) ) ( S ( x ) , V ( x ) , I A ( x ) , I S ( x ) ) uniformly o n D ¯ ,

where S ( x ) > 0 is a constant, V ( x ) , I A ( x ) , I S ( x ) > 0 on D ¯ , and ( S ( x ) , V ( x ) , I A ( x ) , I S ( x ) ) solves

(66) D 2 Δ V = p ( x ) Λ ( x ) ( μ ( x ) + η ( x ) ) V σ β ( x ) V B 1 1 + c 1 B 1 , x D , D 3 Δ I A = θ β ( x ) S B 1 1 + c 1 B 1 + σ β ( x ) V B 1 1 + c 1 B 1 ( μ ( x ) + r ( x ) + δ ( x ) ) I A , x D , D 4 Δ I S = ( 1 θ ) β ( x ) S B 1 + c 1 B + δ ( x ) I A ( μ ( x ) + r S ( x ) ) I S , x D , V ν = I A ν = I S ν = 0 , x D , D μ ( x ) S + β ( x ) S B 1 1 + c 1 B 1 d x = D ( 1 p ( x ) ) Λ ( x ) d x .

Proof

By Theorem 4, the L p -bound of ( S , V , I A , I S ) is valid for all D 1 1 . The first equation of (45) can be rewritten as follows:

D 1 Δ S = ( 1 p ( x ) ) Λ ( x ) μ ( x ) S β ( x ) S B 1 1 + c 1 B 1 , x D , S ν = 0 , x D .

Then, the elliptic L p -theory Sobolev theorem and standard compactness argument guarantee that there exists a subsequence of D S , k , labeled by D k , with D k as k and a corresponding positive solution ( S k ( x ) , V k ( x ) , I A , k ( x ) , I S , k ( x ) ) with D 1 = D k satisfies S k S > 0 in C 1 ( D ) due to Lemma 8( i i ). Further, S solves

Δ S = 0 , x D , S ν = 0 , x D .

Therefore, S > 0 must be a constant. Similar to the previous, with the help of the I A and I S equations of (45), passed to another subsequence if necessary, one has

I A , k I A , I S , k I S on C 1 ( D ¯ ) , as k ,

where I A 0 and I S 0 on D ¯ . As the Harnack inequality of I A also holds, either I A > 0 on D ¯ or I A 0 on D ¯ . Analyzing similarly to Step 2 in the proof of Theorem 4, let’s proceed with the argument by contradiction and suppose that I A 0 on D ¯ . Then from (53), one can derive that I S 0 on D ¯ . Define I ˜ A , k I A , k I A , k L ( D ) . Thus, I ˜ A , k L ( D ) = 1 for k 1 and I ˜ A , k satisfies

D 3 Δ I ˜ A , k = θ γ A ( x ) β ( x ) S k α ( x ) ( 1 + c 1 B 1 , k ) + σ γ A ( x ) β ( x ) V k α ( x ) ( 1 + c 2 B 1 , k ) ( μ ( x ) + r A ( x ) + δ ( x ) ) I ˜ A , k + θ γ S ( x ) β ( x ) S k α ( x ) ( 1 + c 1 B 1 , k ) I A , k L ( D ) + σ γ S ( x ) β ( x ) V k α ( x ) ( 1 + c 2 B 1 , k ) I A , k L ( D ) , x D , I ˜ A , k ν = 0 , x D .

The proof that is similar to the Theorem 4 leads to I ˜ A , k I ˆ A in C 1 ( D ¯ ) as k , where I ˆ A 0 on D ¯ and I ˆ A L ( D ) = 1 . As k , I A , k 0 and I S , k , by sub-supersolution argument that S = D ( 1 p ( x ) ) Λ ( x ) d x D μ ( x ) d x , V = D p ( x ) Λ ( x ) d x D μ ( x ) d x . Thus, I ˆ A solves

D 3 Δ I ˆ A = θ γ A ( x ) β ( x ) D ( 1 p ( x ) ) Λ ( x ) d x α ( x ) D μ ( x ) d x + σ γ A ( x ) β ( x ) D p ( x ) Λ ( x ) d x α ( x ) D μ ( x ) d x ( μ ( x ) + r A ( x ) + δ ( x ) ) I ˆ A , x D , I ˆ A ν = 0 , x D .

It is evident that the Harnack-type inequality also holds for I ˆ A , then I ˆ A > 0 on D ¯ . Define I ˜ S , k I S , k I S , k L ( D ) , it is easy to derive I ˜ S , k I ˆ S in C 1 ( D ¯ ) as k , where I ˆ S 0 on D ¯ and satisfies

D 4 Δ I ˆ S = ( 1 θ ) β ( x ) γ S ( x ) D ( 1 p ( x ) ) Λ ( x ) d x α ( x ) D μ ( x ) d x ( μ ( x ) + r S ( x ) ) + γ S ( x ) β ( x ) α ( x ) × δ ( x ) θ D Λ ( x ) ( 1 p ( x ) ) d x + σ d x Λ ( x ) p ( x ) ( D 3 Δ + ( μ ( x ) + r A ( x ) + δ ( x ) ) ) D μ ( x ) d x I ˆ S , x D , I ˆ S ν = 0 , x D .

Therefore, 0 is the principal eigenvalue of (19) and (23). Combined with (18) and Lemma 4. This contradiction with 0 > 1 . As a result, I A > 0 satisfies (66). This completes the proof.□

As far as Theorem 7 is concerned, the method of proof is very similar to Theorem 6, so that we simply revise it slightly and then omit the details. It turns out to be as follows.

Theorem 7

Suppose that 0 > 1 . Fix D 1 , D 2 , D 4 are positive constants and let D 3   (up to a subsequence), then every solution ( S ( x ) , V ( x ) , I A ( x ) , I S ( x ) ) of (45) meets

( S ( x ) , V ( x ) , I A ( x ) , I S ( x ) ) ( S ( x ) , V ( x ) , I A , ( x ) , I S , ( x ) ) uniformly o n D ¯ ,

where I is the normal number, S , V , I A , , I S , > 0 on D ¯ , and ( S , V , I A , , I S , ) satisfies

(67) D 1 Δ S = ( 1 p ( x ) ) Λ ( x ) μ ( x ) S β ( x ) S B 1 , 1 + c 1 B 1 , , x D , D 2 Δ V = p ( x ) Λ ( x ) ( μ ( x ) + η ( x ) ) V σ β ( x ) V B 1 , 1 + c 1 B 1 , , x D , D 4 Δ I S , = ( 1 θ ) β ( x ) S B 1 , 1 + c 1 B 1 , + δ ( x ) I A , ( μ ( x ) + r S ( x ) ) I S , , x D , S ν = V ν = I S , ν = 0 , x D , D θ β ( x ) S B 1 , 1 + c 1 B 1 , + σ β ( x ) V B 1 , 1 + c 1 B 1 , ( μ ( x ) + r ( x ) + δ ( x ) ) I A , d x = 0 .

7 Numerical simulations

With this section, we use numerical examples to validate our theoretical findings and explore the impacts of individuals movement and spatial heterogeneity on 0 and disease dynamics. Suppose that the spatial domain D is one dimensional and select D = [ 0 , π ] . Specifically, we consider the general incidence functions f ( S , B ) = S B ( 1 + c 1 B ) , g ( V , B ) = V B ( 1 + c 2 B ) , where c 1 and c 2 denote the saturation factors. Let’s fix the parameters D 1 = 0.02 , D 2 = 0.05 , D 3 = 0.008 , D 4 = 0.005 , σ = 0.1 , θ = 0.6 , c 1 = 0.000002 , c 2 = 0.000004 , μ ( x ) = 4.5 × 1 0 5 ( 1 + 0.5 sin 3 x ) , p ( x ) = 0.7 ( 1 + 0.2 sin 3 x ) , Λ ( x ) = 10 ( 1 + 0.5 sin 3 x ) , η ( x ) = 0.0011 ( 1 + 0.5 sin 3 x ) , δ ( x ) = 0.23 ( 1 + 0.1 sin 3 x ) , γ A ( x ) = 20 ( 1 + 0.1 sin 3 x ) , γ S ( x ) = 40 ( 1 + 0.1 sin 3 x ) , and initial data

U ( x ) = 2.24 × 1 0 5 20,000 cos 2 x 6,700 250 cos 2 x 80 + 20 cos 2 x 260 + 80 cos 2 x 2.22 × 1 0 4 + 5,000 cos 2 x , x [ 0 , π ] , U = ( S 0 , V 0 , I A 0 , I S 0 , B 0 ) T .

First, we select r A ( x ) = 0.44 ( 1 + 0.1 sin 3 x ) , r S ( x ) = 0.22 ( 1 + 0.1 sin 3 x ) , α ( x ) = 0.35 ( 1 + 0.1 sin 3 x ) and β ( x ) = 9.9 × 1 0 9 ( 1 + 0.5 sin 3 x ) , the remaining parameters are shown earlier. The approach of calculation from [32], it can be derived that 0 0.9824 < 1 . It follows by the Theorem 1 disease-free steady state 0 is globally asymptotically stable. Indeed, we can observe in Figure 2(a) and (b) that the distributions of asymptomatic individuals I A ( x , t ) and symptomatic individuals I S ( x , t ) eventually converge to 0 as time t increases.

Figure 2 
               The spatio-temporal distribution of 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 A
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 t
                              
                              )
                           
                        
                        {I}_{A}\left(x,t)
                     
                   and 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 S
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 t
                              
                              )
                           
                        
                        {I}_{S}\left(x,t)
                     
                   with 
                     
                        
                        
                           
                              
                                 ℛ
                              
                              
                                 0
                              
                           
                           ≈
                           0.9824
                        
                        {{\mathcal{ {\mathcal R} }}}_{0}\approx 0.9824
                     
                  : (a) 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 A
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 t
                              
                              )
                           
                        
                        {I}_{A}\left(x,t)
                     
                   and (b) 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 S
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 t
                              
                              )
                           
                        
                        {I}_{S}\left(x,t)
                     
                  .
Figure 2

The spatio-temporal distribution of I A ( x , t ) and I S ( x , t ) with 0 0.9824 : (a) I A ( x , t ) and (b) I S ( x , t ) .

If the parameters r A ( x ) = 0.35 ( 1 + 0.9 sin 3 x ) , r S ( x ) = 0.18 ( 1 + 0.9 sin 3 x ) , α ( x ) = 0.3 ( 1 + 0.8 sin 3 x ) and β ( x ) = 1.6 × 1 0 8 ( 1 + 0.5 sin 3 x ) are varied, other parameters are fixed as Figure 2. The direct calculation in this case yields 0 1.1069 > 1 . This is as shown by the Theorem 2 that the epidemic is persistent, and there is at least one epidemic steady state for model (1). It is apparent from Figure 3(a)--(h) that S ( x , t ) , I A ( x , t ) , I S ( x , t ) and B ( x , t ) converge to a positive steady state over the entire spatial region. It is also shown in Figure 3(b), (d), (e), (j) that owing to spatial heterogeneity, there are area differences in the distributions of S ( x , t ) , I A ( x , t ) , I S ( x , t ) , and B ( x , t ) across distinct time scales.

Figure 3 
               The persistence of disease for model (1) with 
                     
                        
                        
                           
                              
                                 ℛ
                              
                              
                                 0
                              
                           
                           ≈
                           1.1069
                           >
                           1
                        
                        {{\mathcal{ {\mathcal R} }}}_{0}\approx 1.1069\gt 1
                     
                  : (a), (c), (e), (g): Spatio-temporal evolution of 
                     
                        
                        
                           S
                           
                              (
                              
                                 x
                                 ,
                                 t
                              
                              )
                           
                        
                        S\left(x,t)
                     
                  , 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 A
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 t
                              
                              )
                           
                        
                        {I}_{A}\left(x,t)
                     
                  , 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 S
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 t
                              
                              )
                           
                        
                        {I}_{S}\left(x,t)
                     
                   and 
                     
                        
                        
                           B
                           
                              (
                              
                                 x
                                 ,
                                 t
                              
                              )
                           
                        
                        B\left(x,t)
                     
                  ; (b), (d), (f), (j): 
                     
                        
                        
                           S
                           
                              (
                              
                                 x
                                 ,
                                 t
                              
                              )
                           
                        
                        S\left(x,t)
                     
                  , 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 A
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 t
                              
                              )
                           
                        
                        {I}_{A}\left(x,t)
                     
                  , 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 S
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 t
                              
                              )
                           
                        
                        {I}_{S}\left(x,t)
                     
                  , and 
                     
                        
                        
                           B
                           
                              (
                              
                                 x
                                 ,
                                 t
                              
                              )
                           
                        
                        B\left(x,t)
                     
                   cross-section curves at various times.
Figure 3

The persistence of disease for model (1) with 0 1.1069 > 1 : (a), (c), (e), (g): Spatio-temporal evolution of S ( x , t ) , I A ( x , t ) , I S ( x , t ) and B ( x , t ) ; (b), (d), (f), (j): S ( x , t ) , I A ( x , t ) , I S ( x , t ) , and B ( x , t ) cross-section curves at various times.

Subsequently, we consider the effects of the spatial heterogeneity parameters on disease propagation. In Figure 4(a)--(c) describe the effect of the vaccination rate p ( x ) on the spatial distribution of susceptible individuals S ( x , t ) , asymptomatic individuals I A ( x , t ) and symptomatic individuals I S ( x , t ) at time t = 500 . At this point, one selects p ( x ) = 0.35 ( 1 + p 1 sin 3 x ) , and p 1 is gradually increased from 0, 0.2, 0.4 to 0.6. With increasing spatial heterogeneity in vaccination rates p 1 , S ( x , 500 ) shows greater regional diversity, whereas I A ( x , 500 ) and I S ( x , 500 ) are not significantly fluctuating. This is due to the fact that vaccinated people are to remain at a high danger of contracting the infection. Thus, mass vaccination should be complemented by a focus on the effectiveness of the vaccine. Moreover, from Figure 4(d)–(f), it is convenient to notice that when the spatial heterogeneity of the recovery rate ( r A ( x ) , r S ( x ) ) = ( 0.35 ( 1 + r 1 sin 3 x ) , 0.18 ( 1 + r 2 sin 3 x ) ) changes, i.e., r 1 and r 2 increase from 0, 0.15, 0.3, 0.45 to 0.6, the heterogeneity of I A ( x , 500 ) , I S ( x , 500 ) , and B ( x , 500 ) decreases. In Figure 4(d) and (e), it can be seen that at the same point, e.g., [ 0 , 1 ] , the peaks of I A ( x , 500 ) ( I S ( x , 500 ) ) decrease with the enhancement of r 1 ( r 2 ) , which indicates that peaks in disease outbreaks can be reduced to some extent by increasing the heterogeneous strength of recovery rates. It can be found from Figure 4(g)–(i) that when the spatial heterogeneity intensity ( γ A ( x ) , γ S ( x ) ) = ( 20 ( 1 + γ 1 sin 3 x ) , 40 ( 1 + γ 2 sin 3 x ) ) of γ 1 ( γ 2 ) increases from 0, 0.15, 0.3, 0.45 to 0.6, the regional differences in I A ( x , 500 ) and B ( x , 500 ) become more apparent, while the regional differences in I S ( x , 500 ) weaken, which is due to the fact that most of the infected individuals are asymptomatic, and asymptomatic individuals take some time to become symptomatic individuals. Meanwhile, it can be noticed from Figure 4(j)–(l) that when the intensity of spatial heterogeneity of α ( x ) = 0.3 ( 1 + α 1 sin 3 x ) is varied, i.e., α 1 is increased from 0, 0.12, 0.24 and 0.3 to 0.4, I A ( x , 500 ) , I S ( x , 500 ) and B ( x , 500 ) are all decreasing in spatial heterogeneity, which further suggests that improving local water sanitation and personal hygiene practices is also one way to control diseases.

Figure 4 
               Effect of spatial heterogeneity parameters on the disease distribution of model (1): (a)–(c) effect of 
                     
                        
                        
                           p
                           
                              (
                              
                                 x
                              
                              )
                           
                           
                              (
                              
                                 x
                              
                              )
                           
                        
                        p\left(x)\left(x)
                     
                   on 
                     
                        
                        
                           S
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        S\left(x,500)
                     
                  , 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 A
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        {I}_{A}\left(x,500)
                     
                   and 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 S
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        {I}_{S}\left(x,500)
                     
                  ; (d)–(f) effect of 
                     
                        
                        
                           
                              
                                 r
                              
                              
                                 A
                              
                           
                           
                              (
                              
                                 x
                              
                              )
                           
                        
                        {r}_{A}\left(x)
                     
                   and 
                     
                        
                        
                           
                              
                                 r
                              
                              
                                 S
                              
                           
                           
                              (
                              
                                 x
                              
                              )
                           
                        
                        {r}_{S}\left(x)
                     
                   on 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 A
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        {I}_{A}\left(x,500)
                     
                  , 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 S
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        {I}_{S}\left(x,500)
                     
                   and 
                     
                        
                        
                           B
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        B\left(x,500)
                     
                  ; (g)–(i) effect of 
                     
                        
                        
                           
                              
                                 γ
                              
                              
                                 A
                              
                           
                           
                              (
                              
                                 x
                              
                              )
                           
                        
                        {\gamma }_{A}\left(x)
                     
                   and 
                     
                        
                        
                           
                              
                                 γ
                              
                              
                                 S
                              
                           
                           
                              (
                              
                                 x
                              
                              )
                           
                        
                        {\gamma }_{S}\left(x)
                     
                   on 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 A
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        {I}_{A}\left(x,500)
                     
                  , 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 S
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        {I}_{S}\left(x,500)
                     
                   and 
                     
                        
                        
                           B
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        B\left(x,500)
                     
                  ; (j)–(k) effect of 
                     
                        
                        
                           α
                           
                              (
                              
                                 x
                              
                              )
                           
                        
                        \alpha \left(x)
                     
                   on 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 A
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        {I}_{A}\left(x,500)
                     
                  , 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 S
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        {I}_{S}\left(x,500)
                     
                   and 
                     
                        
                        
                           B
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        B\left(x,500)
                     
                  .
Figure 4

Effect of spatial heterogeneity parameters on the disease distribution of model (1): (a)–(c) effect of p ( x ) ( x ) on S ( x , 500 ) , I A ( x , 500 ) and I S ( x , 500 ) ; (d)–(f) effect of r A ( x ) and r S ( x ) on I A ( x , 500 ) , I S ( x , 500 ) and B ( x , 500 ) ; (g)–(i) effect of γ A ( x ) and γ S ( x ) on I A ( x , 500 ) , I S ( x , 500 ) and B ( x , 500 ) ; (j)–(k) effect of α ( x ) on I A ( x , 500 ) , I S ( x , 500 ) and B ( x , 500 ) .

Next, let’s explore how diffusion coefficients affect disease propagation. Figure 5(a)--(c) displays the spatial distribution of S ( x , t ) , I A ( x , t ) and I S ( x , t ) at t = 500 as the diffusion rate of the susceptible individuals d 1 varies from 0.0002, 0.002, 0.02 to 0.2. It can be seen that S ( x , 500 ) becomes progressively homogeneous throughout the region as D 1 increases, due to the low other diffusion coefficients ( D 2 , D 3 , D 4 ) , the distributions of I A ( x , 500 ) and I S ( x , 500 ) are reliant on the features of local illness propagation. The Figure 5(d)–(f) reveal that the spatial distribution of I A ( x , 500 ) gradually changes from heterogeneous to homogeneous as the diffusion rate of asymptomatic infected individuals D 3 increases from 0.0008, 0.008, 0.08, 0.8 to 8, I A ( x , 500 ) and I S ( x , 500 ) also show decreasing spatial variability and peaks in disease outbreaks, which suggests that restricting the movement of asymptomatic individuals could decrease the level of danger in high-risk areas to some extent, while increasing the level of danger in low-risk areas, this result is similar to that of Figure 5(g)–(i). In addition, we also found that the movement of susceptible individuals has a small effect on the spatial heterogeneity and peak value of the disease distribution in the spreading process of infection in diverse areas, whereas the spreading of infected individuals I A and I S contributes significantly to the spatial heterogeneity of infection. Therefore, it is possible to appropriately liberalize the mobility of susceptible individuals during the spread of the disease, and focusing on restricting the mobility of infected individuals is a necessary means of controlling the disease.

Figure 5 
               Distribution of different diffusion coefficients for 
                     
                        
                        
                           S
                           
                              (
                              
                                 x
                                 ,
                                 t
                              
                              )
                           
                        
                        S\left(x,t)
                     
                  , 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 A
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 t
                              
                              )
                           
                        
                        {I}_{A}\left(x,t)
                     
                  , and 
                     
                        
                        
                           B
                           
                              (
                              
                                 x
                                 ,
                                 t
                              
                              )
                           
                        
                        B\left(x,t)
                     
                   prevalence at 
                     
                        
                        
                           t
                           =
                           500
                        
                        t=500
                     
                  : (a)–(c) 
                     
                        
                        
                           
                              
                                 D
                              
                              
                                 1
                              
                           
                        
                        {D}_{1}
                     
                  ; (d)–(f) 
                     
                        
                        
                           
                              
                                 D
                              
                              
                                 3
                              
                           
                        
                        {D}_{3}
                     
                  ; (g)–(i) 
                     
                        
                        
                           
                              
                                 D
                              
                              
                                 4
                              
                           
                        
                        {D}_{4}
                     
                  .
Figure 5

Distribution of different diffusion coefficients for S ( x , t ) , I A ( x , t ) , and B ( x , t ) prevalence at t = 500 : (a)–(c) D 1 ; (d)–(f) D 3 ; (g)–(i) D 4 .

In addition, let’s be interested in the correlation between the primary parameters and the local reproduction number ˜ 0 ( x ) . Here we choose p ( x ) = p c ( 1 + 0.2 sin 3 x ) , η ( x ) = η c ( 1 + 0.5 sin 3 x ) , r A ( x ) = r c ( 1 + 0.5 sin 3 x ) , r S ( x ) = r c ( 1 + 0.4 sin 3 x ) , δ ( x ) = δ c ( 1 + 0.1 sin 3 x ) , γ A ( x ) = γ c ( 1 + 0.1 sin 3 x ) , γ S ( x ) = γ c ( 1 + 0.2 sin 3 x ) , α ( x ) = α c ( 1 + 0.8 sin 3 x ) , β ( x ) = β c ( 1 + 0.5 sin 3 x ) , and α ( x ) = α c ( 1 + 0.8 sin 3 x ) . As shown in Figure 6(a), it is not uncommon to find that ˜ 0 ( π ) decreases with the increase of p c and increases with the increase of η c , and when η c [ 1 × 1 0 4 , 2 × 1 0 3 ] , the ˜ 0 ( π ) grows faster, while ˜ 0 ( π ) grows more slowly when η c > 2 × 1 0 3 . This also suggests that simply increasing the effectiveness of immunization is not an effective way to control the disease. Furthermore, in Figure 6(b)–(d) also show that at x = π 2 , δ ( x ) , γ A ( x ) , γ S ( x ) and β ( x ) are positively related to ˜ 0 ( x ) , whereas r A ( x ) , r S ( x ) and α ( x ) is a negative correlation. It also demonstrates that infection prevention can be attained by reducing the transmission rate between individuals and pathogens, the transfer rate from asymptomatic to infected individuals, increasing the cure rate of patients, and thoroughly disinfecting the excreta of patients and carriers. In addition, there is some risk of infection transmission due to pathogens in the environment, i.e., γ A ( x ) and γ S ( x ) are positively correlated with ˜ 0 ( x ) , there is a need to lessen γ A ( x ) and γ S ( x ) to reduce the number of pathogens in the environment and thus eradicate the risk of environmental spread.

Figure 6 
               The correlation between the local reproduction number 
                     
                        
                        
                           
                              
                                 
                                    
                                       ℛ
                                    
                                    
                                       ˜
                                    
                                 
                              
                              
                                 0
                              
                           
                           
                              (
                              
                                 x
                              
                              )
                           
                        
                        {\tilde{{\mathcal{ {\mathcal R} }}}}_{0}\left(x)
                     
                   and some parameters: (a) 
                     
                        
                        
                           
                              
                                 
                                    
                                       ℛ
                                    
                                    
                                       ˜
                                    
                                 
                              
                              
                                 0
                              
                           
                           
                              (
                              
                                 x
                              
                              )
                           
                        
                        {\tilde{{\mathcal{ {\mathcal R} }}}}_{0}\left(x)
                     
                   with 
                     
                        
                        
                           p
                           
                              (
                              
                                 x
                              
                              )
                           
                           =
                           
                              
                                 p
                              
                              
                                 c
                              
                           
                           
                              (
                              
                                 1
                                 +
                                 0.2
                                 sin
                                 3
                                 x
                              
                              )
                           
                        
                        p\left(x)={p}_{c}\left(1+0.2\sin 3x)
                     
                   and 
                     
                        
                        
                           η
                           
                              (
                              
                                 x
                              
                              )
                           
                           =
                           
                              
                                 η
                              
                              
                                 c
                              
                           
                           
                              (
                              
                                 1
                                 +
                                 0.5
                                 sin
                                 3
                                 x
                              
                              )
                           
                        
                        \eta \left(x)={\eta }_{c}\left(1+0.5\sin 3x)
                     
                   at 
                     
                        
                        
                           x
                           =
                           π
                        
                        x=\pi 
                     
                  ; (b) 
                     
                        
                        
                           
                              
                                 
                                    
                                       ℛ
                                    
                                    
                                       ˜
                                    
                                 
                              
                              
                                 0
                              
                           
                           
                              (
                              
                                 x
                              
                              )
                           
                        
                        {\tilde{{\mathcal{ {\mathcal R} }}}}_{0}\left(x)
                     
                   with 
                     
                        
                        
                           
                              (
                              
                                 
                                    
                                       r
                                    
                                    
                                       A
                                    
                                 
                                 
                                    (
                                    
                                       x
                                    
                                    )
                                 
                                 ,
                                 
                                    
                                       r
                                    
                                    
                                       S
                                    
                                 
                                 
                                    (
                                    
                                       x
                                    
                                    )
                                 
                              
                              )
                           
                           =
                           
                              
                                 r
                              
                              
                                 c
                              
                           
                           
                              (
                              
                                 1
                                 +
                                 0.5
                                 sin
                                 3
                                 x
                                 ,
                                 1
                                 +
                                 0.4
                                 sin
                                 3
                                 x
                              
                              )
                           
                        
                        \left({r}_{A}\left(x),{r}_{S}\left(x))={r}_{c}\left(1+0.5\sin 3x,1+0.4\sin 3x)
                     
                  , and 
                     
                        
                        
                           δ
                           
                              (
                              
                                 x
                              
                              )
                           
                           =
                           
                              
                                 δ
                              
                              
                                 c
                              
                           
                           
                              (
                              
                                 1
                                 +
                                 0.1
                                 sin
                                 3
                                 x
                              
                              )
                           
                        
                        \delta \left(x)={\delta }_{c}\left(1+0.1\sin 3x)
                     
                   at 
                     
                        
                        
                           x
                           =
                           π
                           ⁄
                           2
                        
                        x=\pi /2
                     
                  ; (c) 
                     
                        
                        
                           
                              
                                 
                                    
                                       ℛ
                                    
                                    
                                       ˜
                                    
                                 
                              
                              
                                 0
                              
                           
                           
                              (
                              
                                 x
                              
                              )
                           
                        
                        {\tilde{{\mathcal{ {\mathcal R} }}}}_{0}\left(x)
                     
                   with 
                     
                        
                        
                           
                              (
                              
                                 
                                    
                                       γ
                                    
                                    
                                       A
                                    
                                 
                                 
                                    (
                                    
                                       x
                                    
                                    )
                                 
                                 ,
                                 
                                    
                                       γ
                                    
                                    
                                       S
                                    
                                 
                                 
                                    (
                                    
                                       x
                                    
                                    )
                                 
                              
                              )
                           
                           =
                           
                              
                                 γ
                              
                              
                                 c
                              
                           
                           
                              (
                              
                                 1
                                 +
                                 0.1
                                 sin
                                 3
                                 x
                                 ,
                                 1
                                 +
                                 0.2
                                 sin
                                 3
                                 x
                              
                              )
                           
                        
                        \left({\gamma }_{A}\left(x),{\gamma }_{S}\left(x))={\gamma }_{c}\left(1+0.1\sin 3x,1+0.2\sin 3x)
                     
                   and 
                     
                        
                        
                           α
                           
                              (
                              
                                 x
                              
                              )
                           
                           =
                           
                              
                                 α
                              
                              
                                 c
                              
                           
                           
                              (
                              
                                 1
                                 +
                                 0.8
                                 sin
                                 3
                                 x
                              
                              )
                           
                        
                        \alpha \left(x)={\alpha }_{c}\left(1+0.8\sin 3x)
                     
                   at 
                     
                        
                        
                           x
                           =
                           π
                           ⁄
                           2
                        
                        x=\pi /2
                     
                  ; (d) 
                     
                        
                        
                           
                              
                                 
                                    
                                       ℛ
                                    
                                    
                                       ˜
                                    
                                 
                              
                              
                                 0
                              
                           
                           
                              (
                              
                                 x
                              
                              )
                           
                        
                        {\tilde{{\mathcal{ {\mathcal R} }}}}_{0}\left(x)
                     
                   with 
                     
                        
                        
                           β
                           
                              (
                              
                                 x
                              
                              )
                           
                           =
                           
                              
                                 β
                              
                              
                                 c
                              
                           
                           
                              (
                              
                                 1
                                 +
                                 0.5
                                 sin
                                 3
                                 x
                              
                              )
                           
                        
                        \beta \left(x)={\beta }_{c}\left(1+0.5\sin 3x)
                     
                   and 
                     
                        
                        
                           α
                           
                              (
                              
                                 x
                              
                              )
                           
                           =
                           
                              
                                 α
                              
                              
                                 c
                              
                           
                           
                              (
                              
                                 1
                                 +
                                 0.8
                                 sin
                                 3
                                 x
                              
                              )
                           
                        
                        \alpha \left(x)={\alpha }_{c}\left(1+0.8\sin 3x)
                     
                  .
Figure 6

The correlation between the local reproduction number ˜ 0 ( x ) and some parameters: (a) ˜ 0 ( x ) with p ( x ) = p c ( 1 + 0.2 sin 3 x ) and η ( x ) = η c ( 1 + 0.5 sin 3 x ) at x = π ; (b) ˜ 0 ( x ) with ( r A ( x ) , r S ( x ) ) = r c ( 1 + 0.5 sin 3 x , 1 + 0.4 sin 3 x ) , and δ ( x ) = δ c ( 1 + 0.1 sin 3 x ) at x = π 2 ; (c) ˜ 0 ( x ) with ( γ A ( x ) , γ S ( x ) ) = γ c ( 1 + 0.1 sin 3 x , 1 + 0.2 sin 3 x ) and α ( x ) = α c ( 1 + 0.8 sin 3 x ) at x = π 2 ; (d) ˜ 0 ( x ) with β ( x ) = β c ( 1 + 0.5 sin 3 x ) and α ( x ) = α c ( 1 + 0.8 sin 3 x ) .

Finally, we analyze in depth the effect of the proportion of asymptomatic infected individuals as a percentage of the whole infected individuals θ and D 3 on the disease dynamics. In Figure 7(a)--(f), it can be observed that as D 3 increases from 0.0008, 0.008, 0.08, 0.8 to 8, I A ( x , 500 ) , I S ( x , 500 ) , and B ( x , 500 ) will have different spatial distributions at θ = 0.1 and θ = 0.9 , i.e., when θ = 0.1 , the diffusion rate D 3 has only a more significant effect on I A ( x , 500 ) only, and very little effect on the other individuals. In contrast, when θ = 0.9 , D 3 has a large effect on every individuals. It also suggests that an increase in the percentage of asymptomatic individuals intensifies the spread of the disease, making it more difficult to control. Therefore, special attention should be devoted to the proportion of hidden infections to those who contract the disease. When θ is small, a policy of slow liberalization and gradual return to normal life can be adopted for groups in low-risk areas, while for high-risk areas, rapid screening of asymptomatic infected individuals is a priority.

Figure 7 
               The effects of 
                     
                        
                        
                           σ
                        
                        \sigma 
                     
                   and 
                     
                        
                        
                           
                              
                                 D
                              
                              
                                 3
                              
                           
                        
                        {D}_{3}
                     
                   on the spatial distribution of disease at 
                     
                        
                        
                           t
                           =
                           500
                        
                        t=500
                     
                  : (a), (c), (e) the effects on 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 A
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        {I}_{A}\left(x,500)
                     
                  , 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 S
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        {I}_{S}\left(x,500)
                     
                   and 
                     
                        
                        
                           B
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        B\left(x,500)
                     
                   for different values of 
                     
                        
                        
                           
                              
                                 D
                              
                              
                                 3
                              
                           
                        
                        {D}_{3}
                     
                   with 
                     
                        
                        
                           σ
                           =
                           0.1
                        
                        \sigma =0.1
                     
                  ; (b), (d), (f) the effects on 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 A
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        {I}_{A}\left(x,500)
                     
                  , 
                     
                        
                        
                           
                              
                                 I
                              
                              
                                 S
                              
                           
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        {I}_{S}\left(x,500)
                     
                   and 
                     
                        
                        
                           B
                           
                              (
                              
                                 x
                                 ,
                                 500
                              
                              )
                           
                        
                        B\left(x,500)
                     
                   for different values of 
                     
                        
                        
                           
                              
                                 D
                              
                              
                                 3
                              
                           
                        
                        {D}_{3}
                     
                   with 
                     
                        
                        
                           σ
                           =
                           0.9
                        
                        \sigma =0.9
                     
                  .
Figure 7

The effects of σ and D 3 on the spatial distribution of disease at t = 500 : (a), (c), (e) the effects on I A ( x , 500 ) , I S ( x , 500 ) and B ( x , 500 ) for different values of D 3 with σ = 0.1 ; (b), (d), (f) the effects on I A ( x , 500 ) , I S ( x , 500 ) and B ( x , 500 ) for different values of D 3 with σ = 0.9 .

8 Conclusion

To this end, we proposed a model of degenerate cholera with asymptomatic individuals in spatially heterogeneous environments that takes into account the rate of immune loss and general morbidity, and assumes that environmental viruses do not disperse. We studied the existence of global positive solutions to the model and proved that Φ ( t ) is ultimately bounded (Lemmas 1 and 2). Since model (1) lacks tightness, the τ -contraction method is utilized to justify the asymptotic smoothness of Φ ( t ) , which proves the existence of a globally tight attractor for Φ ( t ) (see Lemma 3). Furthermore, we identify the basic reproduction number 0 , give its exact expression, which is a major innovation of this article. By using 0 as the threshold parameter, the dynamical behavior of the model is studied. More precisely, the disease-free steady state is globally asymptotically stable if 0 1 (Theorems 1 and 3); the disease is uniformly persistent if 0 > 1 (see Theorem 2). Finally, it is also discussed that the asymptotic behavior of the endemic steady state as d 1 and d 3 tends to 0 or , respectively (see Theorems 4 and 5, Theorems 6 and 7).

In the numerical simulation section, except for simulating the extinction and persistence of disease (Figures 2 and 3), the impacts of some crucial parameters on the spatial and temporal distributions of the disease are also explored. To be more specific, the spatial heterogeneity of model will lead to a regional variability in the spatial distribution of disease (Figure 4). Furthermore, numerical simulations show that the spread of susceptible individuals does not modify the risk level of local outbreaks when the spread of asymptomatic and infected individuals is relatively small (Figure 5(a)–(c)). This is consistent with the conclusions of [37], which stated that cholera cannot be controlled by limiting the movement of susceptible individuals. We also found the spread of asymptomatic and symptomatic individuals suppresses spatial heterogeneity so that each individual is less differentiated across the region (Figure 5(d)–(i)). Therefore, it is not advisable to entirely limit the mobility of people during epidemic periods of the disease. Further, we also analyze how the proportion of asymptomatic infected individuals to the whole infected individuals θ and d 3 affects the spread of disease (Figure 7(a)–(f)), and the results show that the proportion of asymptomatic individuals θ can change the dynamics of the epidemic and performs a critical role in the process of the spread of the disease. In real life, asymptomatic individuals are asymptomatic, which makes them unaware of whether they are infected, which can eventually lead to large-scale outbreaks of the disease, a finding similar to that in [36]. In Figure 6, we also find that thorough disinfection of the excreta of patients and carriers, as well as the surrounding area are important ways tocontrol and eradicate the risk of environmental spread.

Unfortunately, in this article, we only considered the effects of environment spread on the disease and ignored host-to-host spread (direct transmission), which exists in real situations. However, we know that cholera outbreaks are also related to seasons and temperatures, therefore, a response diffusion cholera model with cyclical and climatic variations would be more meaningful and the results would be more effective in preventing and controlling the disease.



Acknowledgements

We would also gratefully like to thank the editor and anonymous reviewers for their careful review and helpful suggestions, which improved the research in this thesis.

  1. Funding information: This research is partially supported by the Tianshan Talent Training Program (Grant No. 2022TSYCCX0015), the National Natural Science Foundation of China (Grant No. 12361103).

  2. Author contributions: All authors have accepted responsibility for the entire content of this manuscript and consented to its submission to the journal, reviewed all the results and approved the final version of the manuscript. Shengfu Wang: formal analysis, conceptualization, writing-original draft, writing–review & editing. Linfei Nie: funding acquisition, supervision, and editing.

  3. Conflict of interest: The authors state no conflict of interest.

  4. Data availability statement: Data sharing is not applicable to this article as no data sets were generated or analyzed during the current study.

References

[1] J. R. Andrews and S. Basu, Transmission dynamics and control of cholera in Haiti: An epidemic model, Lancet 377 (2011), 1248–1255. 10.1016/S0140-6736(11)60273-0Search in Google Scholar PubMed PubMed Central

[2] J. Arino, C. C. McCluskey, and P. van den Driessche, Global results for an epidemic model with vaccination that exhibits backward bifurcation, SIAM J. Appl. Math. 64 (2003), 260–276. 10.1137/S0036139902413829Search in Google Scholar

[3] N. Bai, C. Song, and R. Xu, Mathematical analysis and application of a cholera transmission model with waning vaccine-induced immunity, Nonlinear Anal-Real. 58 (2021), 103232. 10.1016/j.nonrwa.2020.103232Search in Google Scholar

[4] Z. Bai and L. Han, A partially degenerate reaction-diffusion cholera model with temporal and spatial heterogeneity, Appl. Anal. 102 (2023), no. 11, 3167–3184. 10.1080/00036811.2022.2057302Search in Google Scholar

[5] L. Cai, G. Fan, C. Yang, and J. Wang, Modeling and analyzing cholera transmission dynamics with vaccination age, J. Franklin I. 357 (2020), no. 12, 8008–8034. 10.1016/j.jfranklin.2020.05.030Search in Google Scholar PubMed PubMed Central

[6] V. Capasso and S. Paveri-Fontana, A mathematical model for the 1973 cholera epidemic in the European Mediterranean region, Rev. épidé miol. Santé Publique 27 (1979), no. 2 121–132. Search in Google Scholar

[7] C. T. Codeço, Endemic and epidemic dynamics of cholera: The role of the aquatic reservoir, BMC Infect. Dis. 1 (2001), no. 1, 1–14. 10.1186/1471-2334-1-1Search in Google Scholar PubMed PubMed Central

[8] R. Cui, K. Y. Lam, and Y. Lou, Dynamics and asymptotic profiles of steady states of an epidemic model in advective environments, J. Differ. Equations 263 (2017), 2343–2373. 10.1016/j.jde.2017.03.045Search in Google Scholar

[9] R. Cui and Y. Lou, A spatial SIS model in advective heterogeneous environments, J. Differ. Equations 261 (2016), 3305–3343. 10.1016/j.jde.2016.05.025Search in Google Scholar

[10] Y. Du, R. Peng, and M. Wang, Effect of a protection zone in the diffusive Leslie predator-prey model. J. Differ. Equations 246 (2009), 3932–3956. 10.1016/j.jde.2008.11.007Search in Google Scholar

[11] K. J. Engel and R. Nagel, One-parameter Semigroups for Linear Evolution Equations, Springer-Verlag, New York, 1999. Search in Google Scholar

[12] D. Gilbarg and N. S. Trudinger, Elliptic Partial Differential Equation of Second Order, Springer, Berlin, 2001. 10.1007/978-3-642-61798-0Search in Google Scholar

[13] J. K. Hale, Asymptotic Behavior of Dissipative Systems, American Mathematical Society, Providence, RI, 1988. Search in Google Scholar

[14] R. I. Joh, H. Wang, H. Weiss, and J. S. Weitz, Dynamics of indirectly transmitted infectious diseases with immunological threshold, B. Math. Biol. 71 (2009), no. 4, 845–862. 10.1007/s11538-008-9384-4Search in Google Scholar PubMed

[15] A. A. King, E. L. Ionides, M. Pascual, and M. J. Bouma, Inapparent infections and cholera dynamics, Nature 454 (2008), 877–881. 10.1038/nature07084Search in Google Scholar PubMed

[16] C. M. Kribs-Zaleta and J. X. Velasco-Hernández, A simple vaccination model with multiple endemic states, Math. Biosci. 164 (2000), 183–201. 10.1016/S0025-5564(00)00003-1Search in Google Scholar

[17] Y. Lou and X. Q. Zhao, A reaction-diffusion malaria model with incubation period in the vector population, J. Math. Biol. 62 (2011), 543–568. 10.1007/s00285-010-0346-8Search in Google Scholar PubMed

[18] Y. Lou and W. Ni, Diffusion, self-diffusion and cross-diffusion, J. Differ Equations 131 (1996), 79–131. 10.1006/jdeq.1996.0157Search in Google Scholar

[19] Y. Lou, On the effects of migration and spatial heterogeneity on single and multiple species. J. Differ. Equations 223 (2006), 400–426. 10.1016/j.jde.2005.05.010Search in Google Scholar

[20] P. Magal and X. Zhao, Global attractors and steady states for uniformly persistent dynamical systems, SIAM. J. Math. Anal. 37 (2005), 251–275. 10.1137/S0036141003439173Search in Google Scholar

[21] R. H. Martin and H. L. Smith, Abstract functional differential equations and reaction-diffusion systems, T. Am. Math. Soc. 321 (1990), 1–44. 10.1090/S0002-9947-1990-0967316-XSearch in Google Scholar

[22] E. J. Nelson, J. B. Harris, J. G. Morris, S. B. Calderwood and A. Camilli, Cholera transmission: the host, pathogen and bacteriophage dynamic, Nat. Rev. Microbiol. 7 (2009), no. 10, 693–702. Search in Google Scholar

[23] E. J. Nelson, J. B. Harris, J. GlennMorris Jr., S. B. Calderwood and A. Camilli, Cholera transmission: the host, pathogen and bacteriophage dynamic, Nat. Rev. Microbiol., 7 (2009), 693–702. 10.1038/nrmicro2204Search in Google Scholar PubMed PubMed Central

[24] W. Ni and I. Takagi, On the Neumann problem for some semilinear elliptic equations and systems of activator-inhibitor type, T. Am. Math. Soc. 297 (1986), 351–368. 10.1090/S0002-9947-1986-0849484-2Search in Google Scholar

[25] R. D. Nussbaum, Eigenvectors of nonlinear positive operator and the linear Krein-Rutman theorem, Fixed Point Theory, Lecture Notes in Mathematics, vol. 886, Springer, New York, 1981, p. 309–331. 10.1007/BFb0092191Search in Google Scholar

[26] B. O. Ogola, W. A. Woldegerima, and E. O. Omondi, Parameter and state estimation in a cholera model with threshold immunology: A case study of Senegal, B. Math. Biol. 82 (2020), no. 6, 72. 10.1007/s11538-020-00755-6Search in Google Scholar PubMed

[27] R. Peng, J. Shi, and M. Wang, On stationary patterns of a reaction-diffusion model with autocatalysis and saturation law, Nonlinearity 21 (2008), 1471–1488. 10.1088/0951-7715/21/7/006Search in Google Scholar

[28] H. L. Smith and X. Zhao, Robust persistence for semidynamical systems, Nonlinear Anal-theor. 47 (2001), 6169–6179. 10.1016/S0362-546X(01)00678-2Search in Google Scholar

[29] H. R. Thieme, Spectral bound and reproduction number for infinite-dimensional population structure and time heterogeneity, SIAM J. Appl. Math. 70 (2009), 188–211. 10.1137/080732870Search in Google Scholar

[30] J. H. Tien and D. J. Earn, Multiple transmission pathways and disease dynamics in a waterborne pathogen model, B. Math. Biol. 72 (2010), no. 6, 1506–1533. 10.1007/s11538-010-9507-6Search in Google Scholar PubMed

[31] A. R. Tuite, J. Tien, M. Eisenberg, D. J. D. Earn, J. Ma and D. N. Fisman, Cholera epidemic in Haiti, 2010: using a transmission model to explain spatial spread of disease and identify optimal control interventions, Ann. Intern. Med. 154 (2011), 593–601. 10.7326/0003-4819-154-9-201105030-00334Search in Google Scholar PubMed

[32] W. Wang and X. Zhao, Basic reproduction numbers for reaction-diffusion epidemic models, SIAM J. Appl. Dyn. Syst. 11 (2012), 1652–1673. 10.1137/120872942Search in Google Scholar

[33] M. Wang, Nonlinear Second Order Parabolic Equations, CRC Press, New York, 2021. 10.1201/9781003150169Search in Google Scholar

[34] W. Wang and Z. Feng, Influence of environmental pollution to a waterborne pathogen model: Global dynamics and asymptotic profiles, Commun. Nonlinear Sci. 99 (2021), 105821. 10.1016/j.cnsns.2021.105821Search in Google Scholar

[35] J. Wang and X. Wu, Dynamics and profiles of a diffusive cholera model with bacterial hyperinfectivity and distinct dispersal rates, J. Dyn. Differ. Equ. 35 (2023), no. 2, 1205–1241. 10.1007/s10884-021-09975-3Search in Google Scholar

[36] J. Wang and H. Lu, Dynamics and profiles of a degenerated reaction-diffusion host-pathogen model with apparent and inapparent infection period, Commun. Nonlinear Sci. 125 (2023), 107318. 10.1016/j.cnsns.2023.107318Search in Google Scholar PubMed PubMed Central

[37] J. Wang, F. Xie, and T. Kuniya, Analysis of a reaction-diffusion cholera epidemic model in a spatially heterogeneous environment, Commun. Nonlinear Sci. 80 (2020), 104951. 10.1016/j.cnsns.2019.104951Search in Google Scholar

[38] World Health Organization, “Newsroom Fact sheets Detail Cholera,” WHO. Accessed: Feb. 5, 2021. [Online]. Available: https://www.who.int/en/news-room/fact-sheets/detail/cholera. Search in Google Scholar

[39] Y. Wu and X. Zou, Dynamics and profiles of a diffusive host-pathogen system with distinct dispersal rates, J. Differ. Equations 264 (2018), 4989–5024. 10.1016/j.jde.2017.12.027Search in Google Scholar

[40] Q. Ye and Z. Li, Introduction to Reaction Diffusion Equations, Science Public, Beijing, 2011. Search in Google Scholar

[41] L. Zhang, Z. Wang, and Y. Zhang, Dynamics of a reaction-diffusion waterborne pathogen model with direct and indirect transmission, Comput. Math. Appl. 72 (2016), no. 1, 202–215. 10.1016/j.camwa.2016.04.046Search in Google Scholar

Received: 2024-04-24
Accepted: 2024-11-15
Published Online: 2024-12-18

© 2024 the author(s), published by De Gruyter

This work is licensed under the Creative Commons Attribution 4.0 International License.

Articles in the same Issue

  1. Regular Articles
  2. Existence of normalized peak solutions for a coupled nonlinear Schrödinger system
  3. Orbital stability of the trains of peaked solitary waves for the modified Camassa-Holm-Novikov equation
  4. Global boundedness in a two-dimensional chemotaxis system with nonlinear diffusion and singular sensitivity
  5. Existence and uniqueness of solution for a singular elliptic differential equation
  6. The bounded variation capacity and Sobolev-type inequalities on Dirichlet spaces
  7. Vanishing and blow-up solutions to a class of nonlinear complex differential equations near the singular point
  8. Existence, uniqueness, localization and minimization property of positive solutions for non-local problems involving discontinuous Kirchhoff functions
  9. Concentration phenomena for a fractional relativistic Schrödinger equation with critical growth
  10. Beyond the classical strong maximum principle: Sign-changing forcing term and flat solutions
  11. Monotonicity of solutions for parabolic equations involving nonlocal Monge-Ampère operator
  12. On a nonlinear Robin problem with an absorption term on the boundary and L1 data
  13. Multiple solutions for the quasilinear Choquard equation with Berestycki-Lions-type nonlinearities
  14. Gradient estimates for a class of higher-order elliptic equations of p-growth over a nonsmooth domain
  15. Global existence and decay estimates of the classical solution to the compressible Navier-Stokes-Smoluchowski equations in ℝ3
  16. Global existence and finite-time blowup for a mixed pseudo-parabolic r(x)-Laplacian equation
  17. Multiple positive solutions for a class of concave-convex Schrödinger-Poisson-Slater equations with critical exponent
  18. A minimization problem with free boundary for p-Laplacian weakly coupled system
  19. Multiplicity of semiclassical solutions for a class of nonlinear Hamiltonian elliptic system
  20. k-convex solutions for multiparameter Dirichlet systems with k-Hessian operator and Lane-Emden type nonlinearities
  21. Infinitely many solutions for Hamiltonian system with critical growth
  22. On optimal control in a nonlinear interface problem described by hemivariational inequalities
  23. Variational–hemivariational system for contaminant convection–reaction–diffusion model of recovered fracturing fluid
  24. Potential and monotone homeomorphisms in Banach spaces
  25. Critical fractional Schrödinger-Poisson systems with lower perturbations: the existence and concentration behavior of ground state solutions
  26. Supercritical Hénon-type equation with a forcing term
  27. Concentration of blow-up solutions for the Gross-Pitaveskii equation
  28. Mountain-pass-type solutions for Schrödinger equations in R2 with unbounded or vanishing potentials and critical exponential growth nonlinearities
  29. Regularity for critical fractional Choquard equation with singular potential and its applications
  30. Double phase anisotropic variational problems involving critical growth
  31. Normalized solutions of NLS equations with mixed nonlocal nonlinearities
  32. Nontrivial solutions for resonance quasilinear elliptic systems
  33. Standing waves for Choquard equation with noncritical rotation
  34. Low regularity conservation laws for Fokas-Lenells equation and Camassa-Holm equation
  35. Modified quasilinear equations with strongly singular and critical exponential nonlinearity
  36. Limit profiles and the existence of bound-states in exterior domains for fractional Choquard equations with critical exponent
  37. Nests of limit cycles in quadratic systems
  38. Regularity of minimizers for double phase functionals of borderline case with variable exponents
  39. Boundedness, existence and uniqueness results for coupled gradient dependent elliptic systems with nonlinear boundary condition
  40. Ground state solutions for magnetic Schrödinger equations with polynomial growth
  41. Nonoccurrence of Lavrentiev gap for a class of functionals with nonstandard growth
  42. Dynamics for wave equations connected in parallel with nonlinear localized damping
  43. Boussinesq's equation for water waves: Asymptotics in Sector I
  44. Optimal global second-order regularity and improved integrability for parabolic equations with variable growth
  45. Normalized solutions of Schrödinger equations involving Moser-Trudinger critical growth
  46. Normalized solutions for the double-phase problem with nonlocal reaction
  47. Normalized solutions for Sobolev critical fractional Schrödinger equation
  48. Well-posedness of Cauchy problem of fractional drift diffusion system in non-critical spaces with power-law nonlinearity
  49. Improved results on planar Klein-Gordon-Maxwell system with critical exponential growth
  50. Existence and multiplicity of solutions for a new p(x)-Kirchhoff equation
  51. Quasiconvex bulk and surface energies: C1,α regularity
  52. Time decay estimates of solutions to a two-phase flow model in the whole space
  53. Bounds for the sum of the first k-eigenvalues of Dirichlet problem with logarithmic order of Klein-Gordon operators
  54. The properties of a new fractional g-Laplacian Monge-Ampère operator and its applications
  55. A fractional profile decomposition and its application to Kirchhoff-type fractional problems with prescribed mass
  56. Cauchy problem for a non-Newtonian filtration equation with slowly decaying volumetric moisture content
  57. Multiplicity of normalized semi-classical states for a class of nonlinear Choquard equations
  58. The second Yamabe invariant with boundary
  59. Peaked solitary waves and shock waves of the Degasperis-Procesi-Kadomtsev-Petviashvili equation
  60. Normalized solutions for the Choquard equations with critical nonlinearities
  61. Boundedness and long-time behavior in a parabolic-elliptic system arising from biological transport networks
  62. Initial boundary value problem and exponential stability for the planar magnetohydrodynamics equations with temperature-dependent viscosity
  63. Nonexistence of mean curvature flow solitons with polynomial volume growth immersed in certain semi-Riemannian warped products
  64. Analysis of a vector-borne disease model with vector-bias mechanism in advective heterogeneous environment
  65. Normalized solutions for the Kirchhoff equation with combined nonlinearities in ℝ4
  66. Nonexistence of the compressible Euler equations with space-dependent damping in high dimensions
  67. Multiplicity of solutions for a nonhomogeneous quasilinear elliptic equation with concave-convex nonlinearities
  68. On semilinear inequalities involving the Dunkl Laplacian and an inverse-square potential outside a ball
  69. Besov regularity for the elliptic p-harmonic equations in the non-quadratic case
  70. Uniqueness and nondegeneracy of ground states for Δ u + u = ( I α u 2 ) u in R 3 when α is close to 2
  71. Existence of solutions for a class of quasilinear Schrödinger equations with Choquard-type nonlinearity
  72. Weighted Hardy-Adams inequality on unit ball of any even dimension
  73. Existence and regularity for a p-Laplacian problem in ℝN with singular, convective, and critical reaction
  74. Temporal periodic solutions of non-isentropic compressible Euler equations with geometric effects
  75. Nodal solutions for a zero-mass Chern-Simons-Schrödinger equation
  76. The modified quasi-boundary-value method for an ill-posed generalized elliptic problem
  77. Nonlocal heat equations with generalized fractional Laplacian
  78. Choquard equations with recurrent potentials
  79. Local and global well-posedness of the Maxwell-Bloch system of equations with inhomogeneous broadening
  80. The Riemann problem for two-layer shallow water equations with bottom topography
  81. Global stability and asymptotic profiles of a partially degenerate reaction diffusion Cholera model with asymptomatic individuals
  82. On Kirchhoff-Schrödinger-Poisson-type systems with singular and critical nonlinearity
  83. Energy-variational solutions for viscoelastic fluid models
  84. Stability on 3D Boussinesq system with mixed partial dissipation
  85. Existence and multiplicity results for non-autonomous second-order Hamiltonian systems
  86. Erratum
  87. Erratum to “Normalized solutions for the Kirchhoff equation with combined nonlinearities in ℝ4
Downloaded on 7.9.2025 from https://www.degruyterbrill.com/document/doi/10.1515/anona-2024-0059/html
Scroll to top button