ABSTRACT

The Illinois Junior Academy of Science

This form/paper may not be taken without IJAS authorization.

CATEGORY Environmental Science STATE REGION # 6

SCHOOL Niles North High School IJAS SCHOOL # 6038
CITY/ZIP Skokie, IL SPONSOR CELL PHONE # 8475022672
SPONSOR Christi Camel SPONSOR E-MAIL chrcam@d219.org

MARK ONE: EXPERIMENTAL INVESTIGATION ​☐ DESIGN INVESTIGATION ​✓

NAME OF SCIENTIST* Sharlene Faye Manlongat GRADE 12

NAME OF SCIENTIST GRADE
NAME OF SCIENTIST GRADE
NAME OF SCIENTIST GRADE

* If this project is awarded a monetary prize, the check will be written in this scientist's name, and it will be his/her responsibility to distribute the
prize money equally among all participating scientists.

Math Model the Effects of Deforestation & Anthropization on the Transmission Rates of
PROJECT TITLE Bat-Related Diseases

Purpose: To create a math model that illustrates the effect of deforestation and anthropization on the population of bats and
subsequent transmission of bat-related diseases to humans.

Procedure:

First two subsystems were created based on the basic logistic population growth model and SEIR model for diseases. The
logistic population growth model was redesigned to incorporate rates of deforestation and anthropization and to deliver an
output of varying subpopulations concentrations. The second subsystem incorporated the behavior responses of bats and
delivered a basic reproduction number. These two subsystems were simulated with assumed data from previous research
papers, then redesigned to align with the performance criteria.

Conclusion:

The math model successfully connected the rate of deforestation and anthropization to the transmission rate of diseases. It
effectively delivered varying basic reproduction numbers based on different environmental factors the bats endured. It also
successfully incorporated behaviors of bats and illustrated the change in subpopulation distribution. However, although the
model delivers an estimation for R​0​, the parameters set are still based on several assumptions. This model strongly encourages
more advanced and intricate data collection on the behaviors and population of bats to accurately estimate the R​0​.
1) Limit Abstract to 3 paragraphs (about 250 words or less). a) Purpose - what you set out to investigate; b) Procedure - how you did
it; c) Conclusion - based on your results. Label each paragraph.
2) Must be typed, single-spaced on the front of this form. Do not write on the back of this form.
3) Three copies of your complete paper are required at the State Science Project Exposition.
Four copies of your complete paper are required for the State Paper Session Competition.

This form must be used. ​This form ​must ​be displayed on the front of the exhibitor’s display board. It may be reduced to half a sheet of paper; 8.5 inches (vertical) X 5.5 inche
(horizontal).
 Math Model of the Effect of
Deforestation & Anthropization on the
Transmission Rates of Bat-Related
Diseases
Design Project

Sharlene Faye Manlongat

Niles North High School

 Manlongat 2

Table of Contents
Acknowledgements………………………………………………………………………………..3
Problem…………………………………………………………………………………………...4
Review of Literature……………………………………………………………………………5
Performance Criteria ………….……………………………………………………………….15
Design Plan……………………………………………………………………………………..16
Simulation Model 1 …………………………………………………………………………....22
Summary of Changes ………………………………………………………………………....26
Simulation Model 2 …………………………………………………………………….……....30
Evaluation & Conclusion………………………………………………………………………...33
References………………………………………………………………………………………..37
 Manlongat 3

Acknowledgements
I would like to express my gratitude to my peers, friends, and family who never ceased to

support me throughout the whole process. To my closest confidantes, Lisa Duan, Ana Bojinov,

and Sofia Ford, I would like to say thank you for offering ideas and reassurance repeatedly

throughout the entire process. To my supportive and patient supervisor, Mrs. Christine Camel, I

give my utmost appreciation for without your time and insightful contributions I would not be

able to complete this project. Thank you for all your suggestions and constant support that

enabled me to complete this project.

 Manlongat 4

Problem
Bat-related diseases take up a large proportion of emerging diseases in humans, thus

posing a danger for the future. At the moment, the Coronavirus Pandemic has arisen as a

bat-related disease and has upset the entire world. Specifically, given the innate immunity in

bats, they may harbor more pathogens resulting in greater damaging economic and social costs.

In addition, with the increase in land alterations and reductions in biodiversity, the implications

on disease transmission are imperative to observe. At the moment, there exists a lack of data on

the specific mechanisms between biodiversity reduction through deforestation and anthropization

on viral spillover events for scientists to refer to when anticipating “new” pathogens. A

mathematical model is needed to understand how varying levels of deforestation and

anthropization play a role in the transmission of diseases and the potential for viral spillover

events.
 Manlongat 5

Review of Literature
Zoonotic viruses, viral diseases that are transmitted between vertebrates and humans,

account for nearly ​75% of emerging pathogens from infectious organisms that are harmful to

humans (Gebreyes et al., 2014).​The most notable zoonotic viruses include the Nipah virus,

lyssavirus, Marburg virus, Ebola virus, and coronavirus which are all associated with various bat

species. The increase of frequencies with bat-related disease transmissions can be attributed to a

reduction in biodiversity, brought on by factors like deforestation and anthropization (Kessing et

al., 2020). Not only does biodiversity loss imply extensive negative environmental implications,

but there are severe social and economic consequences as well. Direct costs of zoonoses include

reduction of livestock production and sources of income, which is highly amplified in

low-income countries (Martins, Hasler, & Rushton, 2014). Besides, zoonoses contribute to 1% of

disability-adjusted life years (DALY which is a value of future years lost due to premature death

or poor health from disease) lost to infectious diseases in high-income families and 26% of

DALYs lost to infectious diseases in low-income countries ​(Martins, Hasler, & Rushton, 2014)​.

Along with the rise in diseases comes the rise in vaccination programs and other prevention

methods that can be costly and difficult to implement across political and environmental

boundaries. Given the social, economic, and health implications of viral spillover events, it is

imperative to construct a mathematical model ​that will aid scientists in predicting the fatality,

causes, and factors of zoonotic outbreaks and lead to more effective prevention methods and a

deeper understanding to combat any diseases that may arise.

 Manlongat 6

Zoonotic viral spillovers are formally defined as the event in which there is transmission

of a pathogen between a vertebrate animal to a human. Although they are sporadic, they have

been found to cause nearly two-thirds of emerging diseases (Woolhouse et al., 2012). Despite the

fact each viral spillover event is inherently unique, there are three general aspects of zoonotic

viral spillovers; pathogen pressure, exposure, and the probability of infection within the host

community. Before pathogen pressure, pathogens are distributed and intensify within the

reservoir hosts. Oftentimes, pathogens are not transmitted to humans, and if they are, the result is

rarely highly infectious epidemics or pandemics, thus adding another layer of complexity to our

understanding. After pathogens take root within reservoir hosts, there are three ways in which

they can be “released” from the hosts; excretion, slaughter, and vector-borne (an organism,

usually ones with biting tendencies, transmits pathogens to another animal or plant). Then

pathogenic pressure builds based on the density of reservoir hosts, infection prevalence, and

infection intensity within infected reservoir hosts both in space and time. Both the hosts’ natural

infection history and movement/behavior of original hosts interact with demographic turnover,

spatial distribution, and abundance to determine how efficient the pathogen will spread. The

aforementioned factors culminate in the intensity of the pathogen the host will contract.

Exposure to pathogens, particularly from wildlife organisms, frequently occurs around human

dwellings or in spaces that host occupations that are frequently exposed to animals such as

hunters, veterinarians, researchers, and wildlife management (Johnson et al., 2015). Once the

pathogen penetrates the barriers between hosts, such as skin or physical barriers, the infection

may result in a range of outcomes from the elimination of the pathogenic microorganism to the

death of the new host. In addition, humans who contract a pathogen may contribute to
 Manlongat 7

human-to-human transmission which drives recent public health crises and contributes to longer

spillovers.

One of the foremost species of hosts involved in zoonotic viral spillovers is bats. The

most notable zoonotic viruses include the Nipah virus, lyssavirus, Marburg virus, Ebola virus,

and coronavirus which are all

associated with various bat species

(Letko et al., 2020) (figure 1). Bats

are one of the most diverse

mammalian orders that reside on

almost every continent. Bats can be

described as natural reservoir hosts

for emerging diseases (Letko et al., 2020). The increase in bat-related viral family infections is

rooted in their behavior and biological characteristics. Bats have both innate immunity,

high-energy metabolic flights that increase body temp levels, and adaptive immunity, heightened

intolerance, which explains why bats appear to be asymptomatic in response to infections (Letko

et al., 2020). There is a pattern in bats in which they are persistently infected, thus their immune

competence controls shedding and develops a greater immunity. Given these behavioral qualities

and their widespread population, bat-related viruses may amount to higher chances of viral

spillovers. However, despite the fact that bats are recognized as optimal host reservoirs, host

cycles of infections within bats continue to be difficult to track as data is either lacking or

difficult to interpret. This poses a greater issue coupled with the fact that there is a growing list

of recent bat-related zoonotic spillover events, including direct bat-to-human situations. The
 Manlongat 8

relationship of these events to bats is supported by epidemiological and molecular evidence. In

regards to the SARS-CoV outbreak, the disease has been connected to bats with molecular

evidence from intermediate hosts. Coronaviruses in particular are driven to mutate and adapt to

human hosts’ environments shown through their high rates of genetic recombination (Letko et

al., 2020). The ability of the diseases within bats to rapidly mutate through recombination

enables them to overcome barriers in host species, thus endangering the lives of several species,

including humans.

In relation to deforestation, bats are often subcategorized based on their roosting habits as

cave-roosting or forest-roosting bats, meaning bats will congregate to rest in either a cave or

forest environment. The focus of the research paper relates to forest-roosting bats and their

response to deforestation within their environment. In connection to the transmission of diseases

from bats to humans, oftentimes forest-roosting or forest-foraging bats may contaminate “drip”

zones around trees where they roost and feed by excreting feces, saliva, urine, and other bodily

functions (Han et al., 2015). These drip zones can span for several acres and miles owing to the

flight and foraging behaviors of bats that enable

them to find resources beyond their roosting site.

In a prior research study, a couple particular

forest-roosting bats native to Southeast Asia were

observed to record their responses to deforestation

within the environment. In figure 2, both

forest-roosting (d-f) and cave-roosting (b & c) bat

species were observed and their travel distances

 Manlongat 9

and frequencies were recorded. The distribution of frequency for forest-roosting species shows a

greater skew towards shorter distances than for cave-roosting species. The researchers concluded

that the evidence suggests forest-roosting species forage close to their roosts but on average a

max of 550 m (0.135 acres) (Kingston, 2013). Notably the bats observed are of the species

Rhinolophus, commonly known as Horseshoe bats. Although there is a great deficit in the

collection of data for bats of all species, in terms of population density in varying areas of the

world, a research study in Thailand observed a colony of approximately 300 bats, all of the

species Rhinolophus. Several Rhinolophus bats have been observed to have strains of

SARS-CoV, which will be discussed in the next section. Notably, 6.84% of bats have been

detected and assumed for carrying coronavirus strains, especially Rhinolophus bats with

SARS-related Rhinolophus bat coronaviruses (SARSr-Rh-BatCoV​), ​making this species worth

studying in depth in the up-coming future (Lin et al., 2017).

Aside from behavioral factors, environmental alterations factor into the increasing

likelihood of zoonotic outbreaks. Ecosystems and habitats maintain a level of biodiversity, which

encompasses a diversity of genes and species, to function (Keesing et al., 2010). In recent years,

it has been concluded that a reduction in biodiversity frequently increases disease transmission

(Keesing et al., 2010). Biodiversity loss can be brought on by human interaction that includes

inducing climate change, pollution, deforestation, and anthropization. There is a common

misconception that deforestation and anthropization, the conversion of open spaces for primarily

agricultural reasons, decrease the risk of viral spillovers because it leads to the disappearance of

species. However, this appears to be a common misconception as bats are a prime example to

illustrate how structures and anthropized areas can act as optimal shelter and draw in more prey,
 Manlongat 10

thus attracting bats (Afelt et al., 2018). By attracting more bats to areas in close proximity to

human dwellings, the risk of transmission

by direct contact, domestic animal infection,

or urine/feces contamination is increased

(Afelt et al., 2018). For example, a research

study compiled information throughout

Southeast Asia, where the greatest rate of

deforestation (30% forest loss) has occurred

for the past forty years (figure 3). In

addition, bat species, such as Rhinolophus and Hipposideros bats, that have been detected to

carry Cov strains have been recorded within their respective countries in Southeast Asia in

Figure 3. Between the years of 1990 to 2010, Southeast Asia has recorded a net loss of 1.6

million ha yr, equally a deforestation rate of 0.6% (Estoque et al., 2019). Along with the varying

levels of deforestation, the number of viruses that have developed within each region are

depicted on the map. Regions range from tree loss, to anthropized areas, to intact forest cover.

The drastic rise in the development of diseases in this region can be attributed to the varying

ways deforestation has altered contact between animals (including their specimens) and humans.

Although the specific actions in which the disease had developed and transmitted to humans are

unpredictable, the varying levels of biodiversity reduction through deforestation and

anthropization proves to be an important area of investigation worthy of an individual study. As

the human population grows and deforestation is continued to make room for this growth,

interactions between humans and bats may increase.

 Manlongat 11

Modeling

The basis of the mathematical model that

connects deforestation and disease transmission

includes a basic concept of ecology; population

growth rates. The first equation on Figure 4,

explains the basic growth rate of birth rate -

death rate. The second equation incorporates the concept of an ideal environmental setting in

which organisms and populations are expected to increase unboundedly, without the restrictions

of poor/absent resources. The third equation involves the concept that resources are often lacking

within environments and the existence of differing food availability, competition for resources,

disease, and predation. Therefore, the graph results in a logistic “S” curve. The third equation

involves K, the carrying capacity which the maximum number of individuals an environment can

support, and rmax, a constant variable that equates to the maximum intrinsic per capita growth

rate. These differential rate equations set the basis of population estimation and growth

throughout multiple areas of ecological study, including this model.

Given the preeminent increase of viral spillovers and biodiversity loss, mathematical

models exist to predict or replicate the impact of the viruses on both the environment and the

human population. Math models may concentrate on certain stages of

disease transmission, including contact and exposure, cellular entry,

viral replication and release, and transmission (Allen et al., 2012).

The SEIR (Susceptible-Exposed-Infected-Removed) Model is the

basic epizootic model used to track the rate of transmission for the
 Manlongat 12

four separate populations: susceptible, infected, and removed individuals (SEIR and SEIRS

model, n.d.). Epizootic refers to a single outbreak applied to animal populations. For this model,

conditions must be maintained such as the assumption that once an individual with the pathogen

has recovered, the individual gains immunity; the units S, E, I, and R are individuals; the unit of

time is days; the total size of the population (N = S + E + I + R) does not change. The math

model includes four ordinary differential

equations (ODEs) (Figure 3). The equations

note the rate of transmission equation for

susceptible (dS/dt), exposed (dE/dt), infected

(dI/dt), and recovered (dR/dt) individuals. For

example, a four dimensional model was

recently developed for modeling Covid-19

transmission between bats and people, called

the Bat-Host-Reservoir-People network

which encompasses transmission between four separate SEIR groups (Figure 6) (Chen et al.,

2020). This model seeks to identify the components of transmission from bats to humans through

other components, such as the disease reservoir (an environment in which a disease manifests

and lives) and hosts. Again, it follows the same structure of including susceptible (dS/dt),

exposed (dE/dt), infected (dI/dt), and recovered (dR/dt) individuals for each compartment, while

including more specificities tied to the context of the situation such as the presence of

asymptomatic individuals and the rate of the disease entering the reservoir (a marketplace). Such
 Manlongat 13

a model will be the basis for understanding the reformed Bats to Peoples model developed

through this paper.

In addition, both these models incorporate the basic reproductive number (R​0​) which is an

indicator of a disease’s transmission potential (SEIR and SEIRS model, n.d.). With the basic

SEIR Model, R​0 equals

​ transmission rate/recovery rate (​β/γ​  )​. This number is affected by the rate

of contact, the probability of infection when contact is made, and the infection’s duration. R​0

gives researchers a lot of information such as the potential size of an outbreak or the proportion

of populations that must be vaccinated in order to eliminate the disease from the population. In

addition, if R​0​>1, it is predicted to continue and spread. If R​0​<1, the outbreak is predicted to end,

although the absolute time frame is not indicated. R​0​has been applied to several diseases

including the Ebola virus and HIV. When applied to multi-dimensional and complex models, an

equation for R​0​can be found with next generational matrices defined as “G”. It is termed NGM

due to the presence of generations within epidemic models that are the secondary infection

waves that flow from each previous infection. The first generation of an epidemic is the entirety

of secondary infections resulting from infectious contact with the index case, who is of

generation zero. The construction of these matrices is defined as

the average number of new cases of an infection caused by an

infected individual that is a part of the susceptible population

(Diekmann et al., 2009). To find R​0​, one must determine the

dominant eigenvalue of the matrix G = F V​−1​(Jones, 2007). F is

defined as the amount of new infections, while Vi symbolizes the

transfer of infections from one compartment to another (Figure 7).

 Manlongat 14

The practice of determining R​0​is to assume it to be equivalent to the largest eigenvalue of

multiplying the F matrix and V​-1​matrix (inverse matrix of V). A 2x2 matrix can be formed to

illustrate the amount of infections within the two infectious groups (E, exposed, and I, infectious)

and the rate at which individuals/organisms may enter the different compartments (Exposed bats

to Infected bats) even within subpopulations of bats and humans. Ultimately, it incorporates both

ideas of transmission within subpopulations and distinct species, while evaluating the probability

of an epidemic outbreak.

Ultimately this model could be applied to the impact of biodiversity reduction on viral

spillovers. For example, methods such as vaccination (lowering susceptible count) and culling

(selectively slaughtering animals thus reducing reservoir host density) have been used to observe

how the number of individuals in each group can be impacted (Sokolow et al., 2019). Ultimately

alterations to these counts are targeted towards changing or reducing the rate of transmission of

diseases. These calculations occur once viral spillovers have begun or occurred. For preventative

and anticipatory purposes, math models observing environmental degradation may find how they

influence each sub-population by affecting the amount of direct contact. In the context of this

experiment, differences in levels of deforestation and anthropization have yet to be explored and

applied. By manipulating these variables of amount of deforestation and anthropization and

observing the effect on R​0​, a math model could indicate how a reduction in biological diversity

affects bat-related zoonotic viral spillovers. Although math models are not always 100% accurate

applicable to each virus, they can be complementary to approaches like vaccination, disinfection,

treatment, and chemical control that seek to reduce the instances of viral spillover or its effects

(Sokolow et al., 2019).

 Manlongat 15

Performance Criteria
1. Model has adapted the fundamental logistic population growth model and the general
SEIR model to rates of deforestation and anthropization, and bat-related diseases.

2. Users should be able to apply numbers for different rates of deforestation and
anthropization within a tropical forest and the bat population amount within a set
perimeter. Everything should be based on adjustable parameters.

3. Users will be able to see differences in predicted population growth before and after
deforestation.

4. Model incorporates behaviors specific to varying bat species.

5. Model incorporates bats’ behavioral responses to anthropization along with deforestation

events.

6. Model connects rates of deforestation and anthropization within an area to disease

transmission within a bat species of specified area through two subsystems.

7. Math model delivers an output of a basic reproduction number to evaluate the risk of
disease transmission between a population of bats and humans.

8. Users will be able to see distributions and proportions of sub-populations (susceptible,

exposed, infected, and removed) within the bat, host, and human populations.
 Manlongat 16

Design Plan
 Manlongat 17

Model Version 1:
 Manlongat 18

Subsystem 1: Population Growth & Subpopulation Alteration

 Manlongat 19

Subsystem 2: SEIR Model

Manlongat 20
 Manlongat 21

Assumptions:
1. Assume bat population was at equilibrium, in which it remained at its carrying capacity
before deforestation events and subsequent deforestation rates.
2. Assume deforestation events impact each subpopulation of susceptible, exposed, infected,
and removed individuals equally (same amount, not proportion). Therefore:
​[1 - 𝚫N/N​0​ = (S​B0​+ 𝚫N/4N​0​) + (E​B0​+ 𝚫N/4N​0​) + (I​B0​+ 𝚫N/4N​0​) + (R​B0​ + 𝚫N/4N​0​)]
3. Assume rate of deforestation is constant within this model.
4. Assume diseases from bats are not transmitted to hosts.
5. Assume no other bat colonies are within the specified radius (in acres).
6. Assume the disease is present only within the specified bat species population.
7. Assume transmission between infected persons and susceptible persons remains the same
regardless of deforestation event.
 Manlongat 22

8. Assume deforestation has the same impact on both the population of trees used for shelter
and food
9. Assume per capita birth rate and per capita death rate are constant from before and after
deforestation events.
10. Assume bats do not migrate between winter and summer roosts.

Simulation of Model Version 1:

Assumed Values

t 235 acres Average max distance traveled for Rhinolophus bats was 550 m. This
(trees) becomes the radius of an enclosed circle as an environment for the
bats. Area of this circle = πr​2​= π(550)​2​= 950331​*​m or 235 acres.

N​B 300 bats Average of a colony of Rhinolophus bats within Southeast Asia​†

R​max 0.0167 Female bats produce one pup per reproductive episode. Assuming
female to male ratio is 1:1 and 1 bat pup per reproducing female would
be an increase of approximately 50 bats a year. Life expectancy of bats
is four years, average death rate 70 bats a year(life span =3 years)
(50-45)/(300) = 0.0167

K 300 Assuming bats were at carrying capacity and equilibrium prior to

deforestation event, therefore N = K = 300 bats.

dF/dt -6% or deforestation rate in Southeast Asia: -6% or -0.06 (3950000 acres lost
-0.06
from 1990 to 2010)​‡

E​B​+ I​B 6.84% Prevalence of cov in bats = ​6.84%

(171/2500) E​B​+ I​B​= 0.0684 or (171/2500)​§

S​B 0.9316 - Before deforestation = 0.9316 or (2329/2500)

0.989 - After deforestation = 0.989

R​P 0 SEIR model assumes that the initial population is full of only
susceptibles, reducing R​P​to 0.

*(Kingston, 2013)
†​
(Wacharapluesadee et al​.​, 2021)
‡​
(Estoque et al., 2019)
§​
(Lin et al., 2017)
¶​
(Byrne et al., n.d.)
 Manlongat 23

R​B 0

β​B 0.5 Population parameter is unknown, so to illustrate the model, a

conservative value was chosen.

β​P 0.5 Population parameter is unknown, so to illustrate the model, a

conservative value was chosen.

S​P 1 With SEIR model, it assumes the population is full of susceptibles with
a primary case of infection.

γ​P 13.4 days γ​P​= ​(95% CI 10.9 to 15.8)​¶

γ​B 13.5 days γ​B =

​ 13.5 days

m​B 0.15 m​
​ B=
​ 0.15 45/300

m​P 0.00714

*(Kingston, 2013)
†​
(Wacharapluesadee et al​.​, 2021)
‡​
(Estoque et al., 2019)
§​
(Lin et al., 2017)
¶​
(Byrne et al., n.d.)
 Manlongat 24

Results:
Figure 1.

Figure 2.

Figure 3.
 Manlongat 25

R​0

Ideal Environment 0.0356

Environment suffering from Deforestation 0.0368

 Manlongat 26

Summary of Changes
Change in Model Reason

Change birth rate (B) and death rate (D) to per
capita birth rate and per capita death rate for
bats.
B = Per capita birth rate: (50/300) = 0.1667
D = Per capita death rate: (45/300) = 0.15
Allows for the birth rate and death rate to
remain the same (a required condition for a
SEIR model) while still accounting for the
change in population size due to deforestation
and anthropization. Therefore, although
populations will fluctuate due to deforestation
and anthropization, the estimation of the birth
rate and death rate will be given as a per
capita birth rate and death rate for ease of
model usage.

Add the assumption that female bats and male
bats are affected proportionally during
deforestation events.
For the simplicity in application for this
model, a proportional effect on female and
male bat populations would allow the birth
rate and death rate to remain constant from
before and after the deforestation and
anthropization event.

Add the assumption of migration for bats does Under the scope of the simulation for
not occur. Rhinolophus bats, who do not migrate
between summer and winter roosts unlike
several bats, another assumption will be
added to convey their behavior.

Include new carrying capacity for bats in the
anthropized and deforested area (redefine K in
terms of K for deforestation and K for
anthropization).
dA = anthropization rate in Southeast Asia
= 0.032 (3.2%)
L​A​ = % of land converted into housing from
anthropized area
= 50% or 0.5
H = average amount of housing/buildings
per acre
=5
Influx of peoples into area and favorable
conditions that could potentially support bat
roosts. Show there could possibly be an
increase or preservation of bats in deforested
than anthropized areas.
3.2% anthropized rate value in Southeast
Asia*
0.271 bats per building​†

* (Estoque et al., 2019)

†​
(Winter et al., 2020)
 Manlongat 27

K​BB​= 0.48 bats per building

K​ABB​= carrying capacity for bats in housing
within an anthropized environment
= (dA)(t)(L​A​)(H)((K​BB​)
= 9.024 ≈ 9
K = carrying capacity for bats in an area of
deforestation and anthropization
= K​BD​+ K​ABB
= (b​ts​+ b​tf​) (dF/2) + (dA)(t)(L​A​)(H)((K​BB​)
dN/dT: r​max​N{[ (b​ts​+ b​tf​) (dF/2) +
(dA)(t)(L​ )(H)((K​BB​)]-N}/[ (b​ts​+ b​tf​) (dF/2) +
A​
(dA)(t)(L​ )(H)((K​BB​)]
A​

Include transmission rate difference between Contact rate change because subpopulation
infected bats and susceptible bats after distribution will have changed
deforestation and anthropization
β​B ​= transmission risk * secondary attack rate
= γ​B​* p
γ​B​remains constant: 13.5
P = x / S​B
x = number of secondary cases (remain
constant = 10.370)
S​B​= number of susceptibles in bat population
Before deforestation & anthropization:
S​B​= 280 bats
After deforestation & anthropization:
S​B​= 252 bats
β​B
β​B ​Before deforestation & anthropization:
0.500
β​B After
​ deforestation & anthropization: 0.556

* (Estoque et al., 2019)

†​
(Winter et al., 2020)
 Manlongat 28

Model Version 2:
Manlongat 29
 Manlongat 30

Adjusted Assumptions
1. Assume bat population was at equilibrium, in which it remained at its carrying capacity,
before deforestation events and subsequent deforestation rates.
2. Assume deforestation events impact each subpopulation of susceptible, exposed, infected,
and removed bats equally (same amount, not proportion). Therefore:
​[1 - 𝚫N/N​0​ = (S​B0​+ 𝚫N/4N​0​) + (E​B0​+ 𝚫N/4N​0​) + (I​B0​+ 𝚫N/4N​0​) + (R​B0​ + 𝚫N/4N​0​)]
3. Assume rate of deforestation is constant within this model.
4. Assume diseases from bats are not transmitted to hosts.
5. Assume no other bat colonies are within the specified radius (in acres).
6. Assume the disease is present only within the specified bat species population.
7. Assume transmission between infected persons and susceptible persons remains the same
regardless of deforestation event.
8. Assume deforestation has the same impact on both the population of trees used for shelter
and food
9. Assume per capita birth rate and per capita death rate are constant from before and after
deforestation events.
10. Assume bats do not migrate between winter and summer roosts.

Simulation of Model Version 2:

Adjusted and Additional Assumed values

S​B 0.9316 (Ideal environment)
0.989 (Environment with deforestation)
0.979 (environment with deforestation and anthropization)

K 300 (Ideal)
201 (environment with deforestation)
210 (environment with deforestation and anthropization)

β​B 0.500
0.556
 Manlongat 31

B 0.1667

D 0.15

dA 0.032

L​A 0.5

H 5

K​BB 0.48

K​ABB 9

Figure 4.
 Manlongat 32

Figure 5. ​Figure 6.

​Figure 7.

R​0​in an...

Ideal Environment 0.0356

Environment suffering from Deforestation 0.0368

Environment suffering from deforestation and 0.0385

anthropization
 Manlongat 33

Evaluation & Conclusion

In the initial model, the change in year by year birth rate and death rate due to

deforestation or natural variation would break a condition for the SEIR model, therefore leaving

a hole between the connection in the two subsystems. Therefore the model substituted new

variables of a per capita birth and per capita death rate which satisfies the condition for the SEIR

model stating that the birth rate and death rates do not fluctuate. The redesign of these variables

more closely meets the requirements of Criteria 1 because it adapts the logistic population

growth model to the conditions of the SEIR model, allowing the two to become more connected

theoretically and mathematically. However, in actuality this could pose a problem with true

estimation if per capita death rate and per capita birth rate was drastically different than the birth

rate and death rate at the time, which could potentially be fixed with additional remodeling.

Because of the lack of data and accounts on the unknown fluctuations in birth rate and death

rates of bats in the wild, this assumption was made in order for ease in the usage of this model.

The additional assumption that female and male bats are affected proportionally by

deforestation events coincides with the aforementioned change with birth rate and death rate. It

was reasoned that a severe decrease in females and imbalance between females and males would

lead to severe decrease in birth rate. In real life, these fluctuations are unaccounted for because of

lack of data on bats, therefore it was assumed that the female and bat population would be

affected proportionally, especially within a density-independent event such as deforestation and

anthropization. This change was added to meet both Criteria 1 and 5, since it acknowledges the
 Manlongat 34

bats' response to anthropization in terms of female and male population, in addition to

reaffirming the choice of changing birth rate and death rate.

Because the SEIR model requires that populations remain relatively constant, seeing as

those “removed” or “died” are accounted for as part of the R subpopulation, the assumption that

migration within the bat population does not occur was added. The assumption of no migration

recognizes that certain bats have a behavior of migrating. However, seeing as how this would

further complicate the use of the model and for the sake ease for the user, it was omitted

especially through the data simulation of a bat species that does not migrate. By adding this

assumption, non-migratory bats have now become a requirement for use of this system seeing as

migration would drastically impact population growth over time, which is a failure in this

redesign when applying to bats that potentially migrate. However, the intent of this alteration

was to meet Criteria 4, by acknowledging the specific behaviors of forest-foraging and

forest-roosting bats and adapting the model more closely for a clearer basic reproduction

number.

The choice to alter carrying capacity, K, was to account for the bat responses to both

varying amounts of deforestation and anthropization. The initial design neglected anthropization,

therefore to rectify the failure in considering the pattern of deforestation and anthropization that

normally occur as a pair, the carrying capacity was altered. It acknowledges that anthropized

areas and houses could support bats, thus carrying capacity would not plummet as severely if

deforestation was solely involved. In this manner, it meets Criteria 1 and 6 by considering how

population would change based on rate of deforestation and anthropization, which will then be
 Manlongat 35

used for the SEIR model. This alteration also allows for a more accurate basic reproduction

number because population growth is more accurate thus satisfying Criteria 7.

The final change was to account for the difference in transmission rate once deforestation

and anthropization occurred. The initial model failed to acknowledge that the transmission rates

from infected bats to susceptible bats would be altered due to the change in amount of

susceptible bats from deforestation and anthropization. The subpopulations would have changed,

as modeled in the pie charts, however, these differences were overlooked when computing the R​0

therefore it was altered to account for this failure. By incorporating the difference in the

subpopulation of susceptible bats, the model satisfies Criterias 1, 7, and 8 by formally connecting

the change in subpopulations from the logistic population growth model and the SEIR model to

create a more accurate R​0​.

With these changes, the model was able to successfully connect the deforestation rate and

anthropization rate to bat population changes and the ultimate basic reproduction number

symbolizing the probability of an epidemic to occur. It was successful in demonstrating the

changes in subpopulations of bats when in varying circumstances related to biodiversity loss. It

was also successful in tying in the change in subpopulations to the transmission of diseases from

bats to humans. Although it was successful in producing R​0​s for three different environments

(ideal, deforestation, deforestation and anthropization), the model will still need to be specified

for each and every situation it is applied. All the R​0​values were less than one indicating that if an

epidemic were to occur, they’re more likely to end although the time frame of the epidemic is not

given. In addition, several assumptions were made and estimated to serve the purpose of
 Manlongat 36

illustrating the usage of this model. However, these estimations may be marked by bias from

several different sources and need to be shaped by more accurate data collection. Given the lack

of data on bat populations and more specifically, bat-related diseases, the model may divert from

the true R​0​that has yet to be estimated or measured for several bat-diseases not modeled in this

study.

Finally, deforestation and anthropization has shown to impact the populations of bats

through environmental pressures. Through this model, the expected carrying capacity was altered

due to varying deforestation and anthropization rates derived from previous studies. The carrying

capacity ultimately illustrates how bat populations are highly contingent on their environments,

which are decreasing at an increasing rate. Therefore, this model highly encourages a closer

examination of biodiversity loss and bat populations, knowing there is a potential for diseases to

become transmitted between bats and humans. In addition, it is also important to collect

additional information on the presence of diseases in bats, seeing as how the model was

dependent on factors such as incubation periods and subpopulation distributions. Overall, the

results of this study highly encourage the collection of additional data surrounding bats,

deforestation, environmental alteration, biodiversity loss, and disease transmission in order to

prevent the heavy social and economic effects from events like epidemics or pandemics from

occurring.
 Manlongat 37

References

Afelt, A., Frutos, R., & Devaux, C. (2018, April 11). ​Bats, coronaviruses, and deforestation:

Toward the emergence of novel infectious diseases?

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5904276/

Allen, L. J. S., Brown, V. L., Jonsson, C. B., Klein, S. L., Laverty, S. M., Magwedere, K., Owen,

J. C., & Van Den Driessche, P. (2011, December 30). ​Mathematical modeling of viral

zoonoses in wildlife.​ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3358807/

Byrne AW, McEvoy D, Collins AB, et alInferred duration of infectious period of SARS-CoV-2:

rapid scoping review and analysis of available evidence for asymptomatic and

symptomatic COVID-19 casesBMJ Open 2020;10:e039856. doi:

10.1136/bmjopen-2020-039856

Chen, T.-M., Rui, J., Wang, Q.-P., Zhao, Z.-Y., Cui, J.-A., & Yin, L. (2020, February 24). ​A

mathematical model for simulating the phase-based transmissibility of a novel

coronavirus​. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7047374/
 Manlongat 38

Diekmann, O., Heesterbeek, A. P., & Roberts, M. G. (2009, November 5). ​The construction of

next-generation matrices for compartmental epidemic models​.

https://royalsocietypublishing.org/doi/10.1098/rsif.2009.0386

Estoque, R.C., Ooba, M., Avitabile, V. Hijoka, Y., DasGupta, R., Togawa, T., Murayama, Y. The

future of Southeast Asia's forests. ​Nat Commun​ ​10,​1829 (2019).

https://doi.org/10.1038/s41467-019-09646-4

Gebreyes, W. A., Dupouy-Camet, J., Newport, M. J., Oliveira, C. J. B., Schlesinger, L. S., Saif,

Y. M., Kariuiki, S., Saif, L. J., Saville, W., Wittum, T., Hoet, A., Quessy, S., Kazwala,

R., Tekola, B., Shryock, T., Bisesi, M., Patchanee, P., Boonmar, S., & King, L. J. (n.d.).

The Global One Health Paradigm: Challenges and opportunities for tackling infectious

diseases at the human, animal, and environment interface in low-resource settings.​

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4230840/#:~:text=Zoonotic%20infectiou

s%20diseases%20have%20been,%2C%20socioeconomic%2C%20and%20climatic%20fa

ctors

Han, H.-J., Wen, H.-L., Zhou, C.-M., Chen, F.-F., Luo, L.-M., Liu, J.-W., & Yu, X.-J. (2015,

July 2). ​Bats as reservoirs of severe emerging infectious diseases​.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7132474/

Johnson, C. K., Hitchens, P. L., Evans, T. S., Goldstein, T., Thomas, K., Clements, A., Joly, D.

O., Wolfe, N. D., Daszak, P., Karesh, W. B., & Mazet, J. K. (2015, October 7). ​Spillover

and pandemic properties of zoonotic viruses with high host plasticity.​

https://www.nature.com/articles/srep14830
 Manlongat 39

Jones, J. H. (2007, May 1). ​Notes on R0.​

https://web.stanford.edu/~jhj1/teachingdocs/Jones-on-R0.pdf

Keesing, F., Belden, L. K., Daszak, P., Dobson, A., Harvell, C. D., Holt, R. D., Hudson, P.,

Jolles, A., Jones, K. E., Mitchell, C. E., Myers, S. S., Bogich, T., & Ostfeld, R. S. (2010,

December 1). ​Impacts of biodiversity on​ ​the emergence and transmission of infectious

diseases​. https://www.nature.com/articles/nature09575

Kingston, T. (2013). ​Bat evolution, ecology, and conversation​ (R. A. Adams & S. C. Pedersen,

Eds.). https://doi.org/10.13140/2.1.2121.3125

Letko, M., Seifert, S. N., Olival, K. J., Plowright, R. K., & Munster, V. J. (2020, June 11).

Bat-borne virus diversity, spillover and emergence​.

https://www.nature.com/articles/s41579-020-0394-z

Lin, X. D., Wang, W., Hao, Z. Y., Wang, Z. X., Guo, W. P., Guan, X. Q., Wang, M. R., Wang,

H. W., Zhou, R. H., Li, M. H., Tang, G. P., Wu, J., Holmes, E. C., & Zhang, Y. Z.

(2017). Extensive diversity of coronaviruses in bats from China. ​Virology​, ​507​, 1–10.

https://doi.org/10.1016/j.virol.2017.03.019

Martins, S. B., Hasler, B., & Rushton, J. (2014, August 22). ​Economic aspects of zoonoses:

Impact of zoonoses on the food industry.​

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7120208/

Ridenhour, B., Kowalik, J. M., & Shay, D. K. (2014, February). ​Unraveling R0: Considerations

for public health applications​.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3935673/#fd1
 Manlongat 40

SEIR and SEIRS model.​ (n.d.).

https://docs.idmod.org/projects/emod-hiv/en/latest/model-seir.html

Sokolow, S. H., Nova, N., Pepin, K. M., Peel, A. J., Pulliam, J. R. C., Manlove, K., Cross, P. C.,

Becker, D. J., Plowright, R. K., McCallum, H., & De Leo, G. A. (2019, September 30).

Ecological interventions to prevent and manage zoonotic pathogen spillover​.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6711299/

Wacharapluesadee, S., Tan, C.W., Maneeorn, P. ​et al.​ Evidence for SARS-CoV-2 related

coronaviruses circulating in bats and pangolins in Southeast Asia. ​Nat Commun​ ​12,​972

(2021). https://doi.org/10.1038/s41467-021-21240-1

Winter, R., Mantilla-Contreras, J. & Schmidt, S. Usage of buildings in the life cycle of two

endangered ​Rhinolophus​ species in the Mediterranean region: implications for roost

protection. ​Eur J Wildl Res​ ​66,​38 (2020). https://doi.org/10.1007/s10344-020-01374-1

Woolhouse M, Scott F, Hudson Z, Howey R, Chase-Topping M. ​Human viruses: discovery and

emergence.​ Phil. Trans. R. Soc. B (2012) 367, 2864–2871