The population growth and control of African elephants in Kruger

Regis University Regis University

ePublications at Regis University ePublications at Regis University

Regis University Student Publications

(comprehensive collection)

Regis University Student Publications

Spring 2012

The population growth and control of African elephants in Kruger The population growth and control of African elephants in Kruger

National Park, South Africa:: Modeling, managing, and ethics National Park, South Africa:: Modeling, managing, and ethics

concerning a threatened species concerning a threatened species

William C. Fulton

Regis University

Follow this and additional works at: https://epublications.regis.edu/theses

Part of the Arts and Humanities Commons

Recommended Citation Recommended Citation

Fulton, William C., "The population growth and control of African elephants in Kruger National Park, South

Africa:: Modeling, managing, and ethics concerning a threatened species" (2012).

Regis University

Student Publications (comprehensive collection)

. 560.

https://epublications.regis.edu/theses/560

This Thesis - Open Access is brought to you for free and open access by the Regis University Student Publications

at ePublications at Regis University. It has been accepted for inclusion in Regis University Student Publications

(comprehensive collection) by an authorized administrator of ePublications at Regis University. For more

information, please contact [email protected].

Regis University

Regis College

Honors Theses

Disclaimer

Use of the materials available in the Regis University Thesis Collection

(“Collection”) is limited and restricted to those users who agree to comply with

the following terms of use. Regis University reserves the right to deny access to

the Collection to any person who violates these terms of use or who seeks to or

does alter, avoid or supersede the functional conditions, restrictions and

limitations of the Collection.

The site may be used only for lawful purposes. The user is solely responsible for

knowing and adhering to any and all applicable laws, rules, and regulations

relating or pertaining to use of the Collection.

All content in this Collection is owned by and subject to the exclusive control of

Regis University and the authors of the materials. It is available only for research

purposes and may not be used in violation of copyright laws or for unlawful

purposes. The materials may not be downloaded in whole or in part without

permission of the copyright holder or as otherwise authorized in the “fair use”

standards of the U.S. copyright laws and regulations.

THE POPULATION GROWTH AND CONTROL OF AFRICAN ELEPHANTS IN

KRUGER NATIONAL PARK, SOUTH AFRICA: MODELING, MANAGEMENT,

AND ETHICS CONCERNING A THREATENED SPECIES

A thesis submitted to

Regis College

The Honors Program

in partial fulfillment of the requirements

for Graduation with Honors

William Fulton

May 2012

iii

Thesis written by

William Fulton

Approved by

Thesis Advisor

Thesis Reader

Accepted by

Director, University Honors Program

The Population Growth and Control of African Elephants in Kruger National Park,

South Africa: Modeling, Management and Ethics Concerning a Threatened Species

TABLE OF CONTENTS

LIST OF FIGURES vi

LIST OF TABLES ix

PREFACE and ACKNOWLEDGEMENTS xii

INTRODUCTION 1

ECOLOGICAL BACKGROUND 3

THE MODELING PROCESS 5

BUILDING THE MODEL 11

MANAGEMENT DECISION OPTIONS 14

SHAPING THE MODEL 18

ETHICAL CONSIDERATIONS 21

CONCLUSION 26

BIBLIOGRAPHY 27

LIST OF FIGURES

Figure 1: Elephant Population vs. Time

Figure 1: Based on exponential growth, this graph (plotting total elephant numbers

against years) projects the effects of density-independent growth on the elephant

population of Kruger National Park, South Africa. Exponential growth predicts nearly

1000% growth over a fifty-year period.

20000

40000

60000

80000

100000

120000

140000

160000

180000

200000

1990 2000 2010 2020 2030 2040 2050 2060 2070

vii

Figure 2: Elephant Population vs. Time

Figure 2: Based on exponential growth and contraceptives, this graph (plotting total

elephant numbers against years) projects the effects of density-independent growth on the

elephant population of Kruger National Park, South Africa. This model predicts a lower

(but still large) growth rate and final population increase of nearly 300%.

10000

20000

30000

40000

50000

60000

1990 2000 2010 2020 2030 2040 2050 2060 2070

viii

Figure 3. Elephant Population vs. Time

Figure 3: Based on exponential growth, contraceptives, predicted effects from rainfall

and small-scale population removal of fifty elephants per annum (culling, translocation,

etc.), this graph (plotting total elephant numbers against years) projects the effects of

density-independent growth on the elephant population of Kruger National Park, South

Africa. This model predicts the lowest long-term positive growth rate of any model run,

both those presented in the thesis and those rejected for inaccuracy. This model suggests

an approximate 130% growth rate over 50 years.

2000

4000

6000

8000

10000

12000

14000

16000

18000

1990 2000 2010 2020 2030 2040 2050 2060 2070

Elephants

Years

LIST OF TABLES

Table 1: This table represents the elephant population in Kruger

National Park, South Africa from 1994 to 2009 as determined by

annual aerial survey. Data was not available for years with blank

spaces.

Table 2: This table represents the projected growth of the elephant population in Kruger

National Park, South Africa if not checked by density-dependence or management

actions over the next fifty years. Shaded cells represent the projection. The formula for

calculation was: =[previous]*exp(0.0474)

1994 7806 2008 15811 2022 30213 2036 58667 2050 113919

1995 8064 2009 16315 2023 31680 2037 61515 2051 119448

1996 8320 2010 17107 2024 33218 2038 64501 2052 125247

1997 8371 2011 17937 2025 34830 2039 67632 2053 131326

1998 8869 2012 18808 2026 36521 2040 70915 2054 137701

1999 9152 2013 19721 2027 38294 2041 74358 2055 144385

2000 9596 2014 20678 2028 40153 2042 77967 2056 151394

2001 10062 2015 21682 2029 42102 2043 81752 2057 158743

2002 10459 2016 22735 2030 44145 2044 85720 2058 166448

2003 10967 2017 23838 2031 46288 2045 89881 2059 174528

2004 11454 2018 24995 2032 48535 2046 94244

2005 12467 2019 26209 2033 50891 2047 98819

2006 12427 2020 27481 2034 53361 2048 103615

2007 13050 2021 28815 2035 55951 2049 108645

YEAR E. POP.

1994 7806

1995 8064

1996 8320

1997 8371

1998 8869

1999 9152

2000

2001

2002 10459

2003

2004 11454

2005 12467

2006 12427

2007 13050

2008 15811

2009 16315

Table 3: This table represents the projected growth of the elephant population in Kruger

National Park, South Africa if checked only by contraception efforts over the next fifty

years. Shaded cells represent the projection. The formula for calculation was :

=[previous]*exp(0.0234)

1994 7806 2009 16315 2024 24737 2039 35297 2054 50364

1995 8064 2010 17107 2025 25330 2040 36143 2055 51572

1996 8320 2011 17937 2026 25937 2041 37010 2056 52809

1997 8371 2012 18613 2027 26559 2042 37897 2057 54076

1998 8869 2013 19060 2028 27196 2043 38806 2058 55373

1999 9152 2014 19517 2029 27849 2044 39737 2059 56701

2000 9596 2015 19985 2030 28517 2045 40690

2001 10062 2016 20464 2031 29200 2046 41666

2002 10459 2017 20955 2032 29901 2047 42665

2003 10967 2018 21458 2033 30618 2048 43689

2004 11454 2019 21972 2034 31352 2049 44736

2005 12467 2020 22499 2035 32104 2050 45809

2006 12427 2021 23039 2036 32874 2051 46908

2007 13050 2022 23591 2037 33663 2052 48033

2008 15811 2023 24157 2038 34470 2053 49185

Table 4: This table represents a typical instance of the projected growth of the elephant

population in Kruger National Park, South Africa if checked by contraception efforts and

stochastic rainfall projections over the next fifty years assuming causality between

correlations of peak rainfall/NDVI and conception rates. Shaded cells represent the

projection. The formula for calculation was :

=[previous]*EXP((RANDBETWEEN(1185,2375)*(10^-5))*(1+IF(I59 < 437,-

0.5,0)+IF(I59 > 637,0.25,0))+ (IF(RANDBETWEEN(1,13)=1,RANDBETWEEN(-9,-

5)*0.01,0))) where the range 0.01185 – 0.02375 represents the projected effectiveness of

contraception, the I59 value represents rainfall (generated randomly about the long-term

mean), and the final value represents the chance of a seasonal weather fluctuation severe

enough to cause drought and increased mortality (between 5 and 9%).

1994 7806 2009 16315 2024 16245 2039 17485 2054 18559

1995 8064 2010 15180 2025 16590 2040 17724 2055 18772

1996 8320 2011 15298 2026 16899 2041 17853 2056 19312

1997 8371 2012 15604 2027 17195 2042 18214 2057 19675

1998 8869 2013 15791 2028 17316 2043 18738 2058 19945

1999 9152 2014 15924 2029 17496 2044 19039 2059 20095

2000 9596 2015 16099 2030 17904 2045 19475

2001 10018 2016 16418 2031 18322 2046 18409

2002 10459 2017 16600 2032 17243 2047 17668

2003 10945 2018 16953 2033 15955 2048 17893

2004 11454 2019 16007 2034 16072 2049 18162

2005 12467 2020 16298 2035 16352 2050 18433

2006 12427 2021 16568 2036 16696 2051 18859

2007 13050 2022 16952 2037 16913 2052 19269

2008 15811 2023 16131 2038 17143 2053 18210

xii

Table 5: This table represents a typical instance of the projected growth of the elephant

population in Kruger National Park, South Africa if checked by contraception efforts and

stochastic rainfall projections over the next fifty years assuming causality between

correlations of peak rainfall/NDVI and conception rates combined with annual culling of

50 animals. Shaded cells represent the projection. The formula for calculation was the

same as in Table 4 above, excepting the inclusion of a -50 in the formula.

1994 7806 2010 15786 2026 14866 2042 16067 2058 17031

1995 8064 2011 15793 2027 15100 2043 15440 2059 17085

1996 8320 2012 15906 2028 15145 2044 15711

1997 8371 2013 16111 2029 15204 2045 15953

1998 8869 2014 14713 2030 15371 2046 16104

1999 9152 2015 13849 2031 15506 2047 16186

2000 9596 2016 14019 2032 14622 2048 16322

2001 10018 2017 14056 2033 14813 2049 16610

2002 10459 2018 14198 2034 14851 2050 16718

2003 10945 2019 14363 2035 14979 2051 16871

2004 11454 2020 14558 2036 15073 2052 17131

2005 12467 2021 14785 2037 15309 2053 17420

2006 12427 2022 14928 2038 15445 2054 17611

2007 13050 2023 15142 2039 15700 2055 16337

2008 15811 2024 15206 2040 15915 2056 16724

2009 16315 2025 15368 2041 15937 2057 16845

xiii

PREFACE AND ACKNOLWEDGEMENTS

Writing this thesis has been a bit like riding a roller-coaster: full of ups, downs,

and crazy turns – also, I’ve never done anything like it until now. What I thought it would

be at first has changed radically over the course of the project, and there was work that I

did that doesn’t really belong in this thesis document at the final curtain. I started this

project looking for a genuine mathematical problem that I could address, and found out

more about elephants than I ever thought I could.

I’d like to thank first Dr. Trenary, not only for being my thesis advisor, but also

for being a great teacher and a real inspiration, and for both expecting my best and

accepting my mistakes as learning tools. Without his classes and help, I never would have

explored math long enough to develop my passion for it.

I also need to thank Dr. Kleier for being my reader, even though before I walked

into her office to pitch this project, we hadn’t ever met each other. Her comments and her

encouragement have been top-notch, and her help has been invaluable.

My list of acknowledgements would be incomplete without Dr. Bowie, the

Honors Director. Over the past four years, Dr. Bowie has encouraged us to search for

magis and meaning, from the first time I met him to the present day. We tell ourselves

stories in order to live, and I have been privileged to have such a wise character in my

narrative. I would not have been able to produce this thesis without his advice and

guidance.

xiv

My family have also been instrumental in my success, and I have to single out my

father for allowing me to work long hours – sometimes through the night – at his office,

and my fiancée for being so tolerant of the time I’ve spent ignoring her to focus on

completing my work in Honors, including this thesis.

Finally, thank you to my fellow Honors students. Every time I have spoken to

high school students visiting Regis and considering Honors, the chief virtue that I

extolled was the community, and I would be remiss not to acknowledge the support I

have received from my classmates and peers. Thank you for being part of my life for the

past four years, and for sharing the burden and pleasure of the Honors program.

INTRODUCTION

Elephants are hard to count. Despite the apparent implausibility of such a

statement, it is true (at least in the wild). Although elephants are the world’s largest land

animals, their size necessitates that they (as a species) are spread out over proportionately

large distances, making accurate counts difficult and cost-intensive – many “counts” in

parks with large (n > 50) populations rely on statistical inferences that may or may not be

accurate, based on data collected from aerial surveys conducted from helicopters or

fixed-wing aircraft. Despite the difficulty of obtaining information, we have a vested

interest in gathering these data about elephant populations because elephants are an

endangered species (as of 2012, elephants are classified as “vulnerable” by the IUCN).

Furthermore, as international awareness of and interest in conservation increases, so does

the widespread sense that people generally and the African societies in direct contact with

elephants specifically must act to not only protect the existing population but ensure the

ability of the population to grow to a non-endangered threshold. This goal is complicated

by the concurrent goal of maintaining biodiversity because of the unique “elephant

problem” (Caughley 1976): There are not enough elephants in the world (in the sense that

most conservationists and biologists believe that to guarantee the future of African

elephants, there is a minimum necessary population threshold), and yet where elephants

exist – indeed, thrive – there are too many of them. That is to say that many elephant

populations in wildlife preserves currently are near or exceed the density at which

elephant drastically change their landscapes through grazing, debarking of trees, and

xvi

other ecological impacts (Kerley, 2008; van Aarde, 2008). This landscape-scale impact

often negatively impacts biodiversity by extirpating or threatening the extirpation (local

extinction) of preferred species of tree or aloe (Kerley, 2008). In order to accomplish

both conservation goals, it becomes useful to understand elephant population variations,

trends, and the factors which affect them. Towards that end, I am developing a

mathematical model to explain and predict population variations and outcomes. Part of

evaluating the management decisions involves not only choosing actions which bring

about acceptable consequences in the ecosystem, but are also in and of themselves

acceptable actions to the concerned parties (for example, increasing the land available to

wildlife preserves by demolishing or preventing the construction of buildings or farms

might be beneficial to the elephant population but not be acceptable to the general

public). Therefore, part of assessing the model and the management decisions and their

outcomes must be to consider not only the numerical impacts but also the social

ramifications for elephants and the ethical issues surrounding management.

In this thesis I will discuss an approach to modeling and several possible

applicable models, as well as discussing one particular model that describes and projects

the population changes and constraints in Kruger National Park, South Africa. This

model will include several different management options, with preference placed on

projected effectiveness of implementation and ethical considerations. The ultimate goal

of the modeling process is to obtain a mathematical representation of the elephant

population which can accurately predict the growth or decline of elephant populations for

the purpose of maintaining biodiversity.

xvii

ECOLOGICAL BACKGROUND

The African elephant (Loxodonta africana) is the largest land animal on the

planet. An herbivore, the African savannah elephant is separate species from its cousins

the Asian elephant and the forest elephant (Loxodonta cyclotis), which is also native to

Africa. Although there is strong genetic evidence that the two species are distinct, for

conservation purposes the IUCN has as recently as 2007 classified Loxodonta africana

and Loxodonta cylcotis as the same species (Carruthers 2008). Elephants are what’s

known as megaherbivores, meaning herbivores which on average weigh more than 1,000

kg. Elephants are not picky eaters; they are both browsers and grazers (Owen-Smith

2006). There are approximately between 500,000 and 700,000 African elephants in the

wild as of the 2007 African Elephant Status Report released by the International Union

for Conservation of Nature (Blanc 2007). These elephants have no natural predators as

adults, although predators like lions will attack juvenile elephants given the opportunity

(Loveridge 2006). Rather than being in danger of extinction or extirpation in most areas,

then, the African elephant population is on the whole increasing (Blanc 2007), and this is

particularly true in southern Africa (Carruthers 2008).

Human intervention is the most important factor to a sustainable elephant

population in the wild. Of the four main issues affecting African elephant conservation as

identified in the African Elephant Status Report by the World Conservation Union, three

xviii

were directly caused by humans, and the fourth indirectly: habitat loss/fragmentation,

human-elephant conflict, poaching or hunting and negative localized environmental

impacts, respectively (Blanc 2007). Although hunting and poaching gained notoriety for

decimating elephant populations in the nineteenth and late twentieth centuries, it is not

currently a major factor in elephant population dynamics because of increased

international regulation of the ivory trade and increased policing in African states (Twine

2008). Habitat loss and fragmentation is instead the most immediate problem, along with

its consequence of negative localized impacts (Kerley 2008; Shrader 2010). Although

there are dozens of established parks in Africa both public and private, there are

increasingly few areas available to be turned into parkland. This means that elephant

populations are often physically separated and unable to expand beyond the boundaries of

their parks, particularly when the conservation areas in question are fenced (van Aarde

2008). From the perspective of genetic diversity, this means that either conservationists

must transfer individuals between parks for breeding to ensure genetic diversity or larger

spaces for parks must be obtained.

Because of the enclosed nature of the wildlife preserves in which elephants reside,

their populations are necessarily bounded by the resources inside the preserves. As adult

elephants have no natural predators and hunting/poaching has been largely eliminated,

the size of any given population is limited primarily by the amount of available food and

water; these resources are also consumed by the other animals in the park. This makes

predicting the growth of elephant populations a complicated process.

xix

THE MODELING PROCESS

Mathematical modeling is the process of describing a real-world situation with a

mathematical equation. Modeling almost always involves a simplification of the

processes at hand in order to produce a model which is workable at the expense of some

realism. A model is generally considered valid if it accurately describes the system in

question. There are several stages to building a successful model. One of my two primary

sources of modeling theory suggests the following eight stages: establishment of

goals/objectives, identification of system features/boundaries, development of the

mathematical/simulation model, sensitivity analysis, verification, validation, stability

analysis and finally application (Williams 2002). As many of these as possible are

included in the model presented later/

The first stage (establishment of goals/objectives) can be relatively

straightforward. There are five possible goals of building a model (all of which are to

some extent mutually exclusive): generality, realism, accuracy, identification of

information deficiency, and management decisions (Williams 2002). General population

models are designed to be broadly applicable across many species/environments. Such

models are evaluated based on their ability to highlight general patterns in population

shifts, and are characterized by model simplicity, a lack of biological detail, and low

precision when representing particular biological systems (Williams 2002). Population

models designed to be highly realistic focus on biological mechanisms and thus

incorporate highly detailed descriptions of biological processes, with precise

mathematical terms describing them. The level of detail limits the generality of the model

and may cause imprecision when estimating model parameters. Furthermore, the

precision of terms in the model may cause erroneous assumptions about the validity of

the model to a layperson (Williams 2002). Model accuracy is particularly important for

predictive models, i.e. models that seek to predict population changes under varying

conditions. Predictability is often obtained by limiting the scope of the model, to the

detriment of realism and generality (Williams 2002). Sometimes, the goal of the modeler

is to explore the adequacy of the available data and identify the lacking information

which must be provided to further understanding of a problem or system. Models

developed for this purpose often are broadly conceptual and sometimes consist of

graphical/logical representations of biological interactions (Williams 2002). These

models attempt to forecast the biological impacts of management decisions, accounting

for both population effects and management costs/benefits. A distinguishing

characteristic is that these models include decision variables which influence population

dynamics (Williams 2002). Of these five goals, the one on which this thesis focuses is the

fifth, management decisions.

The second stage of Williams’ approach to modeling is identifying the system

features and boundaries. Some of the primary concerns when building a model are the

selection of what is to be included and what is to be excluded (Williams 2002).In the case

of this thesis, the system will be Kruger National Park, South Africa. The features and

xxi

boundaries will be representative of the park as far as possible, including rainfall, NDVI,

age structure, and existing population size.

The third stage is development of the mathematical/simulation model. In the case

of this thesis, several potential models will be discussed before one is selected.

Essentially, the development of the mathematical model rests on seven things to include:

accumulators, sources, sinks, flow, flow regulators, exogenous variables and artificial

controls (Williams 2002). Accumulators include elephant population, net primary

production of plants (NDVI), water resources, and other resource/population

aggregations. A source represents an input to the system from without, like precipitation

into a water accumulator, whereas a sink represents an output of the system to the

external, i.e. population loss from death or fire. Flow is the internal, directional

movement of material between accumulators, in which one accumulator is depleted and

one is increased, as in birth, death, migration, and transfer of individuals between parks.

Flow regulators (unsurprisingly) regulate the rate at which flows occur, and may

represent birth rate, death rate, etc. Exogenous variables are factors that influence the

movement of material across system boundaries, influencing but not being influenced by

the dynamics of the system, and include sources and sinks (as well as other nonmaterial

information transfers). Controls represent management decisions (adding/removing

artificial water supplies, for example).

Some existing models will now be presented, and their parameters, advantages, and

disadvantages mentioned.

• Exponential Growth Model

xxii

– P(t) = P

• P

is the initial population

• k is the growth/decay rate

• t represents time

– Advantage:

• Simple, easy to compute

– Disadvantage:

• Grossly inaccurate over long periods of time for most (non-

microbial) populations

• The Logistic Model

– dP/dt = rP(1 – P/K)

• P is population size

• K is the environmental carrying capacity

• r defines the growth rate

– Advantage:

• Incorporates a regulatory constraint imposed by the environment

– Disadvantage:

• Generality inhibits precision

• Cohort Model

– Rather than a particular mathematical expression to describe the

population, this is a conceptual approach useful when the population in

question is divided into distinct categories (Ex: populations on distinct

xxiii

game reserves are modeled individually) The model combines the distinct

population models into a metapopulation model. The cohort approach can

be generalized to age structures, genders, etc.

• Lotka-Volterra Predator-Prey Model

– System of differential equations:

• dx/dt = (a - by)x

• dy/dt = (-m + nx)y

• Where a – by is the intrinsic growth rate of the prey and –m

+ nx is the intrinsic growth rate of the predator population,

a,b,m,n >0 and determined by the data available

– Advantages:

• The system is self-regulating and capable of equilibrium

• The system is easily modified to account for competition instead of

predation and can be manipulated to account for density-

dependence

– Disadvantages:

• The model only accounts for two species without regard to outside

factors

After examining the data and attempting to fit several variations on these models to the

existing population data, the model which best fits the data collected in Kruger since

elephant culling ceased in 1994 is the exponential growth model, suggesting that there are

xxiv

few (if any) natural braking effects on the growth rate of the population in Kruger at this

time. The next section will discuss in more detail the constructed model.

xxv

BUILDING THE MODEL

Some preliminary testing suggests that the following are the essential elements to

modeling the “natural” growth of an elephant population, as well as the effects of the

management decisions that can be applied to the population in order to affect the

population’s size or growth rate: Rainfall/water and food availability, current/recent

population size/density, and age structure (which influences breeding and death rates) are

essential elements, while culling, contraception, translocation and property expansion are

potential management tools. These are described below.

The first element that will be included is rainfall/water availability: elephants are

“water-dependent” (Kerley 2008); adult elephants drink approximately 225 liters (or 60

gallons) per day (Blanc 2007); and there is strong evidence that rainfall influences

conception rates (Gough and Kerley 2006). Published studies suggest that most elephants

drink every 1-2 days (Owen-Smith 2007) or, in drier climates, at intervals of at most 5

days (Viljoen 1988). Therefore, droughts increase elephant mortality significantly,

particularly among juvenile elephants younger than twelve (Dudley et al., 2001). This is

not only because the juveniles die of dehydration, but also because the probability of

predation rises significantly (Loveridge 2006). Although grown elephants have only one

natural predator (Homo sapiens), lions will attack juvenile elephants which are

xxvi

undefended by adults. For a variety of reasons, juvenile elephant mortality attributable to

lion predation rises during extended periods of drought (Loveridge 2006).

Current and recent population size and density are also important: The larger an

elephant population that is unconstrained by management actions or carrying capacity is,

the higher the growth rate – however, there is evidence to suggest that birth rates (and

therefore growth rates) rise when density is artificially lowered, as by culling (van Aarde

2008). There is evidence that the converse is true, and growth rates slow naturally when

densities are high (van Aarde et al 1999) but the population density at which growth rates

slow is dependent on several factors, including resource availability, and the relationship

is not well studied (van Aarde 2008).

Age structure is both important and difficult to determine because of the relatively

long lifespan of the African elephant. Although studies of elephant longevity are

uncertain, there is a general consensus that the average age to which an elephant might

live is approximately 60 (Blanc 2007). However, the probabilities that elephants will

reach this age are low, particularly given that poachers tend to target elephants with the

largest tusks because tusk growth is proportional to age (Sukumar et al. 1988) This means

that poachers tend to target the most experienced elephants in a herd, thereby reducing

the average age by eliminating the oldest elephants. This is significant because McComb

et al (2001) show that families with older matriarchs have greater reproductive success,

which may be attributable to greater experience and more nuanced communication ability

(McComb 2001), and therefore the model should incorporate higher growth rates

corresponding to family groups with matriarchs older than some threshold, approximately

xxvii

40-45. However, studies have determined that elephants 15-25 years old contribute the

most to population growth (van Aarde 2008), and also that manipulating the proportion of

juveniles to adults in a population through culling is the most effective way to stabilize

population growth with culling (Woolley 2008).

xxviii

MANAGEMENT DECISION OPTIONS

The most historically used (and most controversial) management option is culling,

which consists of killing elephants and may be applied for various reasons. Culling to

reduce population (almost universally for the purpose of reducing undesirable effects of

high elephant densities) is only effective in the short term, as reducing density may lead

to optimal population growth rates (Caughley 1983; van Aarde 2008), unless culling is

done selectively by age category (Woolley 2008; Slotow 2008). Furthermore, culling to

reduce population is an unpopular management choice among many of the “stakeholders”

in the continued existence of elephants, animal rights groups being some of the most

vocal and easily recognizable. Culling as an intervention in cases of “rogue” elephants –

elephants which are excessively aggressive towards humans or other endangered species

– is philosophically distinct from culling to control population, and is in common use

when elephants pose immediate threats to people or human livelihoods, as may occur

when elephants escape from fenced-in conservation areas and threaten crops, or when an

elephant which has witnessed poaching or culling becomes aggressive towards humans,

to give two examples (For this reason, current best practice is to cull entire herds at once

(Slotow 2008). Culling to destroy aggressive animals does not have a significant effect on

elephant population dynamics (Slotow 2008).

One of the newest management options is also the least tested during long periods

of time: contraception. As a management tool, contraception is relatively new: the first

elephant contraceptive was developed in 1989 for other species and first tested on

elephants in the wild in 1996 (Bertschinger 2008). Therefore, the long-term physical and

xxix

social effects of applying contraceptives to elephants are unknown; however, in the short

term, contraception is effective. In a study by Mackey et al (2009), contraception of 75%

of the female elephant population led to a reduction in population growth rates of

approximately 64% (Mackey 2009). The most common contraceptive in use is PZP

(porcine zona pellucida vaccine), which is preferred over other methods of contraception

like castration for reasons of cost on a large scale and behavioral changes caused by

gonadectomies (Bertschinger 2008). Some potential effects (as identified by Kerley and

Shrader 2007) include increased risk of physical harm to contracepted females due to a

fourfold increase in the frequency of estrous and consequently increased incidences of

sexual attention from bulls, although as Bertschinger points out, this is a controversial

assertion (Bertschinger 2008), as well as potentially increased “male-male aggression

over mating opportunities” (Kerley 2007), fundamental changes in herd dynamics due to

a decreased ratio of adult females to calves, and the potential negative impact on the

practice of “allomothering”, the process by which young female elephants serve a kind of

“motherhood apprenticeship” (Lötter 2008). As every article on contraception notes,

these potential long-term effects may or may not occur, and further research is needed.

Translocation, another management option, is the removal of elephants from one

place to another. This diminishes local elephant densities on a similar scale to culling, but

does not involve the killing of elephants. It may also be undertaken in order to either

introduce elephants to a game reserve – often done because the presence of elephants in a

game preserve increases eco-tourism (Grobler 2008) – or to introduce genetic diversity to

a population (Grobler 2008). Translocation therefore has the benefit that the overall

xxx

population is not diminished. However, translocation of elephants to reduce local density

effects has the same effect as culling in that it creates optimal reproductive conditions,

tending to raise the birthrate and therefore nature compensates for the affected densities.

Furthermore, as a method to curb negative elephant density effects, translocation is a

temporary solution at best. The primary factor limiting translocation as a tool for

population management in light of density effects is the absence of available land to

which elephants may be transported (Grobler 2008); the secondary factor is cost:

technological and innovative advances have been made so that it is not a technical

challenge to translocate any number of elephants over any distance (Grobler 2008). For

social reasons, entire family groups are translocated together.

A fourth management option is water provision/deprivation. This management

option was developed both because elephants are highly water-dependent and because

many conservation areas are naturally dry (Chamaillé-Jammes 2007); however, the

effects of providing or removing artificial waterholes on elephant populations or the plant

life in surrounding areas are not yet well understood (Kerley 2008).

Finally, one last management option is property expansion: One of the most

effective and least feasible management tools, property expansion is simply adding area

to existing wildlife preserves. This is an ideal solution insofar as adding area viable for

elephant populations has the potential to reduce density effects on the local scale while

allowing for a larger total elephant population. This is an impossible solution insofar as

the land which might be annexed is virtually all in use for agricultural, industrial, or

otherwise cultural pursuits. Autocratically displacing the people whose livelihoods are

xxxi

tied up in this land is ethically dubious at best, and funds to buy land are often finite and

low to nonexistent. Therefore property expansion is ideal from a management perspective

and impractical from an economic perspective. This is why private game reserves are

useful for the conservation of elephants.

xxxii

SHAPING THE MODEL

We start with a simple expression of exponential growth, where the rate of growth

is fitted to an existing data set using Microsoft Excel and time is measured discretely in

years. Using this software and the Kruger National Park elephant population data from

the last eighteen years (Table 1), an average rate of 4.74 percent growth per annum was

extrapolated. Using this figure and the assumption of density-independent growth, a first

crude population projection was established for the next fifty years (Figure 1, Table 2),

which suggests that the elephant population will increase tenfold over the timespan of the

model – less than the natural lifespan of an elephant. Although this is a highly unlikely

figure, it is not known at what density elephant populations experience density-dependent

effects (Woolley 2008), and so we let the assumption stand in order to investigate the

consequences of management decisions, which will act as artificial density-dependent

parameters. Our model begins like this:

Next, using data extrapolated from Grobler (2008) and Mackey (2009), we assume

contraception of 80% of the female population will produce a reduction in the growth rate

of 50% (Figure 2, Table 3) which implies the population will still quadruple within fifty

years.

xxxiii

We then refine the model further by taking into account a combined factor of rainfall and

NDVI, which are positively correlated with conception rates (Gough and Kerley, 2006)

by increasing the population growth rate by 125% two years annual rainfall is more than

(approximately) one standard deviation from the annual mean and decreasing the

population growth rate by 75% when annual rainfall is less than one standard deviation

from the annual mean. NDVI was assumed to be proportional to rainfall and did not act

as an independent variable.

R: Rainfall measured in mm/year

Rainfall was determined stochastically based on available data and long-term averages

for Kruger National Park, assuming a normal distribution and periodic drought episodes

resulting in mortality rates of 5-9% every 13-16 years, based on studies of density

dependence and drought mortality (Dudley et al., 2001 and Woolley 2008). The resulting

population growth projection predicts doubling the population inside fifty years (Table 4,

Figure 3). Adding either annual culling of 100 juvenile females or annual translocation of

the same to the model both increases the stability of the model (the range of the projected

population after 50 years is decreased by 18%) and results in an average growth rate of

less than half a percent, resulting in steady but slow growth and an increase in the

population of 130% over the next fifty years (Table 5, Figure 4), which allows for

xxxiv

incremental increases in density and therefore ideal conditions for biological

conservation.

d: Probability of a drought that causes between 5-9% mortality

This is the most effective model within the parameters of the data extrapolated from the

literature because it achieves the biodiversity goals of conservation managers seeking to

balance elephant populations with the changes high elephant densities make to their

environments. However, this model is not ideal. The data necessary to better define some

of the relationships in the model is in some cases unpublished (particularly extensive

historical rainfall records or the precise age structure of the elephant population in Kruger

National Park) or even uncollected (as is the case with the long-term effects of

contraception). Therefore, many of flow regulators in the model are not necessarily

accurate, although they are based on published information available from peer-reviewed

journals and park data when available. Although this uncertainty exists, it should be

noted that the model agrees with the vast majority of literature in emphasizing the effects

of contraception, drought-related mortality and age-based culling (real or simulated by

translocation) as the most effective ways to limit population growth, and that the model

fits the available data.

xxxv

ETHICAL CONSIDERATIONS

The management decisions that are the most effective may not necessarily be the

most ethical ones; this next section will explore the ethical discussions surrounding

various kinds of management, in order from most contentious to least contentious topic:

Culling, contraception, and land acquisition/park creation.

Culling, or killing elephants, is by far the most ethically contentious management

decision. The advocates of culling often approach the issue from an ecosystem-oriented

value perspective, whereas the opponents of culling most often approach the issue from a

perspective that values animal rights. As mentioned above, the model benefits both in

stability and overall growth rate from the inclusion of culling as a management technique,

so it is particularly relevant to bear in mind.

Opponents of culling most often object of grounds of cruelty and animal rights.

As an ethical issue, let us first examine animal rights. Many intellectual approaches to

animal rights have links to the work of Peter Singer, who makes the claim that humans do

not deserve any more or any less than any other member of the natural world, and

therefore have no business claiming ethical superiority over other creatures (Singer,

1985). He justifies his anti-exceptionalist attitude by counting all animals as morally

relevant by virtue of their ability to experience pain or distress and also pleasure (Singer

1985). However, Singer acknowledges that, as humans have more complex and intricate

xxxvi

experiences of these things than do other animals, killing a human can be worse than

killing a snake (Singer 1985) which is essentially a utilitarian approach. This approach

could justify culling if the number of elephants had enough of a negative impact of the

other members of the ecosystem; however, there is not enough information on elephant

impact on surrounding species to determine an appropriate metric with the specificity one

would like (Lötter 2008). After Singer, Tom Regan developed a theory of animal rights in

which was the most prominent theory explicitly valuing individuals over populations

(Lötter 2008).

Regan’s view (or a variation thereof) is the one most commonly held among

opponents of culling today. Although Regan draws on some of Singer, he rejects the

utilitarian conclusions in favor of the opinion that killing individual animals is

unacceptable independent of the outcomes for the other members of the ecosystem

(Lötter 2008), and generally favors a laissez-faire attitude towards human intervention in

the affairs of nature.

Beyond appealing to theories of animal rights, opponents of culling also cite

studies and observations which demonstrate that elephants are, along with dolphins and

primates, some of the most “intelligent” species alive. Physiologically, the volume of an

elephant’s brain is comparable (proportionally) in size and complexity to humans (van

Aarde). Elephants are highly social creatures with well-defined social structures (Gough

and Kerley, 2006). They also exhibit curiosity, playfulness, and apparently grieve for

their dead by exhibiting behavior such as trying to lift recently dead elephants onto their

feet, identifying and examining the carcasses of dead elephants both within and without

xxxvii

family groups, and similar behavior (Douglas-Hamilton et al., 2006). It is this last feature

that critics of culling emphasize the most, along with studies that suggest culling has a

negative effect on the behavior of nearby elephants (Lötter, 2008). Although the

attribution of “intelligence” to elephants based on these behaviors is projection and

inference, there is a strong emotional argument to be made.

In favor of culling, there is the argument that in spite of some apparent social

similarities between humans and elephants, the most significant difference between

elephants and humans is that elephants cannot reason abstractly, and cannot explain or

understand the effects of their behaviors on their surrounding ecosystem (Lötter, 2008).

Furthermore, Regan’s argument is problematic insofar as animals (including elephants)

do not have the same accountability of action (or, in other words, agency) that people do.

As Lötter notes, not only do humans have the greatest agency of all animals, but humans

have also already interfered with nature, and must take responsibility for it, interfering to

conserve the most natural state which can be achieved (Lötter 2008). This is the view

taken by many ecosystem-oriented ethical positions, including the IUCN, WWF and

South African National Parks (SANParks). This position implies a holistic approach to

conservation, in which “all aspects of conservation areas should be protected so as to

allow and enable nature to function, as far as possible, on its own without human

interference or even without benevolent human intervention” (Lötter 2008). Despite the

injunction against interference, this approach tends to favor the use of human intervention

in order to maintain the overall health of the ecosystem.

xxxviii

Another perspective is that of traditional African approaches to elephants, which

can be summarized as respectful, sustainable use (Lötter, 2008). Although there is not a

single, unambiguous interpretation of traditional African use of natural resources

including elephants, there is general consensus that it involves sustainable consumption

(Lötter, 2008). This perspective would support the sustainable killing of elephants for

commercial gain, whether that would take the form of hunting for food, culling and

selling ivory for the benefit of local people, or selling hunting licenses (Schmidtz, 1997;

Lötter 2008). Although this does not directly address the issue of culling from a

management perspective, it does inform various alternatives to culling that serve the

same purpose (i.e. total reduction in numbers).

Contraception, while less contentious than culling, relies on many of the same

interference/non-interference arguments made above. Particular to this issue are the

unknowns related to contraception’s long-term effects on individual elephants and herd

dynamics. Proponents view contraception as an effective and non-lethal management tool

with few downsides, whereas opponents tend object to interference generally, cost, or

potential future effects (Bertschinger, 2008). In particular, many researchers have

suggested that large-scale contraception necessary to reduce growth rates would also

reduce the number of newborn calves in a herd to the point where the formative practice

of allomothering would be severely affected (Bertschinger 2008). As a “motherhood

apprenticeship”, allomothering gives young female elephants the opportunity to learn

how to raise calves in the social setting of the herd (Kerley, 2007; Bertschinger, 2008).

Opponents’ objections to high contraception rates primarily center around the possibility

xxxix

that reduced numbers of calves would lead to a decline in the opportunities for

allomothering or an increase in the size of family groups in order to maintain

allomothering, both of which run counter to the philosophy of non-interference (Lötter,

2008) suggested by animal-rights activists.

On the surface, land acquisition/park creation is an ideal solution to the problem

of high elephant densities; however, ethical issues arise regarding the treatment of

citizens who may be either forced off of the land they live on or forced to change their

way of life in order to adapt to park creation. Elsie Cloete notes that the traditional way to

establish a conservation area is to either evict entirely or conditionally accommodate

humans previously occupying the new conservation space (Cloete, 2008). In the context

of Cloete’s article, conditionally accommodate means to essentially prevent subsistence

farmers from continuing with their way of life by preventing farmers from using more

deterrents than loud noises to drive away elephants, who can eat or trample an entire

year’s worth of crops within a single day. Therefore increasing the land available to

conservation areas displaces indigenous populations either physically or occupationally

(Cloete 2008).

CONCLUSION

There is a clear need for conservation programs to manage not only elephants but

their ecosystems as well. This management should be guided by science and ethical

discussion working in tandem. In this thesis, I presented a model that suggests a

managerial course of action informed by projected population growth linked to

population density. This model suggests widespread contraception and small-scale

culling efforts. When ethics are included, this prescription is modified somewhat to

suggest that of the killing of elephants that should occur, it should be done in such a way

that the community benefits (i.e. the elephants should be processed for meat and tusks,

and the proceeds used for local charities or other community-based and locally-

designated sources) and also so that the elephants experience a minimum of suffering.

xli

BIBLIOGRAPHY

Bertschinger, H., & Delsink, A. (2008). Reproductive control of elephants. In Scholes &

Mennell (Eds.), Elephant Management: A Scientific Assessment for South Africa

(84-145). Johannesburg, South Africa: Wits University Press.

Blanc, J.J., Barnes, R.F.W., Craig, G. C., Dublin, H.T., Thouless, C.R., Douglas-

Hamilton, I. and Hart, J.A. (2007). African Elephant Status Report 2007: an

update from the African Elephant Database. Occasional Paper Series of the IUCN

Species Survival Commission, No. 33. IUCN/SSC African Elephant Specialist

Group. IUCN, Gland, Switzerland. vi + 276 pp.

Carruthers, J., & Boshoff, A. (2008). The elephant in South Africa: history and

distribution. In Scholes & Mennell (Eds.), Elephant Management: A Scientific

Assessment for South Africa (84-145). Johannesburg, South Africa: Wits

University Press.

Caughley, G. 1976. The elephant problem – an alternative hypothesis. East African

Wildlife Journal 14, 265–283.

Cloete, E. Africa’s ‘charismatic megafauna’ and Berlin’s ‘two concepts of liberty’:

postcolony routes to utopia? Politikon 35(3), 257-276.

De Beer, Y., Kilian, W., Versfeld, W, van Aarde, R.J. 2006. Elephants and low rainfall

alter woody vegetation in Etosha National Park, Namibia. Journal of Arid

Environments 64, 412-421.

Derrida, J. The animal that therefore I am (more to follow). 2002. Critical Inquiry, 28(2),

369-418.

Douglas-Hamilton, I., Bhalla, S., Wittemeyer, G., & Vollrath, F. 2006. Behavioural

reactions of elephants towards a dying and deceased matriarch. Applied Animal

Behaviour Science 100, 87-102.

xlii

Dudley, J.P., Criag, G.C., Gibson, D., Haynes, G., & Klimowicz, J. 2001. Drought

Mortality of bush elephants in Hwange National Park, Zimbabwe. African

Journal of Ecology 39, 187-194.

Freeman, E.W., Whyte, I., & Brown, J.L. 2008. Reproductive evaluation of elephants

culled in Kruger National Park, South Africa between 1975 and 1995. African

Journal of Ecology 47, 192-201.

Gaugris, J.Y., & van Rooyen, M.W. 2009. Effects of water dependence on the utilization

pattern of woody vegetation by elephants in the Tembe Elephant Park,

Maputaland, Sounth Africa. African Journal of Ecology 48, 126-134.

Gough, K.F., & Kerley, G.I.H. 2006. Demography and population dynamics in the

elephants Loxodonta africana of Addo Elephant National Park, South Africa: is

there evidence of density dependent regulation? Oryx 40(40), 434-441.

Grobler, D.G., van Altena, J.J., Malan, J.H., & Mackey, R.L. (2008). Elephant

translocation. In Scholes & Mennell (Eds.), Elephant Management: A Scientific

Assessment for South Africa (84-145). Johannesburg, South Africa: Wits

University Press.

Ihwagi, F.W., Vollrath, F., Chira, R.M., Douglas-Hamilton, I., & Kironchi, G. 2009. The

impact of elephants, Loxodonta africana, on woody vegetation through selective

debarking in Samburu and Buffalo Springs National Reserves, Kenya. African

Journal of Ecology 48, 87-95.

Kerley, G.I.H., & Shrader, A.M. 2007. Elephant contraception: silver bullet or potentially

bitter pill? South African Journal of Science 103, 181-182.

Kerley, G., Landman, M., Kruger, L., & Owen-Smith, N. (2008). Effects of elephants on

ecosystems and biodiversity. In Scholes & Mennell (Eds.), Elephant

Management: A Scientific Assessment for South Africa (84-145). Johannesburg,

South Africa: Wits University Press.

Lötter, H.P.P., Henley, M., Fakir, S., & Pickover, M. (2008). Ethical considerations in

elephant management. In Scholes & Mennell (Eds.), Elephant Management: A

Scientific Assessment for South Africa (84-145). Johannesburg, South Africa:

Wits University Press.

Loveridge, A.J., Hunt, J.E., Murindagamo, F., & Macdonald, D.W. 2006. Influence of

drought on predation of elephant calves by lions in an African wooded savannah.

Journal of Zoology 270, 523-530.

xliii

Mackey, R.L., Page, B.R., Duffy, K.J., & Slotow, R. 2006. Modelling elephant

population growth in small, fenced, South African reserves. South African

Journal of Wildlife Research 36(1), 33-43.

Mackey, R.L., Page, B.R., Grobler, D., & Slotow, R. 2009. Modelling the effectiveness

of contraception foe controlling introduced populations of elephant in South

Africa. African Journal of Ecology. 47, 747-755.

Owen-Smith, N, Kerley, G.I.H., Page, B., Slotow, R., & van Aarde, R.J. 2006. A

scientific perspective on the management of elephant in the Kruger National Park

and elsewhere. South African Journal of Science 102, 389-394.

Shrader, A.M., Pimm, S.L., & van Aarde, R.J. 2010. Elephant survival, rainfall, and the

confounding effects of water provision and fences. Biodiversity Conservation 19,

2235-2245.

Slotow, R., Whyte, I., & Hofmeyr, M. (2008). Lethal management of elephants. In

Scholes & Mennell (Eds.), Elephant Management: A Scientific Assessment for

South Africa (84-145). Johannesburg, South Africa: Wits University Press.

Trimble, M.J, Ferreira, S.M., & van Aarde, R.J. 2009. Drivers of megaherbivore

demographic fluctuations: inference from elephants. Journal of Zoology 279, 18-

26.

Twine, W. & Magome, H. (2008). Interactions between elephants and people. In Scholes

& Mennell (Eds.), Elephant Management: A Scientific Assessment for South

Africa (84-145). Johannesburg, South Africa: Wits University Press.

Van Aarde, R., Ferreira, S., Jackson, T., & Page, B. (2008). Elephant population biology

and ecology. In Scholes & Mennell (Eds.), Elephant Management: A Scientific

Assessment for South Africa (84-145). Johannesburg, South Africa: Wits

University Press.

Wittemeyer, G., & Getz, W.M. 2007. Heirarchical dominance structure and social

organization in African elephants, Loxodonta africana. Animal Behaviour 73,

671-681.

Wittemeyer, G., Douglas-Hamilton, I, & Getz, W.M. 2005. The socioecology of

elephants: analysis of the processes creating multitiered social structures. Animal

Behaviour 69, 1357-1371.

xliv

Wittemeyer, G., Rasmussen, H.B., & Douglas-Hamilton, I. 2007. Breeding phenology in

relation to NDVI variability in free-ranging African elephant. Ecography 30, 42-

50.

Woolley, L., Mackey, R.L, Page, B.R., & Slotow, R. 2008. Modelling the effect of age-

specific mortality on elephant Loxodonta africana populations: can natural

mortality provide regulation? Oryx 42, 49-57.

Wu, L.S., & Botkin, D.B. 1980. Of elephants and men: a discrete stochastic model for

long-lived species with complex life histories. The American Naturalist 116(6),

831-849.

Young, K.D., & van Aarde, R.J. 2011. Science and elephant management decisions in

South Africa. Biological Conservation 144, 876-885.

Young, K.D., Ferreira, S.M., van Aarde, R.J. 2009. The influence of increasing

population size and vegetation productivity on elephant distribution in the Kruger

National Park. Austral Ecology 34, 329-342.