(Elephas Maximus Sumatranus) in Sumatra, Indonesia

University of Massachusetts Amherst ScholarWorks@UMass Amherst

Open Access Dissertations

2-2011

Ecology and Conservation of Sumatran Elephants (Elephas maximus sumatranus) in Sumatra, Indonesia

Arnold Feliciano Sitompul University of Massachusetts Amherst

Follow this and additional works at: https://scholarworks.umass.edu/open_access_dissertations

Recommended Citation Sitompul, Arnold Feliciano, "Ecology and Conservation of Sumatran Elephants (Elephas maximus sumatranus) in Sumatra, Indonesia" (2011). Open Access Dissertations. 355. https://scholarworks.umass.edu/open_access_dissertations/355

This Open Access Dissertation is brought to you for free and open access by ScholarWorks@UMass Amherst. It has been accepted for inclusion in Open Access Dissertations by an authorized administrator of ScholarWorks@UMass Amherst. For more information, please contact [email protected].

ECOLOGY AND CONSERVATION OF SUMATRAN ELEPHANTS (ELEPHAS MAXIMUS SUMATRANUS) IN SUMATRA, INDONESIA

A Dissertation Presented

ARNOLD FELICIANO SITOMPUL

Submitted to the Graduate School of the University of Massachusetts Amherst in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

February 2011

Environmental Conservation

ECOLOGY AND CONSERVATION OF SUMATRAN ELEPHANTS (ELEPHAS MAXIMUS SUMATRANUS) IN SUMATRA, INDONESIA

A Dissertation Presented

ARNOLD F. SITOMPUL

Approved as to style and content by:

______Curtice R. Griffin, Co-Chair

______Todd K. Fuller, Co-Chair

______Charles M. Schweik, Member

______Paul R. Fisette, Department Head Department of Environmental Conservation

DEDICATION

To my parents Laurencius and Tiurma Sitompul

ACKNOWLEDGMENTS

This dissertation would not have been possible without generous support and valuable contribution from several institutions and person during my doctoral program.

This research project was funded U.S. Fish & Wildlife Service through the Asian

Elephant Conservation Fund (grants 98210-7-G210), The International Elephant

Foundation, International Foundation for Science, and The Columbus Zoo. I like to thank to generous support from Indonesian-American Cultural Foundation and The

Wildlife Conservation Society during my study in The University of Massachusetts. I also would like to thank to Department Environmental Conservation, University of

Massachusetts for providing me teaching assistantship during my doctoral study.

I owe my deepest gratitude to my major professors; Dr. Curtice Griffin and Dr.

Todd Fuller for their advice and support during my study, data analyses, and dissertation writing. Their contribution during my entire study period in UMASS is invaluable. I also thank to my committee member, Dr. Charles M. Schweik for the constructive review and

I also like to express thanks to my colleague at the Asian Elephant Specialist Group-

International Union for Conservation of Nature (IUCN), especially to Meenakshi “Mini”

Nagendran, and Heidi Riddle for their support during my field work.

In Indonesia, I would like to thank to the Director of Bengkulu Natural Resource

Conservation Agency (BKSDA Bengkulu) especially to Yohanes Sudarto, and Edy

Setiarto and all of their staff. I also thank to all mahouts in Seblat for their support and invaluable assistance in the field. The field work will never be completed without a valuable support from Aswin Bangun, Mugiharto, Rasidin, Slamet and Simanjuntak. All of them provided me excellent assistance for data collection and collaring wild elephant.

I also like to thank to the Erni Suyanti Musabine and Wisnu Wardana for the medical veterinary assistance during the collaring elephant in Seblat and Nazaruddin for his technical assistance on capturing wild elephant. The capturing elephant process would not been success without assistance from all of these people. I would also like to thank to

Paska Iswanto and Elisabet Purastuti for remote sensing and GIS assistance during the data analysis. To my fellow graduate students, I would like to express my sincere gratitude especially to Kyle and Jennifer McCarthy, Kirk Olson, and Doug Sigourney.

Living in Amherst is become much more fun with their sincere friendship. At the end, I would like to thank to my mother Ny. Tiurma Sitompul br. Lumban Tobing, my brothers;

Alexander and Daniel Sitompul, and my sisters; Maurina and Vera Sitompul, for their continuous support and last, but not the least, special thank to my beloved wife Hanny

Napitupulu and my lovely daughter Felicia Sitompul for their pray, patience and love.

ABSTRACT

ECOLOGY AND CONSERVATION OF SUMATRAN ELEPHANTS (ELEPHAS

MAXIMUS SUMATRANUS) IN SUMATRA, INDONESIA

FEBRUARY 2011

ARNOLD F. SITOMPUL, B.S., UNIVERSITY OF INDONESIA

M.S., UNIVERSITY OF GEORGIA, ATHENS

Ph.D., UNIVERSITY OF MASSACHUSETTS AMHERST

Directed by: Professors Curtice R Griffin and Todd K. Fuller

Sumatran elephant (Elephas maximus sumatranus) continue to decline due to habitat loss, poaching and conflict with humans. Yet, developing effective land conservation strategies for elephants is difficult because there is little information available on their foraging ecology, habitat use, movements and home range behaviors.

Using the lead animal technique, 14 free-ranging, tame elephants at the Seblat Elephant

Conservation Center (ECC) were observed for 4,496 hours to describe their foraging ecology and diet. The majority of their daily activity was feeding (82.2 ± 5.0%), followed by moving (9.5 ± 4.0 %), resting (6.6 ± 2.1%) and drinking (1.7 ± 0.6%), and individual activity budgets varied among individuals for all activities. At least 273 plant species belonging to 69 plant families were eaten by elephants and five plant families of

Moraceae, Arecaceae, Fabaceae, Poaceae, and Euphorbiaceae were most commonly consumed. Elephants browsed more frequently than grazed, especially in the wet season.

Levels of crude protein, calcium, phosphorus and gross energy in plants eaten by elephants in Seblat appeared adequate for meeting the nutritional requirements.

vii

Home range size of an adult female elephant in the SECC during 2007-2008, was

97.4 km2 for the MCP and 95.0 km2 for the 95% fixed kernel. There were no relationships between average monthly elephant home range sizes or movement distances with rainfall. Distances to rivers and ex-logging roads had little effect on elephant movements, but vegetation productivity, as measured by the Enhanced Vegetation Index, did affect elephant movements.

We used resource selection and compositional analysis habitat ranking approaches to describe adult female elephant habitat use in the SECC. The elephant used medium canopy and open canopy forests more than expected; however, during the day closed canopy forests were used more than at night.

Locating and capturing wild elephants in tropical rainforest environments are difficult and high-risk tasks. However, using tame elephants improves the search efficiency of finding wild elephants in dense forests and reduces risks to staff and target elephants. Use of experienced veterinarians and standing sedation techniques also greatly reduce the risks of elephant injury while immobilizing elephants.

viii

PREFACE

Despite intensive conservation intervention over the past decades, the Sumatran elephant population is increasingly restricted to fewer habitat fragments and human- elephant conflict is expanding. Continuing habitat loss exacerbates this human-elephant conflict and complicates development of effective land conservation strategies for elephants. Current conservation strategies for Sumatran elephants focus on securing elephant habitat and mitigating human-elephant conflict. However, there is also a critical need to link isolated elephant populations by facilitating elephant movements across the landscape. Yet, developing effective land conservation strategies for elephants is difficult because there is little information on what areas are priority habitats for elephants. More information is needed on elephant ecology, especially their foraging ecology, habitat use, movement and home range behaviors. Further, such information is essential for development of long-term mitigation strategies to reduce crop raiding by elephants.

In Chapter 1, I review Sumatran elephant taxonomy, population status and current distribution. I also describe the factors causing the population decline in the last two decades and how little we know about the ecology and behavior of the species. The continuing problem of human-elephant conflict is also reviewed.

In Chapter 2, I provide a summary of elephant diurnal foraging ecology and the implications for elephant conservation in Sumatra. I also discuss elephant diurnal activity budgets and their wild diet composition, and assess the relationship of elephant foraging behavior and rainfall. I also described the five most important plant families for elephant diet in the wild. Finally, I discuss the nutritional quality and gross energy value of

important items in the elephant diet in their natural habitat and determine seasonal foraging ecology of the Sumatran elephant.

Data I present in Chapter 3 reports movement and home range behaviors of female elephants in Sumatra. I discuss the effect of rainfall on elephant movements and home range sizes. I also assess how environmental factors such as vegetation productivity, and proximity of rivers and roads influence elephant movements.

In Chapter 4, I describe elephant habitat use in five land cover types. I compare the two resource selection techniques and determine if there is consistency of both techniques in describing elephant habitat use. I also assess time-based (day and night) elephant habitat use.

In Chapter 5, I discuss the advantages of using tame elephants to search for and help immobilize wild elephants in the lowland tropical rainforest habitat of Sumatra for telemetry study. I also provide some suggestions to increase success on deploying GPS telemetry units on wild elephants without injury.

Future direction of elephant conservation in Sumatra based on this study is presented in Chapter 6. The overall implication of this study is also discussed and management recommendations are provided.

The results of this study provide important information related to restoring degraded elephant habitat, which can be used as a guide to developing effective land use planning by the government and other stake-holder groups especially in the northern

Bengkulu Province. Even with this study, however, more research on elephants in

Sumatra is clearly needed. More samples from more representative areas in Sumatra are needed to develop more comprehensive conservation strategies for the species at the

landscape level. I still believe we have an opportunity to save the largest living land mammals in Sumatra from extinction. Hopefully this study is one of the first important steps on the long path to conservation of elephants in Sumatra.

TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS ...... v

ABSTRACT ...... vii

PREFACE …………………………………………………………………………...... ix

LIST OF TABLES ...... xvi

LIST OF FIGURES ...... xviii

CHAPTER

1. REVIEW OF THE TAXONOMYAND POPULATION DISTRIBUTION OF SUMATRAN ELEPHANTS (Elephas maximus sumatranus)...... 1

1.1 Introduction ...... 1 1.2 Sumatran Elephant Taxonomy ...... 2 1.3 Population Status and Distribution...... 3 1.4 Human-Elephant Conflict ...... 4 1.5 Literature Cited ...... 6

2. DIURNAL FORAGING ECOLOGY AND DIET OF SUMATRAN ELEPHANTS AT THE SEBLAT ELEPHANT CONSERVATION CENTER ...... 9

2.1 Abstract ...... 9 2.2 Introduction ...... 10 2.3 Study Area ...... 10 2.4 Methods and Analyses ...... 11 2.5 Results ...... 13 2.6 Discussion ………………………………………………………………...... 14 2.7 Literatre Cited……………………………………………………………..... 33

3. MOVEMENT AND HOME RANGE BEHAVIOR OF A SUMATRAN ELEPHANT ...... 37

xii

3.1 Abstract ...... 37 3.2 Introduction ...... 37 3.3 Study Area ...... 38 3.4 Method and Analysis ...... 39 3.5 Results ...... 43 3.6 Discussion ...... 44 3.7 Management Implications ...... 47 3.8 Literature Cited ...... 57

4. HABITAT USE OF AN ADULT FEMALE SUMATRAN ELEPHANT ...... 63

4.1 Abstract ...... 63 4.2 Introduction ...... 63 4.3 Study Area ...... 64 4.4 Methods and Analyses ...... 65

4.4.1 Telemetry...... 65 4.4.2 Land cover classification...... 66 4.4.3 Habitat use analysis ...... 66

4.4.3.1 Manly's Resource Selection ...... 67 4.4.3.2 Habitat ranking using Compositional Analysis ...... 67

4.5 Results ...... 68

4.5.1 Manly's Resource Selection...... 68 4.5.2 Habitat ranking using Compositional Analysis ...... 68

4.6 Discussions ...... 69

4.7 Literature Cited ...... 76

5. USE OF TAME ELEPHANTS TO DEPLOY GPS TELEMETRY UNITS ON WILD ELEPHANTS IN A SUMATRA RAINFOREST ...... 78

5.1 Use of tame elephants to locate wild elephants ...... 79 5.2 Locating wild elephants for immobilization using tame elephants...... 80 5.3 Anesthetizing wild elephants ...... 81 5.4 Summary ...... 83 5.5 Literature Cited ...... 86

xiii

6. CONSERVATION IMPLICATION ...... 88

6.1 The importance of elephant conservation in Sumatra ...... 88 6.2 Foraging ecology and natural diet of Sumatran elephant ...... 89 6.3 Sumatran elephant movement and home range behavior ...... 90 6.4 Sumatran elephant habitat use ...... 91 6.5 Literature Cited ...... 93

BIBLIOGRAPHY ...... 94

xiv

LIST OF TABLES

Table Page

2.1: Sex, size and age of 14 elephants observed at Seblat Elephant Conservation Center, Sumatra. April 2007 to August 2008………...... 18

3.1: Male and female elephant home range sizes (100% minimum convex polygon) reported from studies in Africa and Asia ...... 49

3.2: Summary logistic regression models of elephant locations with vegetation productivity (EVI), and distances to river (RVR) and roads (n=198). Models are ranked from best to worst based using Akaike‟s Information Criterion (AIC), and associated delta (Δ AIC), Akaike weight (ω). AIC is based on –2 x log likelihood and the number of parameters in the model (K)...... 50

3.3: Model-averaged estimate, unconditional standard errors and confidence interval of effect on elephant movement in Seblat Elephant Conservation Center ...... 51

4.1: Habitats descriptions and proportion of study area (335.6 km2) used by elephants at Seblat, Bengkulu, Sumatra...... 71

4.2: Resource selection indices for habitats used by an adult, female elephant from 25 August 2007 to 14 May 2008, at Seblat, Bengkulu, Sumatra...... 72

4.3: Bonferroni confidence intervals for proportions of habitats used by an adult, female elephant from 25 August 2007 to 14 May 2008, at Seblat, Bengkulu, Sumatra...... 72

4.4: Habitat ranking matrix of five habitats used by an adult, female elephant from 25 August 2007 to 14 May 2008, at Seblat, Bengkulu, Sumatra based upon: A). MCP home ranges vs. total study area (2nd order selection) and B). GPS locations vs. MCP home ranges (3rd order selection). Higher ranking indicates greater use compared to availability. Within the matrix (+) represent the row habitat is preferred than column habitat whereas a (-) represent the opposite. Triple sign represent significant deviation from random at P<0.05...... 73

4.5: Habitat ranking matrix of five habitats used by an adult, female elephant from 25 August 2007 to 14 May 2008, at Seblat, Bengkulu, Sumatra based on A) nocturnal activity (0100 hr) and B) diurnal activity (0900 hr). Higher ranking indicates greater use compared to availability. Within the matrix, (+) represents the row habitat is preferred over the column habitat, whereas a (-) represents the opposite. Triple sign represent significant deviation from random at P<0.05...... 74

xvi

LIST OF FIGURES

Figure Page

2.1. Seblat Elephant Conservation Center in Bengkulu Province of Sumatra ...... 19

2.2. Diurnal (0700-1700 hr) activity budget of 14 elephants monitored, Seblat Bengkulu, Sumatra April 2007 to August 2008 ...... 20

2.3. Percent diurnal feeding activity for 14 elephants in Seblat, Bengkulu, Sumatra, April 2007 to August 2008 (numbered points show outlier observations)...... 21

2.4. Percent diurnal moving activity for 14 elephants in Seblat, Bengkulu, Sumatra, April 2007 to August 2008 (numbered points show outlier observations)...... 22

2.5. Percent diurnal resting activity for 14 elephants in Seblat, Bengkulu, Sumatra, April 2007 to August 2008 (numbered points show outlier observations)...... 23

2.6. Percent diurnal drinking activity for 14 elephants in Seblat, Bengkulu, Sumatra, April 2007 to August 2008 (numbered points show outlier observations)...... 24

2.7. Numbers of plant species by taxonomic family consumed by 14 elephants, Seblat, Bengkulu, Sumatra, April 2007 to August 2008 ...... 25

2.8. Percent crude protein content by taxonomic family in plants consumed by 14 elephants, Seblat, Bengkulu, Sumatra, April 2007 to August 2008 ...... 26

2.9. Percent calcium content by taxonomic family in plants consumed by 14 elephants, Seblat, Bengkulu, Sumatra, April 2007 to August 2008 ...... 27

2.10. Percent phosphorus content by taxonomic family in plants consumed by 14 elephants, Seblat, Bengkulu, Sumatra, April 2007 to August 2008 ...... 28

2.11. Gross energy content by taxonomic family in plants consumed by 14 elephants, Seblat, Bengkulu, Sumatra, April 2007 to August 2008 ...... 29

2.12. Percent browsing and grazing (based on plant types consumed) of 14 elephants, Seblat, Bengkulu, Sumatra, April 2007 to August 2008 ...... 30

xvii

2.13. Percent browsing in relation to rainfall (mm) for 14 elephants, Seblat, Bengkulu, Sumatra, April 2007 to August 2008 ...... 31

2.14. Percent crude protein, calcium, and phosphorus in plants for the five most common families consumed by elephants, Seblat, Bengkulu, Sumatra, April 2007 to August 2008. (B = browse, G = grass, except bamboo, red dot = mean value)...... 32

3.1. Location of study area in Bengkulu Province, Sumatra, and land use within a 10 km wide radius of the Seblat Elephant Conservation Center...... 52

3.2. Home ranges [minimum convex polygon estimate (MCP) and fixed kernel density estimate (FKDE) 95%, 90% and 50% contour] for an adult female elephant, August 2007 to May 2008, Bengkulu Province, Sumatra...... 53

3.3. Relationship between monthly home range (km2) 95% fixed kernel home range for an adult female elephant and total monthly rainfall (mm), September 2007 to April 2008, Bengkulu Province, Sumatra...... 54

3.4. Relationship between monthly home range (km2) 95% fixed kernel home range for an adult female elephant and total monthly rainfall (mm), September 2007 to April 2008, Bengkulu Province, Sumatra...... 55

3.5. Distance of locations for an adult female elephant to the nearest stream or river, August 2007 to May 2008, Bengkulu Province, Sumatra...... 56

3.6. Distance of locations for an adult female elephant to the nearest road, August 2007 to May 2008, Bengkulu Province, Sumatra ...... 56

3.7. Enhanced Vegetation Index values at the locations for an adult female elephant, August 2007 to May 2008, Bengkulu Province, Sumatra ...... 56

4.1. Location of study area in Bengkulu Province, Sumatra, and land use within a 10 km wide radius of the Seblat Elephant Conservation Center...... 75

5.1. Collaring wild elephant using “standing sedation” technique in Seblat Sumatra.…………… ...... 85

xviii CHAPTER 1

REVIEW OF THE TAXONOMY, AND POPULATION DISTRIBUTION OF

SUMATRAN ELEPHANTS (Elephas maximus sumatranus)

1.1 Introduction

The Sumatran elephant (Elephas maximus sumatranus) is one of four Asian elephant subspecies and only occurs on the island of Sumatra (Hartl et al. 1996; Fernando et al. 2000; Fleischer et al. 2001). The species is classified as endangered by IUCN

(IUCN, 2010) and listed on Appendix I of the Convention on International Trade in

Endangered Species of Wild Flora and Fauna (CITES) (UNEP-WCMC 2010). Sumatran elephant populations have declined dramatically from habitat loss and degradation due to human settlement and large-scale plantation development (Sukumar 2003; Leimgruber et al. 2003; Blake and Hedges 2004). Additionally, increasing human-elephant conflicts often result in the capture and removal of elephants by the government or poisoning by local people to mitigate the conflict (Hedges et al. 2005).

Despite intensive conservation intervention over the past decade, the Sumatran elephant population is increasingly restricted to fewer habitat fragments (Soehartono et al.2007) and human-elephant conflict is expanding (Soehartono et al. 2007; Uryu et al.

2008). Continuing habitat loss exacerbates this human-elephant conflict (Uryu et al.

2008) and complicates development of effective land conservation strategies for elephants.

Current conservation strategies for Sumatran elephants focus on securing elephant habitat and mitigating human-elephant conflict. However, there is also a critical need to link isolated elephant populations by facilitating elephant movements across the landscape. Yet, developing effective land conservation strategies for elephants is difficult because there is little information on what areas are priority habitats for elephants. More information is needed on elephant ecology, especially their foraging ecology, habitat use, movement and home range behaviors. Further, such information is essential for development of long-term mitigation strategies to reduce crop raiding by elephants.

1.2 Sumatran Elephant Taxonomy

There are four subspecies of elephants (Elephas maximus) across Asia (Fleischer et al. 2001). The subspecies E.m. indicus occurs in Southeast and Southern Asia mainlands, including Malaysia, Thailand, Cambodia, Vietnam, Laos, Myanmar, China,

Bhutan, India and Bangladesh. E.m.maximus occurs on the island of Sri Lanka,

E.m.sumatranus on the island of Sumatra, Indonesia, and E.m. borneensis on the island of

Borneo (Fleischer et al. 2001; Fernando et al. 2003). The Sumatran elephant is considered the most primitive subspecies with 20 pairs of ribs while the other subspecies have only 19 pairs of ribs (Shoshani and Eisenberg 1982). Morphologically, Sumatran elephants have relatively smaller body size, larger ears and longer tusks compared to the other subspecies (Shoshani and Eisenberg 1982).

The Sumatran elephant is considered monophyletic based upon mitochondrial

DNA (mt DNA) analyses, and defined as an Evolutionary Significant Unit (ESU)

(Flesicher et al. 2001). Despite early reports of elephants introduced to Borneo by the

East Indian Trading Company in 1750, Fernando et al. (2003) considered the Bornean elephant a distinct ESU based upon mt DNA analyses. Consequently, Sumatran and

Bornean elephants are considered high priority for elephant conservation (Fernando et al.

2003; Blake and Hedges 2004).

1.3 Population Status and Distribution

In 1980, the total Sumatran elephant population was estimated at 2,800-4,800 individuals in 44 discrete populations (Blouch and Haryanto 1984; Blouch and Simbolon

1985; Santiapillai and Jackson 1990). For the Lampung province in southern Sumatra,

Hedges et al. (2005) reported that nine of 12 elephant populations recorded in this province in 1980 were extirpated by 2000. Two of these remaining elephant populations occurred in national parks with an estimated 498 elephants (95% CI=[373,666]) in Bukit

Barisan Selatan National Park and 180 elephants (95% CI=[144,225] in Way Kambas

National Park (Hedges et al. 2005). No estimates of elephant abundance were available for the third area, Gunung Rindingan, a protected forest. Uryu, et al. (2008) reported that the elephant population in the Riau Province of central Sumatra declined from 1,342 elephants in 1985 to 210 in 2007; they attributed this decline to loss of forest cover.

Mobbrucker (2009) reported 117 elephants (95% CI= [69,196]) in the Tebo District adjacent to the Bukit Tigapuluh National Park in the Jambi Province and also reported 47 elephants (95% CI = [20,108] in another elephant population on the border of Riau-Jambi provinces.

Island-wide, the current Sumatran elephant population is estimated at 2,400-2,800 wild elephants (excluding elephants in camps) in 25 fragmented populations (Soehartono et al. 2007). Most of these populations occur in lowland areas, and upwards of 85% of their habitat is outside of protected areas. All populations are considered vulnerable to continuing habitat loss due to large-scale habitat conversion by agriculture and human settlement and illegal logging and forest fires (Hedges et al. 2005, Soehartono et al 2007;

Uryu et al.2008). Additionally, continuing habitat loss brings elephant populations closer to human settlements, resulting in human-elephant conflict (Sukumar 1992; Leimgruber et al. 2003; Hedges et al. 2005).

1.4 Human-Elephant Conflict

Over the past three decades, human-elephant conflict (HEC) was a major factor contributing to the decline of Sumatran elephants (Nyhus et al. 2000; Sitompul 2004;

Hedges et al.2005). Conflict occurs when elephants enter human settlements and agricultural areas, causing property damage, crop-raiding and injuring/killing people

(Nyhus et al.2000; Sitompul et al. 2004). Since the early 1980s, the response of the

Indonesian Government was to capture “problem elephants” and relocate them into

Elephant Training Centers (ETC) (Santiapillai and Jackson 1990, Lair 1997). By 1996,

570 elephants had been captured and relocated to six ETC‟s across Sumatra (Lair 1997).

In a review of this mitigation strategy, Hedges et al. (2005) reported detrimental effects and recommended termination of capturing and relocating elephants into ETCs. In Riau

Province, Uryu et al. (2008) reported that many of the 224 elephants captured and translocated to ETCs from 2000-2007, died at the capture site or after translocation.

Illegal killing of elephants by villagers, typically by poisoning, as retaliation to HEC is also considered as a serious problem and contributed to the extirpation of some elephant populations (Hedges et al. 2005; Uryu et al. 2008).

In 2008, the Indonesian Government developed new regulations for mitigating

HEC (MOF 2008). As a first response, the new procedures recommend driving elephants from agricultural areas using traditional methods (i.e. fireworks, banging drums, shouting) and guarding fields to detect and deter elephants before they enter agricultural areas. The regulation requires these methods first before capturing or translocating

“problem elephants” (MOF 2008). Despite this new regulation, HEC will likely continue and increase as habitat fragmentation and conversion continue.

An understanding of elephant foraging ecology and diet is needed for developing effective strategies for reducing human-elephant conflict. Furthermore, information on elephant movement, home range behavior and habitat use will also help guide habitat management for elephants on Sumatra.

1.5 Literature Cited

Blake, S and Hedges, S. 2004. Sinking the flagship: The case of forest elephant in Asia and Africa. Conservation Biology, 18: 1191-1202.

Blouch, R.A., and Haryanto. 1984. Elephant in Southern Sumatra. Unpublished report, IUCN/WWF Project 3033, Bogor, Indonesia.

Blouch, R.A., Simbolon, K. 1985. Elephants in Northern Sumatra. Unpublished report, IUCN/WWF Project 3033, Bogor, Indonesia.

Fernando, P., Pfrender, M.E., Enclada, S.E., and Lande, R. 2000. Mithocondrial DNA variation, phylogeography and population structure of the Asian elephant. Heredity, 84:362-372.

Fernando, P., Vidya, T.N.C, Payne, J.,Stuewe, M.,Davison, G., Alfred, R.J., Andau, P., Bosi, E., Kilbourn, A., and Melnick, D.J. 2003. DNA analysis indicate that Asian elephants are native to Borneo and are therefore a high priority for conservation. Plos Biology, 1:001-006.

Fleischer, R.C., E.A. Perry, K. Muralidharan, E.E. Stevens, and C.M. Wemmer. 2001 Phylogeography of the Asian elephant (Elephas maximus) based on mitochondrial DNA. Evolution, 55:1882-1892.

Hartl, G.B., Kurt,F., Tiederman, R., Gmeiner, C., Nadlinger, K., Mar, K.u, and Rubel, A. 1996. Population genetics and systematic of Asian elephant (Elephas maximus): A study based on sequence variation at the Cyt b gene of PCR-amplified mitochondrial DNA from hair bulbs. Zeitschrift für Saugetoerkunde, 61:285-295.

Hedges, S., M. J.Tyson, A.F. Sitompul, M. F. Kinnaird, D. Gunaryadi, & B. Aslan. 2005. Distribution, status and conservation needs of Asian elephant (Elephas maximus) in Lampung Province, Sumatra, Indonesia. Biological Conservation, 124:35-48.

IUCN-2010. IUCN Red List of Threatened Species. Version 2010.1 http//www.iucnredlist.org (Downloaded 31 May 2010).

Lair, R. 1997. Gone Astray: The care and management of the Asian elephant in domesticity, FAO. Rome, Italy.

Leimgruber, P., Gagnon, J.B., Wemmer, C.M., Kelly, D.S., Songer, M.A., Sellig, E.R. 2003. Fragmentation of Asia‟s remaining wild lands: implications for Asian elephant conservation. Animal Conservation, 6:347–359.

Ministry of Forestry-MOF. 2008. Standard protocol for human wildlife conflict mitigation (Peraturan Menteri Kehutanan No. 48 Tahun 2008). Jakarta. Indonesia.

Mobbrucker, A. M. 2009. Zum Status des Sumatra Elefanten (Elephas maximus sumatranus) im Landschaftsraum Bukit Tigapuluh, Sumatra, Indonesien: Verbreitung, Abundanz, Altersstruktur und Gefahrdung. Diploma thesis. Freigburg.

Nyhus, P.J., R. Tilson, and Sumianto. 2000. Crop raiding elephants and conservation implications at Way Kambas National Park, Sumatra, Indonesia, Oryx, 34:262- 274.

Santiapillai, C., and Jackson, P. 1990. The Asian elephant: An action plan for its conservation. IUCN/SSC Asian elephant specialist group. Gland. Switzerland.

Shoshani, J., and Eisenberg, J.F. 1982. Elephas maximus. Mammalian species, 182:1-8.

Sitompul, A.F. 2004. Conservation implication of human-elephant interactions in two nationals in Sumatra. Master of Science, Thesis, University of Georgia, Athens, GA.USA.

Sitompul, A.F., Carroll, J.P., Peterson, J.P and Hedges, S. 2008. Modelling impacts of poaching on Sumatran elephant population in Way Kambas National Park, Sumatra, Indonesia. Gajah, 28:31-40.

Soehartono, T., Susilo, H.D., Sitompul, A.F., Gunaryadi, D., Purastuti, E.M., Azmi, W., Fadhli, N., and Stremme, C. 2007. The strategic and action plan for Sumatran and Kalimantan elephant. Departemen Kehutanan, Jakarta. Indonesia.

Sukumar, R. 1992. The Asian elephant: An ecology and management, second ed. Cambridge University Press, Cambridge, UK.

UNEP-WCMC, 2010. UNEP-WCMC Species Database: CITES-Listed Species. http//www.cites.org/eng/resources/species.html (Downloaded on 31 May 2010).

Uryu, Y., Mott, C., Foead, N., Yulianto, K., Budiman, A., Setiabudi, Takakai, F., Nursamsu, Sunarto, Purastuti, E., Fadhli, N., Hutajulu, C., Jaenicke, J., Hatano, R., Siegert, F., and Stüwe, M. 2008. Deforestation, forest degradation, biodiversity loss and CO2 emssions in riau Sumatra, Indonesia, WWF Indonesia Technical Report, Jakarta, Indonesia.

CHAPTER 2

DIURNAL FORAGING ECOLOGY AND DIET OF SUMATRAN ELEPHANTS

AT THE SEBLAT ELEPHANT CONSERVATION CENTER

2.1 Abstract

Since the early 1980s, hundreds of “problem elephants” were captured and translocated to Elephant Training Centers (ETCs) across Sumatra. Yet, there is little information on the suitability of ETCs for supporting elephants and their nutritional requirements. Using the lead animal technique, 14 free-ranging, tame elephants at the

Seblat Elephant Conservation Center (ECC) were observed for 4,496 hours to describe their foraging ecology and diet. The majority of their daily activity was feeding (82.2 ±

5.0%), followed by moving (9.5 ± 4.0 %), resting (6.6 ± 2.1%) and drinking (1.7 ± 0.6%), and individual activity budgets varied among individuals for all activities. At least 273 plant species belonging to 69 plant families were eaten by elephants, but plants from five plant families (Moraceae, Arecaceae, Fabaceae, Poaceae,and Euphorbiacea) were most commonly consumed. Elephants browsed more frequently than grazed, especially in the wet season. Levels of crude protein, calcium, phosphorus and gross energy in plants eaten by elephants in Seblat appeared adequate for meeting the nutritional requirements of elephants. Thus, the secondary forests of Seblat ECC and surrounding forests appear adequate for meeting the dietary requirements of Sumatran elephants; yet the wild elephants of Seblat are largely isolated from other elephant populations and there is little opportunity for dispersal.

2.2 Introduction

Over the past three decades, human-elephant conflict and habitat loss were the major factors causing the decline of Sumatran elephants (Nyhus et al. 2000; Sitompul

2004; Hedges et al.2005). Conflict occurs when elephants enter human settlements and agricultural areas, causing property damage, crop-raiding and injuring/killing people

(Nyhus et al.2000; Sitompul et al. 2004). Since the early 1980s, the response of the

Indonesian Government was to capture “problem elephants” and translocate them into

Elephant Training Centers (ETCs) (Santiapillai and Jackson 1990, Lair 1997). By 1996,

570 elephants had been captured and moved to six ETC‟s across Sumatra (Lair 1997).

Despite this intensive conservation intervention, there is little information on the suitability of ETCs for supporting translocated elephants and meeting the nutritional requirements of Sumatran elephants. Thus the purposes of this study were to describe the foraging ecology and diet composition of elephant in the Seblat Elephant Conservation

Center on Sumatra, and assess the nutritional quality and gross energy value of important food used by elephants in the lowland forest of Sumatra.

2.3 Study Area

The study was conducted in the Bengkulu Province on the west coast of Sumatra and included the Seblat Elephant Conservation Center (SECC) (lat 03° 03‟12” –

03°09‟24” S, long 101° 39‟18” – 101° 44‟50” E) and surrounding forested and developed areas (Fig. 2.1). Annual rainfall typically exceeds 3000 mm and elevations are < 50 m above sea level. The perennial Seblat River forms the northern boundary of the SECC, providing a reliable water supply for elephants. The SECC comprise 6865 ha of which

70% was in forest cover in 2007. These forests are regenerating following selective logging operations in the late 1980s. Using the land cover map developed by Laumonier et al. (2010), forests comprised 23% of the land cover within a 10 km radius of the SECC with the remainder classified as non-forested. Extensive palm oil plantations, small-scale agricultural areas and human settlements comprise the majority of non-forested lands surrounding the SECC. In addition to 23 elephants captured as part of the government‟s human-elephant conflict mitigation program and housed at the SECC, 40-60 wild elephant are believed to occur on the SECC. With extensive agriculture and human settlements surrounding much of the SECC, there is much human-elephant conflict in the area.

2.4 Methods and Analyses

Fourteen tame elephants (two males and 12 females, Table 2.1) at the SECC were used for the study between April 2007 and August 2008. Although attended by a mahout throughout each observation period, the tame elephants were permitted to forage freely, consuming a natural diet. Using the lead animal technique (Litvaitis 2000) individual elephants or sometimes 2-3 elephants, were observed between 0700 to 1700 hrs. The activity (feeding, moving, resting, drinking) of each elephant was recorded at 5-min intervals (Altman 1974).

Feeding activity was considered all behaviors directly involved with gathering, manipulating, chewing and swallowing food items. Moving was recorded only when elephants traveled from one place to another, but excluded movements while feeding.

Resting was recorded when elephants were standing or laying down and there was no

feeding activity. Drinking was recorded when elephants drank water from streams or ponds.

The daily activities of individual elephants were averaged providing an activity budget for each elephant observed, and pooled by sex. An one-way ANOVA (Sokal and

Rohlf, 1995) was used to test for differences in activity budgets between individual elephants. I used post hoc TUKEY- HSD statistical tests to identify differences between time-activity budgets of individual elephants.

Samples of all food plants eaten by elephants during the study were collected and identified to species/family based upon comparisons with specimens in the herbarium collection in Bogor (Indonesia Institute of Sciences). The most common plants consumed by elephants were analyzed for crude protein (CP), macronutrients (Ca, P) and gross energy produced using the standard Kjeldhal method (Goering & Van Soest 1970). Only leaf and twig materials were sampled for woody plants and the entire plant sampled for grasses.

Proportion of time spent browsing and grazing was based upon the plants utilized.

Grazing occurred when elephants consumed grass and small herbaceous plants on the ground. Browsing occurred when elephants consumed foliage from shrubs, young trees, tree bark, and bamboo. I used a Pearson‟s statistical regressions analysis (Sokal and

Rohlf, 1995) to examine relationships between season and grazing/browsing feeding behaviors. All statistical tests were conducted using SPSS statistical software ver 17.0

(SPSS. Inc)

2.5 Results

A total 4,496 hours of observation were made on the daily activities of 14 elephants. Most of their daily activity was feeding (82.2 ± 5.0%), followed by moving

(9.5 ± 4.0 %), resting (6.6 ± 2.1%) and drinking (1.7 ± 0.6%). Individual activity budgets varied among individuals for all activities (Fig. 2.2; feeding [F=23.55, df=13, P<0.001], moving [F=18.62, df=13, P<0.001], resting [F=21.38, df=13, P<0.001] and drinking

[F=8.23, df=13, P<0.001]). Post hoc Tukey-HSD analyses indicated that the activity budget of each individual elephant differed from at least one other elephant (Fig. 2.3; 2.4;

2.5 and 2.6). Male elephants tended to spend more time feeding (F=48.80, df=1,

P<0.001) and drinking (F=4.93, df=1, P<0.05), but less time moving (F=77.13, df=1,

P<0.001), compared to female elephants.

At least 273 plant species belonging to 69 plant families were eaten by elephants;

(Fig. 2.7). The most common plant taxa consumed were in the Moraceae family

(mulberry family-32 species), followed by Arecaeae (palm family-26 species), Fabaceae

(legume family-25 species), Poaceae (grass family-21 species) and Euphorbiaceae

(spurge family-11 species). Elephants consumed mostly twigs, young leaves and sometimes bark from the Moraceae, Fabaceae, and Euphorbiaceae taxa. Fruit, primarily figs (Moraceae: Ficus sp.),was rarely consumed. The leaves and, petioles of palms including spines, were eaten. Typically, entire grass clumps were consumed. Bamboo species (Schizostachyum sp. and Gigantochloa sp.) were commonly eaten, comprising

19% of the total diet, and 33% of the elephant browse diet.

Elephants tended to browse (56.3%) more than graze (43.1%). Bamboo, shrubs, young trees, rattan and liana were typically browsed, whereas grass species, mainly in the

Poaceae family, dominated the grazed plant taxa (Fig. 2.12). Elephants tended to browse more during the wet months (F=6.35, df=13, P=<0.05) versus the dry months when they tended to graze (F= 6.62, df=13, P<0.05). Further, browsing increased with increasing rainfall (rs = 0.58, df=13, P<0.05; Fig. 2.13).

Nutritional values of the 95 plant species most commonly eaten by elephants averaged 8.8% (SD= 3.0%) for crude protein (CP), 0.70% (SD= 0.41%) for calcium content is 0.70% (SD= 0.41%), 0.21% (SD= 0.07%) for phosphorus content and 2862.9 cal/gram (SD= 249.6 cal/gram) gross energy. The average moisture contents is 74%

(SD= 9.9%). Plants in the family Limnocharitaceae had the highest protein (15.7%) and

Phosphorus content (0.48%) compared to other plant taxa, and plants in the Sapotaceae family had the highest calcium content (1.95%, Fig. 2.8; 2.9; and 2.10). Gross energy was the highest for plants in the Ulmaceae family (3,369.0 kal/gram, Fig. 2.11). There were no differences in protein, calcium and phosphorus content or gross energy in browse

(four most common families of browse plants [Moraceae, Fabaceae, Arecaceae, and

Euphorbiaceae] versus grass diets (Fig. 2.14).

2.6 Discussion

Feeding was the dominant (82%) elephant diurnal activity for Seblat elephants, similar to that reported for African elephant (70%-75% feeding; Lindsay 1994), and

Asian elephants (up to 91.1%; McKay 1973). The higher feeding activity by male elephants probably resulted from their larger size (~10%) and concomitant higher basal metabolic rate (Kleiber 1947). This foraging/size/basal metabolic rate relationship may also partly explain the varying foraging rates we recorded for individual elephants at

SECC for which there was much size variation. In Kenya, Lindsay (1994) also reported that the basal metabolic rate of adult male elephants could reach 1.5 times higher than for adult females.

The five plant taxa families (Moraceae, Arecaceae, Fabaceae, Poaceae,and

Euphorbiaceae) most frequently eaten by Seblat elephants were also reported as important in diets of elephants in Asia and Africa (Buss 1961; MacKay 1973; Guy 1976;

Olivier 1978; Short 1981; Sukumar 1989; White et al. 1993; Chen et al. 2006; Campoz-

Arceiz et al. 2008). Yet, the nutritional values (crude protein, phosphorus, calcium, gross energy and moisture content) of plants from these five taxa were not the highest compared to other plant families sampled. This finding suggests that elephant diet may also be influenced by other factors, such as plant abundance, availability and palatability.

Curiously, the same five plant families most frequently consumed by elephants are also important in the diets of other animals such as orangutans (Pongo pygmaeus), hornbills and many ungulates (Kinnaird et al. 1996, Schaller 1998, Baskin and Danell, 2003,

Russon et al. 2009).

The tendency of elephants to browse more than graze is probably related to the high availability and nutritional value of browse plants in Seblat. Lowland rainforest dominates Seblat and grassland habitat only occurs in small patches within the forest.

Additionally, feeding behaviors of other large herbivores maximize their nutritive value while reducing the ingestion of secondary chemical compounds (Bryant and Kuropat

1980). This condition may occur for Seblat elephants where bamboo, with its low tannin levels (Easa 1989; Shuguang et al. 2009), is important in elephant diets.

In contrast to elephants in India and Africa where elephants tend to switch their foraging strategy from mostly browsing during dry seasons to grazing during the wet seasons (Barnes 1982; Sukumar 1989; Lindsay 1994), Seblat elephants tended to browse more during the wet season. This pattern may be related to higher protein content and fatty acids reported in browse versus grass plant species during the wet season (Dougall, et al. 1964; Field 1971).

The level of crude protein (CP) (range 3.97-15.66%), calcium (range 0.14-

1.95%), and phosphorous (range 0.11-0.48%) elephant diets appeared adequate to support nutritional requirements of Seblat elephants. Olson (2004) reported that adult Asian elephants need at least 8% of CP in their diet with pregnant females up to 14%; calcium concentrations needed for elephant growth 0.5-0.7% with breeding females in early stages of pregnancy about 0.3%; and phosphorus concentrations of 0.3-0.4% with breeding females in early stages of pregnancy 0.2%. Similarly gross energy in the diets of

Seblat elephants appeared to be adequate based on the minimum energy requirement for active metabolic rate (AMR) calculated by Kleiber (1947). Assuming maximum weights of 2,610 kg for adult males and 2,400 kg for adult females, Sumatran elephants need approximately 51,000 kcal and 48,000 kcal per day, respectively. To fulfill these minimum AMR energy requirements, adult male and female Seblat elephants would need to consume 18 kg for adult males and 17 kg for adult females of vegetation (dry matter) or 120 kg and 113 kg (wet matter ~ 85% moisture contents) per day. These food consumption amounts represent less than 1% (dry matter) of body mass for Seblat elephants compared to.1.42% – 1.54% for captive Asian elephants fed on grass hay

(Clauss et al. 2003) and 1.03% - 4.4% for African elephants fed palm leaves (Dierenfeld

2006).

Although selectively logged 20 years ago, the Seblat ECC and the surrounding forested areas provide adequate nutritional quality for supporting elephant reproduction and growth. Thus, secondary forests of similar age should be considered suitable habitat in conservation planning for Sumatran elephants, and habitat management in other disturbed elephant habitats should focus on restoring/providing plants important in elephant diet. Yet, despite the suitable nutritional value of its lowland forests, the Seblat

ECC is relatively small and the forested areas adjacent to the center are unprotected and at risk of conversion to palm plantations and human settlements. The loss of these elephant habitats outside of the SECC will undoubtedly reduce the capacity of the area to sustain its current wild elephant population (~40-60 elephants) and will increase human- elephant conflicts in the area. Further, loss of these unprotected forested areas will also exacerbate the isolation of the SECC wild elephant population from the nearest other elephant populations in the northern part of the Bengkulu Province. Consequently, there is a critical need to protect the former logging concessions around SECC, including the

Production Forest Air Rami, Production Forest Air Teramang, Limited Production Forest

Lebong Kandis, Limited Production Forest Air Ipuh 1 and Air Ipuh 2. Additionally, recent human settlements on the eastern border of the SECC pose a significant potential barrier to elephant dispersal to the east and north from the Seblat ECC.

Table 2.1. Sex, size and age of 14 elephants observed at Seblat Elephant Conservation Center, Sumatra. April 2007 to August 2008.

Elephant name Age/sex Shoulder ht Approx. wt Approx. age (cm) (kg) (yrs)

Fatma AF 205 2140 21

Darmi AF 220 2050 24

Tria AF 225 2320 32

Natalia AF 220 2280 36

Yanti AF 205 1740 24

Sari AF 225 2240 31

Mori AF 215 2400 39

Aswita AF 220 2280 21

Gia AF 220 2010 23

Desi AF 215 2300 24

Paula AF 205 1800 19

Eva AF 215 2080 19

Nelson AM 240 2610 30

Ucok SAM 220 1960 14

A = adult, SA = subadult, F = female, M = male

Figure 2.1. Seblat Elephant Conservation Center, Bengkulu Province, Sumatra

Figure 2.2. Diurnal (0700-1700 hr) activity budget of 14 elephants monitored, Seblat Bengkulu, Sumatra April 2007 to August 2008.

100.00

80.00

60.00

Drinking

40.00 Resting Feeding Moving 20.00

0.00

Figure 2.3. Percent diurnal feeding activity for 14 elephants in Seblat, Bengkulu, Sumatra, April 2007 to August 2008 (numbered points show outlier observations).

Figure 2.4. Percent diurnal moving activity for 14 elephants in Seblat, Bengkulu, Sumatra, April 2007 to August 2008 (numbered points show outlier observations).

Figure 2.5. Percent diurnal resting activity for 14 elephants in Seblat, Bengkulu, Sumatra, April 2007 to August 2008 (numbered points show outlier observations).

Figure 2.6. Percent diurnal drinking activity for 14 elephants in Seblat, Bengkulu, Sumatra, April 2007 to August 2008 (numbered points show outlier observations). .

Figure 2.7. Numbers of plant species by taxonomic family consumed by 14 elephants, Seblat, Bengkulu, Sumatra, April 2007 to August 2008.

Number of Species 30

Cyperaceae

Tectaria Orchidaceae Myrsinaceae Urticaceae Burseraceae Arecaceae

Cyatheceae Gesneriaceae Asteraceae Moraceae

Fagaceae Icacynaceae Lauraceae Fabaceae

Cecropiaceae Cucurbitaceae

Leeaceae Myristicaceae Dipterocarpaceae Lecythidaceae Myrtaceae

Bombacaceae Linaceae Maranthaceae

Hippocrateaceae Araceae Clausiaceae Dioscoreaceae Gnetaceae Tiliaceae

Apocynaceae Rutaceae Zingiberaceae Sterculiaceae

Meliaceae

Polygalaceae Aspleniaceae Liliaceae PterisGr (Blechnaceae) Schisandraceae Ulmaceae Chrysobalanaceae Connaraceae Poaceae

Polypodiaceae Aslepiadaceae Passifloraceae Vitaceae Rubiaceae Mimosaceae Rhamnaceae

Sapotaceae Verbenaceae Piperaceae Convolvulaceae

Flagellariaceae Araliaceae Menispermaceae Annonaceae

Amaranthaceae Solanaceae Violaceae Pandanaceae

Limnocharitaceae Euphorbiaceae

Sapindaceae

Commelinaceae Selaginellaceae

Family

Figure 2.8. Percent crude protein content by taxonomic family in plants consumed by 14 elephants, Seblat, Bengkulu, Sumatra, April 2007 to August 2008.

Limnocharitaceae Ulmaceae Verbenaceae Sapotaceae Euphorbiaceae Annonaceae Fabaceae Menispermaceae Cucurbitaceae Cyatheceae Violaceae Meliaceae Myrsinaceae Burseraceae Poaceae

Plant Family Plant Solanaceae Fagaceae Chrysobalanaceae Araceae Bombacaceae Rhamnaceae Moraceae Hippocrateaceae Convolvulaceae Arecaceae Cyperaceae Maranthaceae Zingeberaceae Flagellariaceae Myristicaceae Tiliaceae

0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00

Protein (%)

Figure 2.9. Percent calcium content by taxonomic family in plants consumed by 14 elephants, Seblat, Bengkulu, Sumatra, April 2007 to August 2008.

Sapotaceae Cucurbitaceae Maranthaceae Araceae Limnocharitaceae Solanaceae Moraceae Violaceae Zingeberaceae Myristicaceae Fabaceae Rhamnaceae Convolvulaceae Chrysobalanaceae Euphorbiaceae Cyperaceae Fagaceae Tiliaceae Arecaceae

Plant Family Plant Verbenaceae Burseraceae Myrsinaceae Bombacaceae Meliaceae Poaceae Annonaceae Menispermaceae Flagellariaceae Ulmaceae Cyatheceae Hippocrateaceae

0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80 2.00 2.20

Calcium (%)

Figure 2.10. Percent phosphorous content by taxonomic family in plants consumed by 14 elephants, Seblat, Bengkulu, Sumatra, April 2007 to August 2008.

Limnocharitaceae Zingeberaceae Verbenaceae Cyatheceae Rhamnaceae Burseraceae Solanaceae Menispermaceae Maranthaceae Sapotaceae Cyperaceae Violaceae Euphorbiaceae Fabaceae Araceae Chrysobalanaceae Annonaceae Plant Family Plant Fagaceae Ulmaceae Myrsinaceae Cucurbitaceae Poaceae Moraceae Flagellariaceae Arecaceae Convolvulaceae Bombacaceae Hippocrateaceae Tiliaceae Myristicaceae Meliaceae

0.00 0.10 0.20 0.30 0.40 0.50 0.60

Phosphorus(%)

Figure 2.11. Gross energy content by taxonomic family in plants consumed by 14 elephants, Seblat, Bengkulu, Sumatra, April 2007 to August 2008.

Ulmaceae Fagaceae Tiliaceae Myrsinaceae Convolvulaceae Hippocrateaceae Bombacaceae Chrysobalanaceae Myristicaceae Rhamnaceae Arecaceae Burseraceae Fabaceae Menispermaceae Zingeberaceae Euphorbiaceae Maranthaceae

Poaceae Plant Family Plant Moraceae Araceae Cyperaceae Violaceae Limnocharitaceae Cucurbitaceae Verbenaceae Solanaceae Meliaceae Flagellariaceae Sapotaceae Annonaceae Cyatheceae

0.0 500.0 1000.0 1500.0 2000.0 2500.0 3000.0 3500.0 4000.0

Gross Energy (Calorie/gram)

Figure 2.12. Percent browsing and grazing (based on plant types consumed) of 14 elephants, Seblat, Bengkulu, Sumatra, April 2007 to August 2008.

100.0

80.0

60.0

Browse 40.0 Graze

20.0

0.0

Figure 2.13. Percent browsing in relation to rainfall (mm) for 14 elephants, Seblat, Bengkulu, Sumatra, April 2007 to August 2008.

Figure 2.14. Percent crude protein, calcium, and phosphorus in plants for the five most common families consumed by elephants, Seblat, Bengkulu, Sumatra, April 2007 to August 2008. (B = browse, G = grass, except bamboo, red dot = mean value).

% 10

0 Protein (B) Protein (G) Calcium (B) Calcium (G) Phosphorus (B) Phosphorus (G)

2.7 Literature Cited

Altman, J. 1974. Observational study of behavior: sampling methods. Behaviour 49: 227- 267.

Barnes, R.F.W. 1982. Elephant feeding behaviour in Ruaha National Park, Tanzania. African Journal of Ecology 20: 123-136.

Baskin, L., and Danell, K. 2003. Ecology of ungulates: a handbook of species in Eastern Europe and Northern and Central Asia. Springler Verlag. Berlin. Germany.

Bryant, J.P., and Kuropat, P.J. 1980. Selection of winter forage by subartic browsing vertebrates: the role of plant chemistry. Ann. Rev. Ecol. System 11: 261-285.

Buss, I.O. 1961. Some observations on food habits and behavior of the African elephant. Journal of Wildlife Management 25: 131-148.

Campoz-Arceiz, A., Lin , Z.T., Htun, W., Takatsuki, S and Leimgruber, P. 2008. Working with mahouts to explore the diet of work elephants in Myanmar (Burma). Ecological Research 23: 1057-1064.

Chen, J., Deng, XB, Zhang L, Bai ZL. 2006. Diet composition and foraging ecology of Asian elephants in Shangyong, Xishuaangbanna, China. Acta Ecologica Sinica 26: 309-316.

Clauss, M., Loehlein, W., Kienzle, E., and Weisner, H. 2003. Studies on feed digestibility in captive Asian elephant (Elephas maximus). J.Anim. Physiol.and Anim.Nutr. 87: 160-173.

Dougall, H.W., and Sheldrick, D.L.W. 1964. The chemical composition of a day‟s diet of an elephant. East Africa Wildlife Journal 2:51-59

Dierenfeld, E.S. 2006. Nutrition. In: Biology, medicine, and surgery of elephants. Fowler, M.E and Mikota, S (eds). Blackwell Publishing.Iowa.USA

Easa, P.S. 1989. Certain aspects of ecology and ethology of the Asian elephant (Elephas maximus Linn.) in Parambikulam Wildlife Sanctuary, South India. PhD Thesis. Univerisity of Kerala. India.

Field, C.R. 1971. Elephant ecology in the Queen Elizabeth National Park, Uganda. East Africa Widlife Journal 9:99-123.

Goering, H.K., Van Soest, P.J. 1970. Forage fibre analysis (Apparatus, reagents, procedures and some applications). USDA Agriculture Handbook, pp 1–21, no. 379.

Guy, P. R. 1976. The feeding behaviour of elephants Loxodonta africana in the Sengwa area, Rhodesia. South African Journal of Wildlife Research 6:55-63.

Hedges, S., Tyson, M.J. Sitompul, A.F., Kinnaird, M.F., Gunaryadi, D. & Aslan, B. 2005. Distribution, status and conservation needs of Asian elephant (Elephas maximus) in Lampung Province, Sumatra, Indonesia. Biological Conservation, 124:35-48.

Kinnaird, M.F., O‟Brien, T.G.O and Suryadi, S. 1996. Population fluctuation in Sulawesi Red-Knobbed Hornbills: Tracking Figs in space and time. The Auk 113: 431-440.

Kleiber,M. 1947. Body size and metabolic rate. Physiological Review 27: 511-541.

Lair, R. 1997. Gone Astray: The care and management of the Asian elephant in domesticity, FAO. Rome, Italy.

Laumonier, Y., Uryu, Y., Stüwe, M., Budiman, A., Setiabudi, B., and Hadian, O. 2010. Eco-floristic sectors and deforestation threats in Sumatra: identifying new conservation area network priorities for ecosystem-based land use planning Biodiversity Conservation, 19: 1153-1174.

Lindsay, K. 1994. Feeding ecology and population demography of African elephants in Amboseli, Kenya. Ph.D. thesis. University of Cambridge, Cambridge, U.K.

Litvaitis, J.A. 2000. Investigating food habits of teresterial vertebrates. In: Research Techniques in Animal Ecology: Controversies and Consequences. Boitani,L and Fuller, T.K. (eds). Columbia University Press. New York. USA.

McKay, G.M. 1973. Behavior and ecology of the Asiatic elephant in southeastern Ceylon. Smithsonian Contribution to Zoology 125:1-113.

Nyhus, P.J., R. Tilson, and Sumianto. 2000. Crop raiding elephants and conservation implications at Way Kambas National Park, Sumatra, Indonesia, Oryx, 34:262- 274.

Olivier, R.C.D. 1978. On the ecology of the Asian elephant. Ph.D. thesis. University of Cambridge, Cambridge, U.K.

Olson, D. 2004. Elephant Husbandry Resource Guide. Azle, Texas, International Elephant Foundation. 209-217

Russon, A.E. Wich, S.A.,Ancrenaz, M.A., Kanamori, T., Knott, C.D., Kuze, N., Morrogh-Bernard, H.C.,Pratje, P., Ramlee, H., Rodman, P., Sawang, A., Sidiyasa, K.,Singleton, I., and van Schaik, C. 2009. Geographic variation in orangutan diets. In: Orangutans: Geographic variation in Behavioral Ecology and Conservation. Wich, S.A., Suci Utami Atmoko, S., Mitra Setia, T., van Schaik, C.P. (eds). Oxford University Press, New York. USA.

Santiapillai, C., and Jackson, P. 1990. The Asian elephant: An action plan for its conservation. IUCN/SSC Asian elephant specialist group. Gland. Switzerland

Schaller, G.B. 1998. Wildlife of the Tibetan Steppe. The University of Chicago Press. Chicago.IL.USA.

Short, J.C. 1981. Diet and feeding behaviour of the forest elephant. Mammalia 45: 177- 185.

Shuguang, W., Xiaulan, P., Yulong, D., and Xianchong, W. 2009. Aboveground biomass and bamboo shoot nutrients of high altitude bamboos (Fargesia yunnanensis Hsueh et Yi) from different sites in Yunan province, China. Journal of Tropical Agriculture 47: 48-53.

Sitompul, A.F. 2004. Conservation implication of human-elephant interactions in two nationals in Sumatra. Master of Science, Thesis, University of Georgia, Athens, GA.USA.

Sokal, R.R., and Rohlf, F.J. 1995. Biometry. New York: W.H. Freeman and Co. USA.

SPSS Statistic ver 17. SPSS Inc. Chicago, IL. USA

Sukumar, R. 1989. The Asian elephant: Ecology and management. Cambridge University Press, Cambridge, U.K.

White, L.J.T., Tutin, C.E.G. and Fernandez, M. 1993. Group composition and diet of forest elephants, Loxodonta Africana cyclotis Matschie 1900, in Lopé Reserve, Gabon. African Journal of Ecology 31: 181-199.

CHAPTER 3

MOVEMENTS AND HOME RANGE BEHAVIOR OF A SUMATRAN

ELEPHANT

3.1 Abstract

Increasingly, habitat fragmentation by agricultural and human development has forced Sumatran elephants into relatively small areas; yet, there is no information on the movements and home range behaviors of elephants on Sumatra. Using a GPS collar, we determined the home range sizes of an adult female elephant in the Seblat Elephant

Conservation Center (SECC), Bengkulu Province of Sumatra in 2007-2008. We used autocorrelation analyses to assess the level of autocorrelation among elephant locations, and correlation and logistic regression analyses to examine relationships between elephant movements and monthly rainfall, vegetation productivity and distance to roads and rivers. Home range size was 97.4 km2 for the MCP and 95.0 km2 for the 95% fixed kernel. There were no relationships between average monthly elephant home range sizes or movement distances with rainfall. Distances to rivers and ex-logging roads had little effect on elephant movements, but vegetation productivity, as measured by the Enhanced

Vegetation Index, did affect movements with elephants occurring predominately in forests with intermediate canopy cover versus closed canopy forests. Consistent food and water availability in the lowland forests of the SECC in combination with high human development surrounding the center probably affect the small home range size. The remaining forested areas surrounding the SECC need protection from expanding palm

plantations and human settlements, and to enhance potential dispersal to other elephant populations in the province.

3.2 Introduction

2400-2800 wild elephants (excluding elephants in camps) in 25 fragmented populations

(Soehartono et al. 2007). Most elephant populations occur in lowland areas with upwards of 85% of their range outside of protected areas, and all populations are considered vulnerable to continuing habitat loss from large-scale habitat conversion by agriculture, human settlement, illegal logging and forest fires (Hedges et al. 2005, Soehartono et al.

2007, Uryu et al.2008). Additionally, continuing habitat loss brings elephant populations closer to human settlements, resulting in human-elephant conflict (Sukumar 1992;

Leimgruber et al. 2003; Hedges et al. 2005). These human-elephant conflicts often result in the capture and removal of elephants by the government or poisoning by local people to mitigate the conflict (Hedges et al. 2005).

Current conservation strategies for Sumatran elephants focus on securing elephant habitat and mitigating human-elephant conflict. However, there is also a critical need to link isolated elephant populations by facilitating elephant movements across the landscape (Soehartono et al. 2007). Yet, developing effective land conservation strategies for elephants is difficult because there is no information on the movements and home range behaviors of elephants on Sumatra. Most studies of Asian elephant

movement and home range behaviors were conducted on Indian elephants (Sukumar

1989; Desai 1991; Williams et al. 2001), and there are a few in Southeast Asia. Olivier

(1978) provided limited information on elephant movements and home range behaviors from his radio telemetry study in Taman Negara National Park in Malaysia.

Subsequently, Stüwe et al. (1998) reported a home range size of 350 km2 for a male elephant and 7000 km2 for a female elephant that were tracked with satellite telemetry after translocation to Taman Negara National Park.

The absence of information on Sumatran elephant movements and home range behaviors has hampered development of effective land conservation strategies for elephants on Sumatra. Consequently, land use planning and protected area management in and around elephant habitats remain ineffective. Further, fragmentation of elephant habitats into relatively small areas also complicates elephant conservation programs on

Sumatra (Santiapilai and Jackson 1990; Leimgruber et al. 2003). Thus, the purpose of this study is to report on the movement and home range behaviors of a female elephant in a lowland rainforest of Sumatra. Although only one elephant was tracked, this study provides the only information available on elephant movements on Sumatra. I also report on relationships between monthly rainfall and elephant movements, and assess if vegetation productivity and distance to roads and rivers influenced elephant movements.

3.3 Study Area

The study was conducted in the Bengkulu Province on the west coast of Sumatra and included the Seblat Elephant Conservation Center (SECC) (lat 03° 03‟12” -

03°09‟24” S, long 101° 39‟18” - 101° 44‟50” E) and surrounding forested and developed

areas (335.6 km2; Fig. 3.1). Annual rainfall typically exceeds 3000 mm and elevations are

< 50 m above sea level. Using the land cover map developed by Laumonier et al. (2010), forests comprised 23% of the land cover within a 10 km radius of the SECC with the remainder classified as non-forested. These forests are regenerating following selective logging operations in the late 1980s. Extensive palm oil plantations, small-scale agricultural areas and human settlements comprised the majority of non-forested lands. In addition to 23 elephants captured as part of the government‟s human-elephant conflict mitigation program and housed at the SECC, a population of 40-60 wild elephants is believed to occur on the SECC. With extensive agriculture and human settlements surrounding much of the SECC, there is much human-elephant conflict in the area.

3.4 Method and Analysis

On August 24, 2007, one adult (~25 years old) wild female was darted from elephant back by a veterinarian (see Chapter 5) using 7 ml of xylazine (100 mg/ml) in a dart fired from tranquilizer gun. The elephant was further sedated using 4 ml ketamine hydrocloride 100 mg/ml. intra muscular. She was fitted with a GPS collar (Africa

Wildlife Tracking, Inc, Pretoria, South Africa) and observed until fully recovered from the anesthesia. The duty cycle of the unit was set to download three GPS fixes per 24- hour period, one every eight hours (0100, 0900, 1700 hrs) from August 24, 2007 to May

14, 2008. A 9-minute GPS login time period was used for each monitoring interval.

Home range sizes were estimated using the minimum convex polygon (MCP)

(Mohr 1947) and the fixed kernel (FK) methods (Powell 2000). Despite its limitations

(Powell 2000; Osborn 2004), I used the 100% MCP estimate to facilitate comparisons

with other elephant telemetry studies. I chose to use the fixed kernel method, rather than the adaptive kernel method, because there is lower bias and better surface fit (Seaman et al. 1999) and is more reliable for estimating the outer contours and centers of activity of home ranges (Kernohan et al. 2001). For kernel estimates, I defined the area within the

95%, 90% and 50% isopleths as the „95% kernel‟, „90% kernel‟, „50% kernel‟, respectively. All home range sizes were calculated using the Hawths Tool extension in

ArcGIS 9.2 (ESRI, Inc.2007). I allowed the program to automatically select the appropriate smoothing parameters. A Spearman correlation test was used to assess the relationship between monthly elephant home range size and rainfall.

As an index of daily movement, I measured the linear distance between locations on consecutive days. Total monthly movement was calculated based on the summation of these daily movements. A Spearman correlation test was used to investigate the relationship between monthly elephant movement and rainfall.

I used a univariate correlogram (Legendre and Legendre 1998), plotting distance classes between point locations (Cliff and Ord 1981), and Moran‟s I autocorrelation coefficient (Moran 1950) to assess the level of autocorrelation among elephant locations.

GeoDaTM spatial autocorrelation analysis software (Anselin 2003) was used for all autocorrelation analyses.

I used the Information Theoretic Approach (Burnham and Anderson 2002) to examine the effects of vegetation productivity and distance to roads and rivers on elephant movements. Vegetation productivity was determined using the Enhanced

Vegetation Index (EVI) from the NASA Moderate Resolution Imaging

Spectroradiometer (MODIS) sensor. The spatial resolution of EVI MODIS was 500 m

with time series of 16 days obtained from the U.S, Geological Survey

(http://glovis.usgs.gov). The EVI provides a radiometric measure of vegetation structure and condition, providing an index (0.0 – 1.0) to variations in vegetation productivity (Gao et al. 2000). Distance to roads (ROAD) and rivers/streams (RIVER) were determined by measuring the closest distance of elephant locations to these two features. All roads in the study area were abandoned logging roads no longer used by vehicles. I also generated 99 random points as „non-elephant‟ location within the elephant home range. I used these non-elephant locations (0) in combination with observed elephant locations (1) to create a binomial dataset for the logistic regression model.

The EVI, ROAD and RIVER variables were used to develop a logistic regression model. For the regression analyses, I developed seven combinations of models to determine what variables best explained elephant movements. The interaction effect of the variables was not used in the model because I considered none of the possible interactions ecologically meaningful in explaining elephant movements. I used the 95% confidence interval to assess the effect of each variable in the model. An Akaike

Information Criteria (AIC) value was calculated for each model using log-likelihood and total number of parameters used in the model (Burnham and Anderson 2002). I also calculated model-averaged parameter estimates, and unconditional standard errors for each parameter (Burnham and Anderson 2002). I used the lowest AIC value and highest

Akaike‟s weight (ω) to determine the best model. I used R- open source statistical software (http://cran.r-project.org/) for all statistical analysis.

3.5 Results

I recorded 358 locations for the collared adult female elephant between August

2007 and May 2008. Her home range size was 97.4 km2 for the MCP and 95.0 km2 for the 95% fixed kernel (Figure 3.2). Average monthly home range size between September

2007 and April 2008 was 34.6 km2 (range = 12.4 km2 – 51.7 km2) for MCP, and 47.2 km2

(range = 28.7 km2 - 65.2 km2) for the 95% fixed kernel. There was no relation between average monthly elephant home range sizes and rainfall (rs= 0.19; P = 0.65) (Fig. 3.3).

The mean daily movement distance of the elephant was 1.5 km ± 0.3 km (1.2 –

1.9 km). Average monthly elephant movement was 36.6 km ± 4.6 km (30.7 – 43.6 km).

There was no correlation between monthly elephant movement distances and rainfall (rs =

0.55; P = 0.16) (Fig. 3.4). Over half (57%, n = 204) of elephant locations were inside the

SECC, and 41% (n = 147) in undeveloped forested areas surrounding the SECC. Only

2% (n = 7) of the locations occurred in palm plantations.

There was much autocorrelation between elephant locations (n = 350, I = 0.1268, p < 0.001), but little autocorrelation when I re-sampled the data to include only locations separated by 48-hours (I = 0.06, p < 0.07). I used these 99 re-sampled locations for the logistic regression analyses.

The mean distances of elephant locations to rivers were 286 m ± 210 (SD) and

291 m ± 198, respectively, for the complete (Fig. 3.5) and re-sampled data sets. Mean distance of elephants to roads were 686 m ± 524 and 734 m ± 494, respectively, for the complete (Fig. 3.6) and re-sampled data sets. The mean EVI value was 0.53 ± 0.09 (SD)

0.08 (Fig. 3.7). Vegetation productivity (EVI) had the largest effect on elephant

ˆ movements in the regression model ( j = -2.4871, SE = 1.792), although none of the

three parameters were significant (Table 3.3). The negative parameter estimate for the

EVI suggests that this elephant tended to occur predominately in forests with intermediate canopy cover versus closed canopy forests. The very small parameter estimates for river and roads suggest these variables may be less important factors affecting the movements of this elephant (Table 3.3).

3.6 Discussion

Although the movements of only one elephant were followed in this study, we believe this elephant represented the movements of most of the wild elephants in the

Seblat ECC. On eight occasions, we were contacted by plantation mangers when elephants were crop-raiding palm plantations. On each of these occasions, the single

GPS-collared female elephant in the study was in close vicinity of the location where elephants were reportedly crop-raiding. This coincidence of GPS locations with crop- raiding instances suggests that there may be only one breeding elephant herd in Seblat, an observation further supported by rangers who report seeing no more than one breeding herd on their regular patrols throughout the SECC.

The elephant home range size in this study was relatively small compared to ranges reported for Asian elephant studies in India, but larger than the home ranges of the four bulls tracked in Taman Negara, Malaysia (Table 3.1). In contrast, home ranges of

African elephants are substantial larger than those reported for Asian elephants (Table

3.1). The small ranges of the Sumatran and Malaysian elephants compared to Indian and

African elephants are probably most affected by the stability of environmental conditions.

In dry areas, such as the savanna and deciduous forest elephant habitats of India and

Africa, elephants tend to increase their home range sizes seasonally in search of food and water (Sukumar 1989, Lindeque and Lindeque 1991, Thouless 1995, 1996, Leggett

2006). In contrast, annual rainfall is stable and relatively high (> 3000 mm/year) in

Sumatra, providing more consistent water availability, and density and quality of palatable plants for elephants (Chapter 2). Thus, there is less need for elephants in

Sumatra to increase their home range size in search of water or food. The absence of a relationship between elephant home range size and rainfall in Sumatra further supports this hypothesis.

Concurrently, high human activity in areas surrounding the SECC may also be restricting elephant home range sizes. Numerous studies report the significant effects of human settlements and illegal hunting on elephant movement patterns (Barnes et al.

1991; Ruggiero 1992; Tchamba et al.1995; Sitati et al. 2003). The extensive palm oil plantations, land clearing for human settlements, and illegal logging around the SECC over the past 30 years pose significant barriers to elephant movements. The near absence of elephant locations in palm plantations and human settlements strongly suggest the avoidance of these areas. With few exceptions, all elephant locations occurred within the

SECC or the forested areas surrounding the SECC. Despite the occurrence of forests extending to the east and north of the SECC (Fig. 3.1), no elephant locations were recorded. Further, no elephant sign was observed in this forested area on surveys conducted by the Bengkulu Natural Resource Agency in 2007/08 (Aswin Bangun, pers. comm.). The settlement of about 200 families on the eastern border of the SECC probably blocks elephants from entering this forested area from the SECC. Further, 53% of the forest cover of the northwestern region of Bengkulu Province (including the SECC) was

lost from 1985 to 2007 (Laumonier et al. 2010), fragmenting elephant distribution into relatively small isolated populations. Thus, it is difficult for elephants to move between these isolated forest fragments.

The small home range size of the elephant in this study may not be indicative of the home range needs of elephants on Sumatra. Her movements appeared to be much restricted by human development, and there is little opportunity for dispersal out of the

SECC. Thus, additional research is needed on the home range movements of elephants in the SECC and in other elephant populations before establishing habitat management goals for elephant conservation on Sumatra.

Distances to rivers and roads did not appear to affect the elephant‟s movements in the SECC. This result contrasts greatly with African elephant movements that are greatly affected by water availability (Redfern et al. 2003, Leggett, 2006, Chamille-Jammes et al., 2007, Lee and Graham, 2006, Cushman et al. 2010), especially in semi- and arid environments. Similarly, Cushman et al. (2010) reported that elephants avoided roads in their satellite telemetry study in southern Africa. The high availability of water in the

Seblat and Air Rami rivers, abundance of tributary streams, high rainfall and small area of the SECC combined with the small home range of the monitored elephant suggests that she was never far from a water source. Similarly, distance to the abandoned logging roads also did not appear to affect the elephant‟s movements in the SECC; however, she regularly occurred near roads (Fig. 3.6). Roads may facilitate her movements to feeding areas and water. African elephants often use traditional elephant trails to move between important resources (Blake et al. 2008). Yet, the use of roads by elephants in SECC may

also increase their risk to poachers. There were at least four elephants killed by poachers on the SECC between 2007-2009.

Vegetation productivity, as measured by the EVI, was probably the factor most affecting elephant movements on the SECC. The negative value of the EVI parameter in

ˆ the model ( j = -2.4871) suggests that this elephant frequently utilized areas with more open canopies than closed canopy forests. Similarly, Chen et al. (2006) reported that elephants used early successional habitats with more abundant food resource in China.

3.7 Management Implications

Despite its limited sample size, this study provides useful information for developing an elephant conservation strategy for the SECC and surrounding areas, an important habitat for elephants on the island of Sumatra (IUCN 2008). Although the elephant habitats of the SECC are relatively secure since the logging concession closed in the 1980‟s, the remaining forested areas surrounding the SECC remain unprotected and are rapidly being converted to palm plantations and human settlements. Further, the

SECC and adjacent forested habitats appear to be fully occupied by elephants; thus, the loss of elephant habitats outside of the SECC will undoubtedly reduce the capacity of the area to sustain its current elephant population (~40-60 elephants) and will increase human-elephant conflicts in the area. Further, loss of these unprotected forested areas will also exacerbate the isolation of the SECC elephant population from the nearest other elephant populations in the northern part of the Bengkulu Province. Consequently, there is a critical need to protect the former logging concessions around SECC, including the

Production Forest Air Rami, Production Forest Air Teramang, Limited Production Forest

Lebong Kandis, Limited Production Forest Air Ipuh 1 and Air Ipuh 2. The developing carbon credit programs, such as the Reducing Emissions from Deforestation and Forest

Degradation (REDD) program, offer much opportunity for protecting these forested areas. Under this program, communities would receive direct benefits from protecting forests. District, provincial and central government agencies in combination with local communities and NGOs must work together to develop and implement these carbon credit programs.

With elephants already ranging beyond the boundaries of the SECC, the long- term viability of the SECC elephant population is uncertain. Further, recent human settlements on the eastern border of the SECC appear to pose a barrier to elephant dispersal to the east and north. The legal rights of these settlements are in question, and need to be resolved. Further, resettlement of these new communities may be an option, but may require compensation. Notwithstanding the future status of the human settlements, there is a critical need to restore the functionality of this forested corridor for elephant dispersal, thereby helping to sustain population viability, genetic variability and reducing the potential human-elephant conflicts.

Table 3.1. Male and female elephant home range sizes (100% minimum convex polygon) reported from studies in Africa and Asia.

Annual No of Location Sex Size (km2) rainfall Reference elephants (mm) Asia

South India Male 2 170-320 900 Sukumar 1989

Malaysia Male 4 32-60 2500 Olivier 1978

South India Female 2 105-115 900 Sukumar 1989

Sumatra Female 1 97 3005* This study

Africa

Lindeque and Namibia Female 7 5800-8700 315 Lindeque 1991 Amboseli NP- Western and Female 6 2756 350 Kenya Lindsay 1984

Laikipia-Kenya Female 4 450-500 750 Thoules 1996

Hwange NP- Conybeare Male 7 1300-2981 632 Zimbabwe 1991 Sengwa- Male 9 322 688 Osborn 1998 Zimbabwe Queen Elisabeth Male 6 500 900 Abe 1994 NP-Uganda

(*) Rainfall from April 2007-March 2008 (12 months)

Table 3.2. Summary logistic regression models of elephant locations with vegetation productivity (EVI), and distances to river (RVR) and roads (n=198). Models are ranked from best to worst based using Akaike‟s Information Criterion (AIC) , and associated delta (Δ AIC), Akaike weight (ω). AIC is based on –2 x log likelihood and the number of parameters in the model (K).

Model K AIC Δ AIC ωi

EVI 2 276.52 0.00 0.3399

RVR 2 278.14 1.62 0.1512

EVI + RVR 3 278.23 1.71 0.1446

ROAD 2 278.48 1.96 0.1276

EVI + ROAD 3 278.50 1.98 0.1263

RVR + ROAD 3 280.12 3.60 0.0562

EVI + RVR + ROAD 4 280.19 3.67 0.0543

Table 3.3. Model-averaged estimate, unconditional standard errors and confidence interval of effect on elephant movement in Seblat Elephant Conservation Center.

95% CI

a ˆ Parameter j SE Upper Lower

Intercept 0.9435 1.0311 2.9644 -1.0775

EVI -2.4781 1.7923 1.0348 -5.9910

RVR -0.0003 0.0006 0.0008 -0.0014

ROAD -3.51e-05 0.0003 0.0006 -0.0006

aParameter descriptions. EVI- Enhanced Vegetation Index, RVR-distance to the nearest river, ROAD-distance to the nearest ex-logging road

Figure 3.1 Location of study area in Bengkulu Province, Sumatra, and land use within a 10 km wide radius of the Seblat Elephant Conservation Center.

Figure 3.2. Home ranges [minimum convex polygon estimate (MCP) and fixed kernel density estimate (FKDE) 95%, 90% and 50% contour] for an adult female elephant, August 2007 to May 2008, Bengkulu Province, Sumatra.

Figure 3.3. Relationship between monthly home range (km2) 95% fixed kernel home range for an adult female elephant and total monthly rainfall (mm), September 2007 to April 2008, Bengkulu Province, Sumatra.

Figure 3.4. Relationship between monthly movement (km) for an adult female elephant and total monthly rainfall (mm), September 2007 to April 2008, Bengkulu Province, Sumatra.

Figure 3.5. Distance of locations for an adult female elephant to the nearest stream or river, August 2007 to May 2008, Bengkulu Province, Sumatra.

1,200

1,000

800

600

Distance (m) 400

200

1 7

91 13 19 25 31 37 43 49 55 61 67 73 79 85 97

175 277 103 109 115 121 127 133 139 145 151 157 163 169 181 187 193 199 205 211 217 223 229 235 241 247 253 259 265 271 283 289 295 301 307 313 319 325 331 337 343 349 355 Point locations

Figure 3.6. Distance of locations for an adult female elephant to the nearest road, August 2007 to May 2008, Bengkulu Province, Sumatra.

3000

2500

2000

1500 (m) Distance Distance 1000

500

1 7

13 19 25 31 37 43 49 55 61 67 73 79 85 91 97

133 223 307 103 109 115 121 127 139 145 151 157 163 169 175 181 187 193 199 205 211 217 229 235 241 247 253 259 265 271 277 283 289 295 301 313 319 325 331 337 343 349 355 Point locations

Figure 3.7. Enhanced Vegetation Index values at the locations for an adult female elephant, August 2007 to May 2008, Bengkulu Province, Sumatra.

0.90

0.80

0.70

0.60

0.50

0.40

Vegetation Vegetation Index 0.30

0.20

0.10 Enhanced

0.00

1 7

97 13 19 25 31 37 43 49 55 61 67 73 79 85 91

229 235 103 109 115 121 127 133 139 145 151 157 163 169 175 181 187 193 199 205 211 217 223 241 247 253 259 265 271 277 283 289 295 301 307 313 319 325 331 337 343 349 Point locations

3.8 Literature Cited

Abe, E. 1994. The behavioural ecology of elephants in the Queen Elizabeth National Park, Uganda. Unpublished PhD thesis. University of Cambridge.

Anselin, L. 2003. GeoDaTM 9.0 Users Guide.Spatial Analysis Laboratory. Department of Agricultural and consumer economics. University of Illinois Urbana, IL and Center for Spatially Integrated Social Science. http://geodacenter.asu.edu/ Downloaded on 23 September 2010.

Barnes, R.F.W., Barnes K.L. Alders, M.P.T. and Blom, A. 1991. Man determines distribution of elephant in rainforest of north-eastern Gabon. African Journal of Ecology 29:54-63.

Blake, S., Deem, S.L., Strinberg, S., Maisels, F., Momont, L., Isia, I-B., Douglas- Hamiltion, I., Karesh, W.B., and Kock, M.D. 2008. Roadless wilderness area determines forest elephant movements in the Congo Basin. PLos ONE 3:1-9.

Burnham, K. P. and D. R. Anderson. 2002. Model selection and inference. A practical information theoretic approach. Springer-Verlag. New York.

Chamaille-Jammes, S., Valiex, M., and Fritz, H. 2007. Managing heterogeneity in elephant distribution: interaction between elephant population density and surface-water availability. Journal of Applied Ecology 44: 625-633.

Cliff, A.D. and.Ord, J.K. 1981. Spatial processes: models and applications. Pion, London. England.

Cushman, S.A., Chase, M., and Griffin, C. 2010. Mapping landscape resistance to identify corridors and barriers for elephant movement in Southern Africa. In: Spatial complexity, informatics and wildlife conservation. Chusman, S.A and Huettmann, F. (eds).2010. Springer. Tokyo.Japan.

Conybeare, A.M. 1991. Elephant occupancy in vegetation change in relation to artificial water points in a Kalahari sand area of Hwange National Park. Unpublished PhD Dissertation. University of Zimbabwe. Zimbabwe.

Desai, A.A. 1991. The home range of elephants and its implications for the management of the Mudumalai Wildlife Sanctuary, Tamil Nadu. Journal of Bombay Natural History Society 88: 145-156

Fernando, P., Pfrender, M.E., Enclada, S.E., and Lande, R. 2000. Mithocondrial DNA variation, phylogeography and population structure of the Asian elephant. Heredity, 84:362-372.

Fleischer, R.C., Perry, E.A. Muralidharan, K. Stevens, E.E and Wemmer, C.M. 2001 Phylogeography of the Asian elephant (Elephas maximus) based on mitochondrial DNA. Evolution, 55:1882-1892.

Gao, X., Huete, A.R., Ni, Wi., and Miura, T. 2000. Optical-biophysical relationship of vegetation spectra without background contamination. Remote Sensing of Environment. 74: 609-620.

Hedges, S., Tyson, M., Sitompul, A.F., Kinnaird, M.F., Gunaryadi, D., and Aslan. 2005. Distribution, status and conservation needs of Asian elephant (Elephas maximus) in Lampung Province, Sumatra, Indonesia. Biological Conservation 124: 35-48.

IUCN 2008. IUCN range-wide mapping and strategic conservation planning workshops for Asian Elephants, Cambodia, October 20–24th, 2008

Kernohan, B. J., R. A. Gitzen, J. J. Milspaugh. 2001. Analysis of animal space use and movements. Pages 126-166 in J. J. Milspaugh and J. M. Marzluff (eds). Radio tracking and animal populations.Academic Press. San Diego. California. USA.

Lee, P.C. and Graham, M.D. 2006. African elephant Loxodonta africana and human- elephant interactions: Implications for conservation. Int Zoo Yearbk 40-9-19.

Legendre, P. and L, Legendre 1998. Numerical ecology. Development in environmental modeling, 20- Elsevier

Leggett, K.E.A. 2006. Home range and seasonal movement of elephant in the Kunene Region, northwestern Namibia. African Zoology 41:17-35.

Lindeque, M and Lindeque, P.M.1991. Satellite tracking of elephants in northwestern Namibia. African Journal of Ecology 29: 196-206.

Mohr 1947. Table of equivalent populations of North American small mammals. American Midland Naturalist 37: 223-249.

Moran, P.A.P. 1950. Notes on continuous stochastic phenomena. Biometrika 37:17-23.

Olivier, R.C.D. 1978. On the ecology of the Asian elephant. Unpublished Ph.D thesis. University of Cambridge, Cambridge, UK.

Osborn, F.V.1998. The ecology of crop-raiding elephants in Zimbabwe. Unpublished Ph.D thesis. University of Cambridge. UK.

Osborn, F.V.2004. The concept of home range in relation to elephants in Africa. Pachyderm 37:37-43

Powell, R.A.2000. Animal home ranges and territories and home range estimators. In: Research technique in animal ecology. Boitani, L and Fuller, T.K.(eds).Columbia University Press. New York.

R.ver 2.11. The R project for statistical computing. http://www.r-project.org/ Downloaded on 4 May 2010.

Redfern, J.V., Grant, R., Biggs, H., and Getz, W.M. 2003. Surface-water concerns on herbivore foraging in the Kruger National Park, South Africa. Ecological Society of America 84: 2092-2107.

Ruggiero, R.G. 1992. Seasonal forage utilization by elephants in central Africa. African Journal of Ecology 30: 137-148.

Santiapilai, C., and Jackson, P. 1990. The Asian elephant: An Action Plan for its Conservation. IUCN/SSC Asian Elephant Specialist Group. Gland, Switzerland.

Seaman, D.E. J. J. Millspaugh, B. J. Kernohan, G. C. Brundige, K. J. Raedeke, and R. A. Gitzen. 1999. Effects of sample size on kernel home range estimates. Journal of Widlife Management 63: 739-747.

Sitati, N.W., Walpole, M.J., Smith, R.J., and Leader-Williams, N. 2003. Predicting spatial aspects of human-wildlife conflict. Journal of Applied Ecology 40: 157- 164.

Stüwe, M., Abdul, J.B., Nor B.M., and Wemmer, C.M. 1998. Tracking the movements of translocated elephants in Malaysia using satellite telemetry. Oryx 32, 68-74.

Sukumar, R. 1989. Ecology of the Asian elephant in Southern India.I. Movement and Habitat Utilization Patterns. Journal of Tropical Ecology 5: 1-18.

Sukumar, R. 1989. The Asian elephant: ecology and management. Cambridge University Press. Cambridge, UK.

Sukumar, R. 1992. The Asian elephant: An ecology and management, second ed. Cambridge University Press, Cambridge, UK.

Tchamba, M.N., Bauer, H. and Dejongh, H.H. 1995. Application of VHF radio and satellite telemetry techniques on elephants in northern Cameroon. African Journal of Ecology 33:335-346.

Thouless, C.R. 1995. Long distance movements of elephants in northern Kenya. African Journal Ecology 33:321-334

Thouless, C.R. 1996. Home ranges and social organization of female elephants in northern Kenya. African Journal of Ecology 34:284-297.

USGS. 2009. Global Visualization Viewer. http://glovis.usgs.gov . Downloaded 30 October 2009.

Western, D. and Lindsay, W.K. 1984. Seasonal herd dynamics of a savanna elephant population. African Journal of Ecology 22:229-244.

Williams. A.C., Johnsingh, A.J.T., and Krausman, P.R. 2001. Elephant-human conflicts in Rajaji National Park, northwestern India. Widlife Society Bulletin 9: 1097- 1104.

CHAPTER 4

HABITAT USE OF AN ADULT FEMALE SUMATRAN ELEPHANT

4.1 Abstract

Increasingly, habitat fragmentation by agricultural and human development has forced Sumatran elephants into relatively small areas; yet, there is no information on the habitat use of elephants on Sumatra. Using a GPS collar and a land cover map developed from TM imagery, we determined the habitats used by an adult female elephant in the

Seblat Elephant Conservation Center (SECC), Bengkulu Province of Sumatra in 2007-

2008. We used resource selection and compositional analysis habitat ranking approaches to describe habitat use. The elephant used medium canopy and open canopy forests more than expected; however, during the day closed canopy forests were used more than at night. This elephant tended to avoid open areas. Effective elephant conservation strategies in Sumatra need to focus on forest restoration of cleared areas and providing a forest matrix that includes various canopy types.

4.2 Introduction

Since the 1990s, the Sumatran elephant (Elephas maximus sumatranus) has declined by approximately 35% to an estimated 2,400-2,800 elephants in the wild

(Soehartono et al. 2007). Elephants occur in 25 fragmented populations in lowland areas, and all populations are considered vulnerable to continuing habitat loss from large-scale habitat conversion by agriculture, human settlement, illegal logging and forest fires

(Leimgruber et al. 2003, Hedges et al. 2005, Soehartono et al. 2007, Uryu et al.2008).

Additionally, continuing habitat loss brings elephants closer to human settlements, often resulting in the capture and removal of elephants by the government or poisoning by local people (Hedges et al. 2005).

Current conservation strategies for Sumatran elephants focus on securing elephant habitat and mitigating human-elephant conflict. Yet, developing effective land conservation strategies for elephants is difficult because there is no information on the habitats used by Sumatra elephants. Thus, the purpose of this study is to report on the habitats used by a single satellite-tagged female elephant in a lowland rainforest of

Sumatra. Although only one elephant was used, this study provides the only information available on elephant habitat use on Sumatra. I also compare two different habitat use analysis approaches for describing resource selection by this female elephant in Sumatran rainforests.

4.3 Study Area

The study was conducted in the Bengkulu Province on the west coast of Sumatra and included the Seblat Elephant Conservation Center (SECC) (lat 03° 03‟12” -

03°09‟24” S, long 101° 39‟18” - 101° 44‟50” E) and surrounding forested and developed areas (335.6 km2; Fig. 3.1). Annual rainfall typically exceeds 3000 mm and elevations are

< 50 m above sea level. Using the land cover map developed by Laumonier et al. (2010), lowland rainforests (Pesisir-Indrapura-Talamau ecofloristic sector) comprised 23% of the land cover within a 10 km radius of the SECC with the remainder classified as non- forested. These forests are regenerating following selective logging operations in the late

1980s. Extensive palm oil plantations, small-scale agricultural areas and human

settlements comprised the majority of non-forested lands. In addition to 23 elephants captured as part of the government‟s human-elephant conflict mitigation program and housed at the SECC, a population of 40-60 wild elephants is believed to occur on the

SECC. Other endangered large mammals such as Sumatran tiger (Panthera tigris sumatrae), Malayan tapirs (Tapirus indicus) and Malayan sun bears (Helarctos malayanus) also occur on the SECC. With extensive agriculture and human settlements surrounding much of the SECC, there is much human-elephant conflict in the area.

4.4 Methods and Analyses

4.4.1 Telemetry

On August 25, 2007, one adult (~25 years old) wild female elephant was darted from elephant back by a veterinarian (see Chapter 5) using 7 ml of xylazine (100 mg/ml) in a dart fired from tranquilizer gun. The elephant was further sedated using 4 ml ketamine hydrocloride 100 mg/ml. intra muscular. She was fitted with a GPS collar

(Africa Wildlife Tracking, Inc, Pretoria, South Africa) and observed until fully recovered from the anesthesia. The duty cycle of the unit was set to download three GPS fixes per

24-hour period, one every eight hours (0100, 0900, 1700 hrs) from August 25, 2007 to

May 14, 2008. A 9-minute GPS login time period was used for each monitoring interval.

All elephant locations were plotted on LANDSAT Thematic Mapper (TM) satellite images and later entered into GIS format using ArcView GIS version 3.3 (ESRI).

4.4.2 Land cover classification

I created a land cover map for the study area using Landsat TM 2005 satellite images. I assigned land cover classes to the image using supervised classification techniques using ERDAS IMAGINE 8.4. I initially, 20 categories were classified based on reflectance excluding bands 6 and 8. I re-classified these 20 categories into five land cover types: closed canopy forest, medium canopy forest, open canopy forest, open area and water (Table 4.1). These five broad categories were selected considering the accuracy of land cover classification, ecological significance for elephants and subsequent habitat management by resource managers. To avoid problems of including habitat that the GPS- tagged elephant may not have access to in the study area (Garshelis 2000), I restricted the analyses to all available habitats within a 10-km radius of a central circular point statistic of all elephant location data (Fig 4.1).

4.4.3 Habitat use analysis

I used two analytical approaches for describing habitat use, the resource selection statistical approach (Manly et al. 1993) and the habitat ranking statistical approach using compositional analysis (Johnson 1980; Aesbischer et al. 1993). Both approaches are based on the proportion of time that animals spend in each habitat type in comparison to the relative available habitat (e.g., Neu et al. 1974; Johnson 1980; Manly et al. 1993:

Aebischer et al. 1993).

4.4.3.1 Manly’s Resource Selection

Using the procedures of Manly et al. (1993), I calculated the Resource Selection

Index (ŵi) by comparing the observed number of elephant GPS locations in each habitat type to habitat availability (expected use based upon proportions of each habitat in study area). The resource selection ratios (ŵi) were then standardized (B) and chi-square goodness of fit tests used to identify if there was significant use of a habitat category.

Chi-squared values were then compared with a chi-squared distribution statistic with k-1 degrees of freedom. When a significant difference is detected, I then used Bonferroni Z- statistic to determine habitat selection ratios (the habitat type used more or less frequently than expected (a =0.05). If the confidence interval for resource selection ratios does not contain the value of 1, then selection for that habitat is inferred

4.4.3.2 Habitat ranking using Compositional Analysis

I examined 2nd order resource selection and 3rd order resource selection (Johnson

1980) using Compositional Analysis (Aebischer et al. 1993). Second order selection is defined as selection of a home range within the study area and third order selection is selection of habitat types within the home range (Johnson 1980). I then assessed significant deviation of habitat use from random use and ranked habitat types from most to least used, at each level of habitat selection using multivariate analysis of log ratio test

(Aebischer et al. 1993). I also used a similar procedure for 3rd order selection to determine whether elephant habitat use varied by time using locations at 0900 hrs for diurnal activity and locations at 0100 hrs for nocturnal activity.

For all compositional analyses, I calculated habitat use on a monthly basis; therefore, this analysis is not representative of a population-level analysis. I selected the first five days of elephant locations in each month to ensure independent observations

(Aebischer et al. 1993). I calculated habitat composition in the total study areas in the elephant‟s home range (Minimum Convex Polygon [MCP]), and as the proportion of elephant locations within each habitat type using ArcView GIS ver 3.3 (ESRI Inc.). Prior to the compositional analysis, I replaced zero values with the value of 0.001%, an order of magnitude less than the smallest recorded nonzero proportion (Aesbischer et al. 1993).

4.5 Results

4.5.1 Manly’s Resource Selection

The single female elephant used habitats significantly different from random in proportion to availability (χ2 = 21.512, df = 4, P<0.001). Medium canopy and open canopy area tended to be used more compared to the other three habitat types (Tables 4.2,

4.3). Closed canopy, open area and water habitat categories were used less than expected, but were not significant

4.5.2 Habitat ranking using Compositional Analysis

The 2nd order compositional analysis of the female‟s home ranges within the study area showed that her habitat use was non-random (Λ = 0.1497, χ2 = 15.191, randomized P< 0.001). Elephant habitats ranked from most to the least use included medium canopy forest > open canopy forest > closed canopy forest > open area > water

(Table 4.4). Similarly, 3rd order habitat use within the MCP ranges also differed from

random use (Λ = 0.2271, χ2 = 11.856, randomized P < 0.001) with habitats ranked as follows: medium canopy forest > open canopy forest > water > open area > closed canopy forest (Table 4.4). Elephant habitat use also differed from random use for diurnal

(Λ = 0.0015, χ2 = 52.108, randomized P< 0.001) and nocturnal (Λ = 0.06, χ2 = 21.594, randomized P< 0.001) locations. However, this female used open canopy forest more during the night in contrast to her use of medium and closed canopy forests during the day (Table 4.5).

4.6 Discussions

Although the habitat use of only one elephant was described in this study, we believe this elephant represented the habitat use of most of the wild elephants in the

Seblat ECC. On eight occasions, we were contacted by plantation mangers when elephants were crop-raiding palm plantations. On each of these occasions, the single

Overall, the female elephant in Seblat used medium canopy and open canopy forests more often than expected, and these results were consistent for both of the analytical methods used despite differing statistical approaches. High use of medium canopy and open canopy forests may be related to food availability. Medium canopy forests appeared to have abundant browse including bamboos and rattan, while open

canopy forests had abundant grasses (e.g Poaceae family). Chen et al. (2006) reported reduced availability of many important elephant food plants such Dendorcalamus spp.

(Poaceae), Musa acuminata (Musaceae) and Microstegium ciliatum (Poaceae) with the loss of secondary and early successional forests in Xishuangbanna Nature Reserve in

China. The lower rank of closed canopy forests in elephant use in Seblat may be related to the relatively low abundance of elephant foods in closed canopy forests as indicated by the reportedly low densities of elephants in tropical forests (Olivier 1978; Sukumar 1989;

Hedges et al. 2005). The low use of open area and water habitats probably reflects their infrequent use by elephants for water and minerals.

The high use of closed canopy habitat during the day in contrast to night is probably related to thermal regulation and the shade provided by the closed canopy during the day. Thermoregulation was also observed in other herbivore species where shaded areas are preferred when solar radiation is maximum (Demarchi and Bunnel 1993;

Beyer and Haufler 1994). Valeix et al (2007) reported that giraffe, buffalos and zebra occurred more often in closed canopy forest during the hottest period of the day, shortening the time period of access to water because they avoided staying in open area to protect them from direct solar radiation.

In conclusion, Sumatran elephants use a variety of forest types, ranging from open to closed canopy forests. Open and medium canopy forests are probably the most important habitats for feeding, whereas closed canopy forests may be most important for thermoregulation. Yet, elephants tended to avoid open areas. Thus, effective elephant conservation strategies in Sumatra need to focus on forest restoration of cleared areas and providing a forest matrix that includes various canopy types.

Table 4.1 Habitats descriptions and proportion of study area (335.6 km2) used by elephants at Seblat, Bengkulu, Sumatra.

Prop. of Habitat Class Description study area

Area with closed canopy forest and dense tree Closed canopy 0.4549 vegetation.

Area with broken canopy or rare standing tree vegetation. This area also mainly covered with Medium canopy 0.2827 secondary vegetation or tall shrub vegetation including bamboo vegetation.

Area with no tree vegetation and dominated Open canopy with secondary vegetation, shrub or Alang- 0.2264 alang (Imperata cylindrica).

This area mostly bare ground or area with rare Open area small vegetation mostly grass (e.q Poaceae 0.0174 family) or small shrub.

Water Water body including ponds, stream or river. 0.0185

Table 4.2. Resource selection indices for habitats used by an adult, female elephant from 25 August 2007 to 14 May 2008, at Seblat, Bengkulu, Sumatra.

Population Sample Expected Manly Selection Selection Habitat Proportion count count standardize ratio (ŵ) level (π) (u) (π *ut) Index (B)

Closed Canopy 0.4549 126 162.413 0.7758 0.1687 "-"

Medium Canopy 0.2827 135 100.929 1.3376 0.2908 "+"

Open Canopy 0.2264 87 80.837 1.0762 0.2340 "+"

Open area 0.0174 5 6.220 0.8038 0.1748 "-"

Water 0.0185 4 6.601 0.6059 0.1317 "-"

1 357 4.5994 1

Table 4.3. Bonferroni confidence intervals for proportions of habitats used by an adult, female elephant from 25 August 2007 to 14 May 2008, at Seblat, Bengkulu, Sumatra.

Bonferroni Sample Used sample Selection confidence limits Habitat count proportion Sig P<0.05 ratio(ŵ) (u) (o) Lower Upper

Closed Canopy 0.776 126 0.353 0.75 2.46 NS

Medium Canopy 1.338 135 0.378 1.32 4.30 S

Open Canopy 1.076 87 0.244 1.06 3.46 S

Open area 0.804 5 0.014 0.80 2.60 NS

Water 0.606 4 0.011 0.61 1.96 NS

4.599 357

Table 4.4. Habitat ranking matrix of five habitats used by an adult, female elephant from 25 August 2007 to 14 May 2008, at Seblat, Bengkulu, Sumatra based upon: A). MCP home ranges vs. total study area (2nd order selection) and B). GPS locations vs. MCP home ranges (3rd order selection). Higher ranking indicates greater use compared to availability. Within the matrix (+) represent the row habitat is preferred than column habitat whereas a (-) represent the opposite. Triple sign represent significant deviation from random at P<0.05.

A).

Closed Open Medium Open Water Rank Canopy Area Canopy Canopy

Closed Canopy + - - +++ 2

Open Area - - --- + 1

Medium Canopy + + + +++ 4

Open Canopy + +++ - + 3

Water ------0

B).

Closed Open Medium Open Water Rank Canopy Area Canopy Canopy

Closed Canopy ------0

Open Area + - - - 1

Medium Canopy +++ + + + 4

Open Canopy +++ + - + 3

Water + + - - 2

Table 4.5. Habitat ranking matrix of five habitats used by an adult, female elephant from 25 August 2007 to 14 May 2008, at Seblat, Bengkulu, Sumatra based on A) nocturnal activity (0100 hr) and B) diurnal activity (0900 hr). Higher ranking indicates greater use compared to availability. Within the matrix, (+) represents the row habitat is preferred over the column habitat, whereas a (-) represents the opposite. Triple sign represent significant deviation from random at P<0.05.

A).

Closed Open Medium Open Water Rank Canopy Area Canopy Canopy

Closed Canopy +++ ------+ 2

Open Area ------0

Medium Canopy +++ +++ --- +++ 3

Open Canopy +++ +++ +++ +++ 4

Water - + ------1

B).

Closed Open Medium Open Water Rank Canopy Area Canopy Canopy

Closed Canopy +++ --- + +++ 3

Open Area ------0

Medium Canopy +++ +++ + +++ 4

Open Canopy - +++ - +++ 2

Water --- + ------1

Figure 4.1. Location of study area in Bengkulu Province, Sumatra, and land use within a 10 km wide radius of the Seblat Elephant Conservation Center.

4.7 Literature Cited

Aebischer, N.J., Robertson, P.A., and Kenward, R.E. 1993. Compositional analysis of habitat use from animal radio-tracking data. Ecology 74: 1313-1325.

Beyer, Jr, D.E., and Haufler, J.B. 1994. Diurnal versus 24-hour sampling of habitat use. Journal of Wildlife Management 58:178-180.

Chen, J., Deng, XB, Zhang L, Bai ZL. 2006. Diet composition and foraging ecology of Asian elephants in Shangyong, Xishuaangbanna, China. Acta Ecologica Sinica 26: 309-316.

Demarchi, M.W. and Bunnell, F.L. 1993. Estimating forest canopy effects on summer thermal cover for Cervidae (deer family). Canada Journal of Forest Resources 23: 2419-2426.

Garshelis, D.L. 2000. Delusions in habitat evaluation: Measuring use, selection, and Importance. In Boitani, L. & Fuller, T.K. (eds) Research Techniques in Animal Ecology: Controversies and Consequences. Columbia University Press. Pp. 111– 164.

Johnson, D.H. 1980. The comparison of usage and availability measurement for evaluating resource preference. Ecology 61: 65-71.

Manly, B.F.J., Mc Donald, L.L., Thomas, D.L. 1993. Resource selection by animals, statistical design and analysis for field studies. Chapman & Hall, London, UK.

Neu, C.W., Byers, C.R., and Peek, J.M. 1974. A technique for the analysis of utilization- availability data. Journal of Wildlife Management 38: 541-545.

Olivier, R. 1978. On the ecology of the Asian elephant. Unpublished Ph.D thesis, University of Cambridge.

Sukumar, R. 1989. The Asian elephant: Ecology and Management. Cambridge University Press. Cambridge.UK

Valeix, M., Fritz, H., Matsika, R., Matsvimbo, F., Madzikanda, H., 2007. The role of water abundance, thermoregulation, perceived predation risk and interference competition in water access by African herbivore. African Journal of Ecology 46: 402-410.

CHAPTER 5

USE OF TAME ELEPHANTS TO DEPLOY GPS TELEMETRY UNITS ON

WILD ELEPHANTS IN SUMATRAN RAINFOREST

In many elephant ranges in Africa and South Asia, where elephants live in open semiarid-savannas habitats, researchers immobilize study elephants from helicopters

(Osofsky 1993, Chase 2008, Kikoti 2009), vehicles (Kikoti 2009) or on foot (e.g. in

Williams et al. 2001). Open areas with long-range visibility provide optimum conditions for researchers to identify target animals and increase darting success rate. This situation, however, is uncommon in rainforest habitat. Dense vegetation and closed canopy cover limits researchers‟ ability to find animals. In high-density elephant areas such as Dzanga-

Ndoki National Park, Central Africa, researchers search for forest elephants on foot and dart the animal from the ground (Blake et al. 2001). In the tropical rainforests of Sumatra, the elephant population is relatively low density, and occurs mostly in dense secondary and primary forests, sometimes in hilly terrain.

Thus, the purposes of this paper are to report on the use of tame elephants to search for and help immobilize wild elephants for attachment of telemetry collars in the lowland tropical rainforest habitat of northern Bengkulu Province, Sumatra. I also provide suggestions for increasing deployment success of the GPS units and reducing the potential of injury to personnel, and the tame and wild elephants.

5.1 Use of tame elephants to locate wild elephants

In 1986, the policy of the Indonesian Government was to capture “problem elephants” and hold them at Elephant Training Centers (ETCs) (Santiapillai and Ramono,

1993; Lair 1997). The original purposes of ETCs were to reduce human-elephant conflicts, increase ecotourism activities and use tame elephants to patrol protected areas

(McNeely 1978; Lair 1997). Most of the ETCs in Sumatra are located near protected elephant habitat such as national parks, wildlife sanctuaries or nature reserves. The formerly wild elephants held at an ETC are trained by mahouts for regular daily activities

(e.g., drinking, feeding, bathing) and periodic health care checks. Some of the adult tamed elephants at the ETCs that are well controlled by mahouts can be used to search for wild elephants, providing several advantages for darting wild elephants in the rainforest.

First, tame elephants can help researchers cover larger areas of forest and increase researcher visibility when searching for wild elephants. Second, wild elephants are less wary of people when riding an elephant versus a person walking, thereby increasing the probability of finding wild elephants. Third, tame elephants are often able to approach within a closer distance of a wild elephant, facilitating more accurate darting. Lastly, more field supplies to facilitate darting and collaring can be brought into the field on elephant back.

5.2 Locating wild elephants for immobilization using tame elephants

The use of 3-4 adult tame elephants, either male or female, is optimum for immobilizing wild elephants. Typically, there are 7-9 people involved in the darting and collaring, including a mahout for each elephant, veterinarian, ranger with the dart rifle, and the researcher. Considering the high potential for injury to the tame elephants and personnel, it is critical that the mahout is experienced and able to maintain the confidence of his tame elephant when faced with the aggressive behavior of a wild elephant. Team members must also remain calm and maintain their balance atop the elephant. If they fall off, there is a high likelihood that the wild elephant will quickly attack them.

We encountered four wild adult female elephants, two in August 2007 and two in

April 2008, and deployed two collars. Of these four wild elephants we approached for darting, one charged, presumably the matriarch, hitting the lead tame elephant repeatedly with her head and trunk. One of the three tame elephants used in this first encounter was scared and ran away from the attacking wild elephant, with the mahout eventually regaining control of this elephant within 10 min. The mahouts on the other two tame elephants were able to stand their ground against the attacking elephant. The attacks of the single wild elephant continued for upwards of 10 min before the combined efforts of the three tame elephants and shouting by the team stopped the attacks and the wild elephant retreated. No attempts were made to dart this elephant. Within 15 min, another adult female was encountered. We approached from behind within 20 m without her turning to observe our approach. A dart was fired, she trumpeted and ran away, and we did not pursue her for approximately 10-15 min. We followed her trail via trampled vegetation, but encountered several other elephants. Within 30 min, we re-sighted her

within 200 m of the darting site. She was standing, but immobilized. The third elephant encountered for collaring in April 2008 was an adult female with an ~ 2 yr old calf.

Similar to the second elephant, we approached from behind to within 20 m without her turning. A dart was fired; she trumpeted and ran to her calf < 30 m away. Stopping briefly to look at us, raise her trunk to smell the air, before running away with her calf.

We noticed that the dart had hit her rump at an angle, appearing to not be fully inserted.

We were unable to relocate this darted female after searching for 30 min and suspect that drug was not fully injected. While searching for this female, we encountered another sub-adult female. She did not turn to look at us upon our approach. Without a second loaded dart for the rifle, we approached within 2 m and used an aluminum jab stick to injected her. She trumpeted and ran away. After waiting for 10 min, we followed her trail, and re-sighted her on the top of a ridge leaning against a tree, approximately 200 m from the immobilization location.

5.3 Anesthetizing wild elephants

All elephants were darted from behind in the posterior part of dorsal ilium to avoid potential injury to internal organs. Our goal was to conduct a “standing sedation” in which the elephant still stands after initial immobilization (Fowler et al. 2000) to avoid respiratory depression that may result from the elephant laying on its side or sternum.

Hypoxemia is significant risk factor for sternally-recumbent (Harthoorn 1973, Honeyman et al. 1992) immobilized elephants, and it is near impossible to move an elephant from a sternal to a lateral position once fully sedated and recumbent on the forest floor. A

standing sedation also reduces the risk of drowning if a darted elephant runs to a river or falls on its sternum in steep terrain.

When first darted, we used 7 ml xylazine (Rompun® ; 100 mg/ml) (Hsu, 1981) in a dart fired from a long-range rifle tranquilizer gun or a jab stick. First signs of immobilization of elephants by xylazine usually involve a combination of slow movement or stopping, snoring, mild head weaving, and for bull elephants, penis relaxation (Cheeran 2008). For the two elephants relocated after darting, one was standing and the other standing against a tree. To prevent the immobilized elephant from falling to the ground, a tame elephant was positioned to one side of the wild elephant

(Fig. 1). If there is risk to falling to the other direction, a second tame elephant can be positioned on the other side, or an adjacent tree used for added support. To reduce the risk of attack from other elephants in the vicinity, the other tame elephants may be used to guard the team members.

Once the position of the standing wild elephant was secure with the tame elephants, the veterinarian confirmed whether the elephant was fully immobilized (i.e. unable to move its trunk or legs). If not, then a second sedative of 4 ml ketamine hydrocloride (100 mg/ml) is administered intra muscular using local injection (Wisnu

Wardana, pers comm.). Both of the elephants we collared received ketamine injections.

Sterile ophthalmic ointment was applied to the eyes of elephant prior to blindfolding the elephant with a soft towel (Osofsky 1993). While the team attached the collar, the veterinarian followed an anesthesia protocol, including: cardiothoracic auscultation, palpation of auricular pulse for quality and regularity, checking of rectal temperature, and monitoring respiratory and heart rates (Osofsky 1993). We also injected 500,000 IU

potassium penicillin G and1000,000 IU procaine penicillin, and applied 500 mg neomycin at the injection and dart locations to reduce risk of infection (Osofsky 1993).

Other immobilzation drugs are typically used with elephants in Africa [ethorphine hydrochloride with hyaluronidase (M99); Kock et al. (1992), Osofsky (1993)] and in

India [ethorpine with ACP (Immobilon LA®); Cheeran (2008)]. However, these drugs typically cause the elephant to fall down preventing a standing sedation. Considering the risks of injury to sedated elephants posed by hilly terrain and close proximity of rivers in our study area, we used the standing sedation technique.

Additionally, morphometric measurements were made to estimate elephant age

(circumference of front and rear leg, shoulder height, total body length and chest circumference). A blood sample was also taken for genetic and parasitological study.

Once the collar was attached, yohimbine (2 mg/ml) 0.125 mg/kg body size was administered to reverse the effect of xylazine (Hatch et al. 1985; Jessup et al. 1983).

After administering the reversal drug, the tame elephants are repositioned 10-20 m from the wild elephant for the team to confirm recovery before leaving the area.

5.4 Summary

Locating and capturing wild elephants in tropical rainforest environments are difficult and high-risk tasks. However, using tame elephants improves the search efficiency of finding wild elephants in dense forests and reduces risks to staff and target elephants during the immobilization process. Use of experienced veterinarians and standing sedation techniques greatly reduce the risks of elephant injury while immobilizing elephants. Tame elephants with experienced mahouts and veterinarians

increase the success of elephant collaring studies in forested areas, the safety of wild elephants and personnel during immobilization, and the value of tame elephants for elephant conservation programs.

Figure 5.1. Collaring wild elephant using “standing sedation” technique in Seblat Sumatra.

5.5 Literature Cited

Blake, S., Douglas-Hamilton, I. and Karesh, W.B. 2001. GPS telemetry of forest elephants in central Africa: result of preliminary study. African Journal of Ecology 39: 178-186.

Chase, M. 2007. Home ranges, transboundary movement s and harvest of elephants in northern Botswana and factors affecting elephant distribution and abundance in the Lower Kwando River Basin. Unpublished Ph.D Dissertation. University of Massachusetts, Amherst, MA. USA.

Cheeran, J.V. 2008. Drug Immobilization: Yesterday, Today and Tommorow. The IUCN Asian Elephant Specialist Group Journal. Gajah. 28: 35-38

Hatch R.C., Kitzman, J.V., Zahner, J.M. and Clark, J.D. 1985. Antagonism of xylazine sedation with yohimbine, 4-aminopyridine, and doxapram in dogs. American Journal of Veterinary Research. 46: 371-375.

Harthoorn, A.M. 1973. Review of wildlife captures drugs in common use. In The capture and care of wild animal. E Young (editor). Human and Rousseau Publishers Ltd, Pretoria, Republic of South Africa. 14-34.

Honeyman, V.L., Pettifer, G.R., and Dyson, D.H. 1992. Arterial blood pressure and blood gas values in normal standing and laterally recumbent African (Loxodonta africana) and Asian (Elephas maximus) elephant. Journal of Zoo and Wildlife Medicine 23: 205-210.

Hsu W.H. 1981.Xylazine-Induced Depression and Its Antagonism by Alpha Adrenergic Blocking Agents. The Journal of Pharmacology and Experimental Therapeutics. 218: 188-192.

Jessup, DA, Clark, W.E., Gullet, P.A.,Karen, R., and Jones, B.S. 1983. Immobilization of mule deer with ketamine and xylazine and reversal of immobilization with yohimbine. Journal of the American Veterinary Medical Association 183:1339- 1340.

Kikoti, A. 2009. Seasonal home range sizes, transboundary movements and conservation of elephants in northern Tanzania. Unpublished Ph.D Dissertation, University of Massachusetts, Amherst, MA.USA.

Kock, M.D. 1992. Use of hyalurodinase and increased etorphine dosage (M99) to improve induction time and reduced captured-related stress in the chemical immobilization of free-ranging of black rhinoceros (Diceros bicornis) in Zimbabwe. Journal of Zoo and Wildlife Medicine 23: 181-188.

Lair, R.C. 1997. Gone Astray: The care and management of Asian elephant in domesticity. FAO, Rome. Italy.

Osofsky, S. A. 1997. A practical anesthesia monitoring protocol for free-ranging adult African elephants (Loxodonta africana). Journal of Wildlife Diseases 33: 72-77.

Santiapillai, C. and Ramono, W.S. 1993. Reconciling elephant conservation and economic development in Sumatra. The IUCN Asian Elephant Specialist Group Journal. Gajah 10: 11-19.

Williams. A.C., Johnsingh, A.J.T., and Krausman, P.R. 2001. Elephant-human conflicts in Rajaji National Park, northwestern India. Wildlife Society Bulletin 9: 1097- 1104.

CHAPTER 6

CONSERVATION IMPLICATIONS

Despite decades of conservation efforts for Sumatran elephants (Elephas maximus sumatranus), little information was available on their foraging ecology, movement and home range behaviors, and habitat use before this study. Considering that most Sumatran elephants live outside of protected areas, conservation action for Sumatran elephants should focus on protecting remaining unprotected elephant habitats and habitat restoration to improve elephant habitat quality in the future. Furthermore, recommendations to develop corridors connecting protected areas may have little effect for elephant conservation given the relatively small numbers of elephants in protected areas. Thus, immediate and urgent action is needed to save as much elephant habitat as possible through better land use planning that incorporates elephant habitats outside of protected areas.

6.1 The importance of elephant conservation in Sumatra

Why conserve elephants? What is in it for the people of Sumatra? These two questions are fundamental and often raised from the local people to the politicians and government agencies in Sumatra. The importance of conserving elephant populations in

Sumatra involves several aspects including, ecological, socio-economical, and ethical perspective. From the ecological point of view, as large mammals, elephants play important role in the ecosystem. Several studies show that elephants are important seed dispersal agents in rain forest habitat (Lieberman et al. 1987, Powell 1997). Furthermore, elephants play important roles in creating habitats for other species. From the socio-

economic aspect, Sumatran elephants live in large areas with complex ecosystem functions. These ecosystems are critically important for the people of Sumatra, providing important resources for local culture (e.g traditional medicine, or dye for coloring traditional cloths), food and clean water, stabilizing local climate, and protecting humans from natural disasters (e.g., flooding, severe drought and landslide) and human-wildlife conflict. Thus protecting elephant habitat is also protecting people. Finally, the elephant is a charismatic mega-fauna and considered a flagship species because of their distinctiveness and attractiveness. Therefore, the Sumatran people have a responsibility to the global community to protect the species and to prevent its extinction.

6.2 Foraging ecology and natural diet of Sumatran elephant

Ex-logging concession area such as the Seblat Elephant Conservation Center is still considered important habitat for elephant. This study shows that nutritional quality of the majority of elephant diet is in the optimum level to support elephant to growth.

Elephant range in Sumatra that is part of ex-logging concession across Sumatra should be carefully managed. Currently, at least 1.6 million ha of elephant habitat is part of ex- logging concession across Sumatra (Soehartono et al. 2007). Habitat conversion into large scale monoculture plantation such as palm oil and Acacia.sp tree for pulp and paper industry or mining activities should be prohibited. Instead, management of these areas should be focus on habitat restoration that tie into the Reducing Emission from

Deforestation and forest Degradation (REDD) program (UN-REDD, 2009). Indonesia is one of the nine pilot countries designated as pilot project for the United Nations-REDD program in 2008 and eligible for developing an alternative finance mechanism to

maintain the integrity of remaining forested area in Sumatra (UN-REDD, 2009). The

Indonesian government and other stakeholders (conservation NGO‟s, community groups) should take the opportunity and use the program to better manage the ex-logging concession areas in Sumatra in a sustainable way (e.g., see Sitompul and Pratje 2009).

Sustainable payment scheme for ecosystem services should also be explored for future management of these ex logging concession areas (Redford and Adams 2009). At the local level this approach has been successfully implemented in Africa as alternative strategy to promoting biodiversity conservation (Nelson et al. 2009), and should be tested in Sumatra.

6.3 Sumatran elephant movement and home range behavior

The small home range size of elephants in the SECC area suggests that this population is isolated and the area may not be large enough to support elephant population in the long term. Management of the elephant habitat in SECC and surrounding habitat should focus on maintaining habitat connectivity of Production

Forest area in the western side of the Kerinci Seblat National Park. Further, conversion of remaining forests in the Production Forest areas into plantations, mining or human settlement should not be allowed. With increasing habitat loss around the SECC, several impacts may occur, including: 1) increased contact between elephants and agriculture/human settlements that will increase human-elephant conflicts and increase the risk of elephant mortality; 2) increased potential for inbreeding depression; 3) higher risks of locak extinction due to low population viability. The result of this study also shows that elephant movements more affected by the vegetation productivity compared

to the relative distance to rivers and roads. Therefore management of elephant habitat in future needs to balance these environmental characters in order to facilitate optimal elephant movement on their ranges.

6.4 Sumatran elephant habitat use

Understanding on the habitat use for Sumatran elephant is critically important to develop effective conservation strategies of the species across the island. Further habitat protection clearly need detail information on habitat otherwise the designation of protected area for elephant will be mislead. For example, in the past, designation of protected area in Sumatra did not include elephant habitat as a parameter to determine the delineation of conservation area (Soehartono et al. 2007). Therefore, as a results most of the current elephant distribution area in Sumatra is not protected.

The results of this study indicate that Sumatran elephants extensively use open canopy and medium canopy forests. However closed canopy forests are also important as shade and possible protection from anthropogenic threats such as hunting. Conserving elephant habitats outside of protected areas in Sumatra should focus on areas containing these various habitat types. Further, ex-logging concession areas with medium canopy forest should not be allowed to be converted to the large scale development activities such as monoculture plantation, mining or human settlement. Many of these areas may provide suitable habitats for elephant. On the other hand, allocating all the conservation effort to save remaining forested area in Sumatra might not necessarily help elephant conservation in the future. Therefore, conservation strategy for Sumatran elephant should be carefully developed balancing the species conservation (e.g. Sumatran elephant) and

ecosystem services (e.g. water source, clean air, etc) when designing land use planning in the region. Combination of habitat mosaic that incorporates different habitat configuration should take into account to protect elephant habitat in a landscape level.

6.5 Literature Cited

Lieberman, D., Lierberman, M., and Martin,C. 1987. Notes on seed in elephant dung from Bia National Park, Ghana. Biotropica 19:365-369.

Nelson, F., Foley, C., Foley, L.S., Leposo, A., Loure, E., Peterson, D., Peterson, M., Peterson, T., Sachedina, H., Williams, A. 2009. Payment for ecosystem services as a frame work for community-based conservation in northern Tanzania. Conservation Biology 24: 78-85.

Powel, J.A. 1997. The ecology of forest elephants (Loxodonta africana cyclotis Matschie 1900) in Bangyang-Mbo and Korup forests, Cameroon, with particular reference to their role as seed dispersal agents. Unpublished Ph.D. thesis, University of Cambridge, Cambridge, U.K.

Redford, K.H., and Adams, W.M. 2009. Payment for ecosystem services and challenge of saving nature. Conservation Biology 23: 785-787.

Sitompul, A.F., and Pratje, P. 2009. The Bukit Tigapuluh Ecosystem Conservation Implementation Plan. Frankfurt Zoological Society, Frankfurt. Germany.

United Nation-REDD, 2009. From the web: http://www.un-redd.org/ downloaded September 11, 2010.

BIBLIOGRAPHY

Abe, E. 1994. The behavioural ecology of elephants in the Queen Elizabeth National Park, Uganda. Unpublished PhD thesis. University of Cambridge.

Aebischer, N.J., Robertson, P.A., and Kenward, R.E. 1993. Compositional analysis of habitat use from animal radio-tracking data. Ecology 74: 1313-1325.

Altman, J. 1974. Observational study of behavior: sampling methods. Behaviour 49:227- 267.