PHYLOGEOGRAPHY AND DIFFERENTIATION OF WHITE-NEST SWIFTLET (Aerodramus fuciphagus)

IN MALAYSIA

Goh Wei Lim

Master of Science (Molecular Ecology)

2007

Pbsat Khidmat Maklumat Akademik üNIVERSITI MALAYSIA SARAWAK

P. KHIDMAT MAKLUMAT AKADEMIK

uiiuuiýlýiýuia 1000246088

PHYLOGEOGRAPHY AND DIFFERENTIATION OF WHITE-NEST SWIFTLET (AERODRAMUS FUCIPHAGUS)

IN MALAYSIA

Goh Wei Lim

A thesis submitted in fulfillment of the requirement for the degree of Master of Science

Faculty of Resource Science and Technology UNIVERSITY MALAYSIA SARAWAK

2007

Acknowledgement

This MSc thesis is the result of two-year-work whereby many people and a few institutions

have supported me.

I am especially grateful to my main supervisor, Associate Professor Dr Mustafa Abdul

Rahman. Two years ago when I was only a fresh graduate, it was him who responded to my enquiry

and encouraged me to do postgraduate study in molecular ecology and evolution. I appreciate his

guidance to me through out these years. I would also like to express my sincere gratitude to my

co-supervisors, Dr Lim Chan Koon, who gave me the opportunity to go into the world of swiftlet. He

is willing to teach and to help me with his expertise in swiftlet. Many other academic staff in FSTS

had been very helpful to me. I learnt a lot from them and appreciate their guidance.

Thanks to the Unversiti Malaysia Sarawak and its Faculty of Resource Science and

Technology for providing me the financial support by Zamalah Unimas. The project is funded by

Unimas short-term grant 248/2001(7).

I am indebted to Dr. Charles Leh, Malaysia Birds' Nest Merchants Association, especially Mr

Lim, Mr. Lee, and Mr. Yap, for providing the samples of this study.

During these years in the Land of Hornbill, my life has been cheered up by a lot of good

coursemates and friends. Thanks to Dency, Imelda, Wani, Shima, Then, Jong, Lee San, Lee Tse, Jolly,

Xiang Ru, Eunice, Cindy, Tiong, Joanna, Wai Hwa and all of the friends in FSTS, who strived in the

laboratories days and nights with me. Thanks to all my housemates for their companionship.

This is a special experience in my life. I would not have the courage to travel alone without

the support my family members as well as my old friends in Peninsular Malaysia. I wish I could know

the best way to show my deepest gratitude to all of them.

11

Phylogeography and Differentiation of White-nest Swiftlet

(Aerodramusfuciphagus) in Malaysia

Abstract

The edible birds'-nest builder, white-nest swiftlet (Aerodramus fuciphagus), has received

much attention in Malaysia over past few decades. It has become one of the important natural

resources in Malaysia and contributed to the profitable birds'-nest industry. Due to the great market

demand, wild populations of white-nest swiftlet are declining. In recent years, swiftlet farming

becomes a new trend in edible bird nests industry in Malaysia and also in the Southeast Asia.

Phylogeographic study of the white-nest swiftlet populations could provide the information

of the white-nest swiftlet subspecies identification, recent gene flow among the white-nest swiftlet

populations and the recent population expansion. Thus, this study is aimed to: (a) resolve the dispute

over the classification within species; and (b) investigate the patterns and structure of the relationships

among the populations in Peninsular Malaysia, Sarawak and Sumatra, in which the resulted

information will be useful in the strategy planning for white-nest swiftlet conservation and for the

swiftlet farming.

In this study, six populations had been studied. They are middle Baram, Sibu, Sitiawan,

Selangor, Kuantan, Endau-Rompin and Sumatra. Based on phylogenetic analyses of the 558 bp of the

mitochondrial cytochrome b and 462 bp of the control region, there are two major clades with a deep

divergence among all the white-nest swiftlet populations studied. Clade II is closely related to the

Aerodramus fuciphagus vestitus while Clade I is closely related to A. f germani. The grouping of

different clades is inconsistent with the geographical distribution. However, Clade II is not participated

by the individual from the east coast of Malaysia, while Clade I is not participated by the individual

from middle Baram. The statistical analyses supported the phylogenetic analyses. The results suggest

111

that there is no substantial gene flow among the populations in Clade I, Clade II and `Garman

Cluster'. The possible Pleistocene refugia were suggested based on the test of population expansion.

Further investigation by using longer DNA sequence and involving larger sample size is

needed to confirm whether the Clade I and Clade II could be classified into different subspecies. Both

cytochrome b and control region have shown their potential in the phylogeographic study of the

white-nest swiftlet.

Keywords: white-nest swiftlet, phylogeographic study, cytochrome b, control region, Southeast Asia.

iv

Filogeografi dan Perbezaan di antara Burung Layang-layang

(Aerodramusfuciphagus) di Malaysia

Abstrak

Burung layang-layang ('swiftlet ) adalah salah satu sumber semulajadi yang penting di Malaysia

kerana industri sarang burung memberi hasil yang lumayan. Walau bagaimanapun, permintaan yang

tinggi terhadap sarang burung menyebabkan populasi semula jadi burung layang-layang terganggu.

Kebelakangan ini, penternakan burung layang-layang ('swiftlet farming) telah menjadi satu industri

yang baru di Malaysia and juga di Asia Tenggara.

Kajian filogeografi terhadap burung layang-layang. dapat memberi informasi tentang identifikasi

sub-spesies, aliran gen di antara populasi-populasi, dan pengembangan populasi. Oleh itu, objektif

kajian ini adalah: (a) menjelaskan klasifikasi di peringkat subspesies dan (b) mengkaji struktur

hubungan antara populasi-populasi di Semenanjung Malaysia, Sarawak dan Sumatra. Informasi yang

didapati adalah penting untuk konservasi populasi semula jadi dan juga untuk penternakan burung

layang-layang.

Dalam kajian ini, enam populasi telah dikaji, iaitu Baram Tengah, Sibu, Sitiawan, Selangor,

Kuantan, Endau-Rompin dan Sumatra. Analisisfilogenetik telah dijalankan terhadap 558 bp daripada

gen 'cytochrome b' dan 462 bp daripada gen 'control region' di dalam DNA mitokondria. Dua

kumpulan yang major telah didapati. Kumpulan II mempunyai hubungan yang lebih rapat dengan

sub-spesies Aerodramus fuciphagus vestitus manakala Kumpulan I mempunyai hubungan yang lebih

rapat dengan A. f germani. Kumpulan II tidak mempunyai individu daripada pantai timur

Semenanjung Malaysia. Kumpulan Ipula tidak mempunyai individu daripada Baram Tengah. Analisis

stastistik menyokong keputusan analisis filogenetik. Hasil kajian ini mencadangkan bahawa

V

pengaliran genetik tidak berlaku di antara Kumpulan I, Kumpulan II dan `Germani Cluster'.

Beberapa kemungkinan kawasan hutan 'refugia'pada zaman Pleistocene telah dicadangkan.

Kajian yang seterusnya boleh dijalankan dengan menggunakan jujukan DNA yang lebih panjang

dan saiz sampel yang lebih besar. Dengan ini, klasifikasi sub-spesies mengikut Kumpulan I dan

Kumpulan II mungkin dapat dibuat. Kedua-dua gen 'cytochrome b' dan gen 'control region'

masing-masing mempunyai potensi untuk digunakan dalam kajianfilogeografi burung layang-layang..

Kata-kata kunci: burung layang-layang, kajian filogeografi, 'cytochrome b, 'control region, Asia

Tenggara.

V1

Pusat Khidmat Maklumat Akademik UVNERSITI MALAYSIA SARAWAK

Table of Content Acknowledgement

Abstract

Table of Content

List of Tables and Figures

List of Abbreviations and Symbols

11

lll

vii

X

xiii

Chapter One - Literature Review

1.1 Phylogeography 1

1.1.1 Contribution of phylogeography 2

1.1.2 Phylogeeography of Southeast Asia - special focus on Peninsular 3

Malaysia and Borneo

1.1.2.1 Historical biogeography of Southeast Asia 5

1.1.2.2 Genetic variation in relation to the historical biogeography in Southeast Asia 10

1.2 Avian Mitochondria) DNA

1.2.1 Recombination of mitochondrial DNA

1.2.2 Control region 1.2.3 Cytochrome b

1.3 The Study Species

1.3.1 Introduction to the Apodiformes

1.3.2 White-nest swifllet (Aerodramusfuciphagus)

1.3.2.1 Description

1.3.2.2 Geographic distribution

1.3.3 Taxonomy and evolution of white-nest swiftlet 1.3.3.1 Tribal and generic levels

1.3.3.2 Species and sub-species levels

1.3.4 Economic Importance and Conservation of White-nest Swiftlet

1.3.4.1 Bird nests industry

1.3.4.2 Swiftlet farming

1.4 Objectives of the Study

12

14

14

17

18

18

19

20

20

21

22

23

27

27

29

31

vii

Chapter Two - Materials and Methods

2.1 Sampling Location 32

2.2 Samples and Preservation 35

2.3 DNA Extraction 34

2.4 Primers Design 37

2.4.1 Cytochrome b 37

2.4.1 Control Region 37

2.5 PCR Amplification, PCR Production Purification and Cycle Sequencing 40

2.6 Data Analyses 41

2.6.1 Sequence Editing and Alignment 41

2.6.2 Phylogenetic Analyses 41

2.6.3 Congruency of Combined Data 42

2.6.4 Genetic Distance and Minimum Spanning Tree 42

2.6.5 Statistical Analyses 43

2.6.5.1 Pairwise comparison FsT and analyses of molecular variance (AMOVA) 43

2.6.5.2 Mismatch distribution 43

Chapter Three - Results

3.1 Cytochrome b 44

3.1.1 Cytochrome b sequence 44

3.1.2 Phylogenetic and geographic distribution of the cyt b haplotypes 47 3.1.3 Statistical analyses of the genetic structure based on cytochrome b 52

3.2 Control region 54

3.2.1 Control region sequence 54

3.2.2 Phylogenetic and geographic distribution of the control region haplotypes 62

3.2.3 Statistical analyses of the genetic structure based on control region 66

3.3 Control Region and Cytochrome b 68

3.3.1 Partition homogeneity test 68

3.3.2 Phylogenetic and geographic distribution of the combined control region and 69

cyt b haplotypes

3.3.3 Statistical analyses of the genetic structure based on the combined control 74

region and cyt b sequence

3.3.4 Test for an expanding population 78

Vlll

Chapter Four - Discussions

4.1 Cytochrome b and Control Region as the Genetic Markers 81

4.2 Inta-specific Taxonomy of the White-nest Swiftlet 82

4.2.1 Current status and possible scenario 82

4.2.1.1 Clade II 86

4.2.1.2 Clade I 86

4.2.2 Gene flow and population structure of the white-nest swiftlets in Malaysia and 88

Sumatra

4.3 Speculation of Historical Gene flow and Forest Refugia 89

Chapter Five - Conclusion 95

References 99

Appendices 109

Appendix I 109

Appendix II 119

ix

List of Tables and Figures

Figure 1.1: Wallace's Line, Huxley's Line, Weber's Line in Sunda Shelf to mark the area of greatest faunal changes in Southeast Asia.

Figure 1.2a: Palaeogeographic reconstruction of the succession of island separation during one cycle of glaciation/ deglaciation in the Malay Archipelago.

Figure 1.2b: Expected phylogenetic relationships of a species of a monophyletic group which speciated through successive vicariance of a widespread ancestor, based on the allopatric-speciation mode of evolution.

Figure 1.2c: Expected phylogenetic relationships for a group which speciated after successive dispersal from the continent.

Figure 1.3: Schematic structure of typical avian mitochondrial genome.

Figure 1.4: A different mitochondrial gene arrangement for the orders Piciformes, Cuculiformes, and Falconiformes and for suboscine Passeriforms.

Figure 1.5: Schematic structure of avian mitochondrial DNA control region.

Figure 1.6: Distribution of the White-nest Swiftlets, Aerodramusfuciphagus.

Figure 1.7: The distribution of the recognized subspecies of Collocalia fuciphaga by Medway, 1966.

Figure 2.1: Geographical locations of the sampling sites. Indonesia is indicated by the bold border lines.

Figure 2.3: The universal positions of the primers.

Figure 3.1: Variable sites of the cytochrome b region from 11 haplotypes in white-nest swiftlets.

Figure 3.2: Neighbour Joining bootstrap tree based on cytochrome b using MEGA v. 3.0.

Figure 3.3: Maximum parsimony bootstrap consensus tree (bootstrap = 1000) based on cyt b using PAUP v. 4.10.

Figure 3.4: Minimum spanning tree of 11 haplotypes found in mtDNA cyt b region in white-nest swiftlet.

Figure 3.5: Variable sites of the control region from 39 haplotypes in white-nest swiftlets.

Figure3.6: The alignment of the A. fuciphagus mtDNA sequence obtained in this study with three other avian mtDNA sequences.

Figure 3.7: Neighbuor joining consensus tree (bootstrap = 1000 rep) based on the partial control region of white-nest swiftlet. Only bootstrap values more than 50 % were shown.

X

Figure 3.8: Maximum parsimony consensus tree (bootstrap = 1000 rep) based on the partial control region sequence of white-nest swiftlet.

Figure 3.9: Minimum spanning tree of 39 haplotypes found in mtDNA control region in white-nest swiftlet.

Figure 3.10: Neighbuor joining consensus tree (bootstrap = 1000 rep) based on the combined control region and cyt b sequence of white-nest swiftlet.

Figure 3.11: Maximum parsimony consensus tree (bootstrap = 1000 rep) based on the combined control region and cyt b sequence of white-nest swiftlet.

Figure 3.12: Minimum spanning tree of 36 haplotypes found in mtDNA (the combined sequence of cyt b and CR) in white-nest swiftlet.

Figure 3.13: Frequency distribution of pairwise sequence differences among all the individuals in this study.

Figure 3.14: Frequency distribution of pairwise sequence differences among the Clade I populations in this study.

Figure 3.15: Frequency distribution of pairwise sequence differences for the Clade II individuals.

Figure 3.16: Frequency distribution of pairwise sequence differences for the Clade I after omitting the individuals from `Germani Cluster' (Clade Ia).

Figure 4.1: Illustration to show the overlapping zone between Clade I and Clade II. Sibu population is eliminated for this diagram.

Figure 4.2: The location of the possible refugia in Southeast Asia suggested by Taylor et al. (1998) and Morley (1998).

Table 1.1: The White-nest swiftlets subspecies in Malaysia.

Table 2.1: Global Positioning System (GPS) coordinates, names of the sampling locations and the dates of sample collection.

Table 2.2: The list of the cyt b primers used in the present study.

Table 2.3: The primers used to design the primers of control region in this study.

Table 2.3. The arrows indicate the direction of the primers.

TWO. 1: Summary of standard and molecular diversity indices based on the cyt b sequence of the white-nest swiftlets populations in Malaysia and Sumatra.

Table 3.2: Summary of the mtDNA cytochrome b haplotypes distribution in white-nest swiftlet.

Table 3.3: Matrix of pairwise FST values among six populations of white-nest swiftlet based on cyt b sequence.

X1

Table 3.4: Analysis of molecular variance (AMOVA) of cyt b haplotypes of white-nest swiftlet populations.

Table3.5: Summary of standard and molecular diversity indices based on the control region sequence of the white-nest swiftlets populations in Malaysia and Sumatra.

Table 3.6: Summary of the mtDNA control region haplotypes distribution in white-nest swiftlet.

Table 3.7: Matrix of pairwise FST values among six populations of white-nest swiftlet based on control region sequence.

Table 3.8: Analysis of molecular variance (AMOVA) of control region haplotypes of white-nest swiftlet populations.

Table 3.9: Summary of the mtDNA haplotypes (combined sequence of cytochrome b and control region) distribution in white-nest swiftlet.

Table 3.10: Matrix of pairwise FST values among the re-defined populations of white-nest swiftlet based on the combined sequence of control region and cyt b region.

Table 3.11: Analysis of molecular variance (AMOVA) based on the haplotypes of the combined control region and cyt b sequence of white-nest swiftlet populations.

xii

List of Abbreviations and Symbols

bp - base pair

DNA - Deoxyribonucleic acid

kb - kilo base pair

KCl - Pottasium chloride

MgC12 - Magnesium choride

Mya - Million years ago

µl - microlitre

xiii

CHAPTER ONE

Introduction

1.1 Phylogeography

With the development of the molecular methodology such as polymerase chain reaction (PCR)

and automated sequencing technology, geneticists are able to shift their research focus from typically

phylogenetic studies of higher level taxa to phylogeography studies that address within-species

variability (Emerson et al., 2001). The discipline of `phylogeography' is placed as a branch of

biogeography, which is on a par with `ecogeography' (Avise et al., 1987; Avise et al., 1994).

With particular emphasis on mitochondrial DNA (mtDNA) data, Avise et al. (1987)

introduced the term `intraspecific phylogeography', which focuses on the examination of two aspects:

a) within-species phylogenetic interrelationships among the DNA molecules and b) geographic

distribution of the phylogenetic groupings (Avise et al., 1987). In other words, phylogeography

reflects the demographic history and the evolutionary processes in relation to geographic structure of

the populations. Unlike the traditional population genetic using summary-statistics methods, which

start with the Wright-Fisher equation, phylogeography is gene-tree-based (Hey and Machado, 2003).

There are three phylogeographic hypotheses (Avise et al., 1987), which might serve as the

`stimulus' for the further investigations: a) Most species are composed of geographical populations

whose members occupy different branches of an intraspecific phylogenetic tree; b) Species with

limited or `shallow' phylogeographic population structure have life histories conducive to dispersal

and have occupied ranges free of long-standing impediments to gene flow; and, c) Monophyletic

groups distinguished by large phylogenetic gaps usually arise from long-term extrinsic

(biogeographic) barriers to gene flow.

Within the discipline of phylogeography, mitochondrial data serves as a useful introduction to

the types of intraspecific population structure. The development in phylogeography is mainly relied on

1

the use of animal mtDNA. Mitochondrial DNA is commonly used because of the following features

(Avise et al., 1987; Kocher et al., 1989; Avise et al., 1994; Randi, 2000):

a. It is maternally inherited;

b. It has non-recombining nature;

c. It has relatively low rate of base substitution; and

d. It is relatively ease to be isolated and analyzed

1.1.1 Contribution of Phylogeographic Study

The knowledge of phylogeography structure of a species would help the biogeographers to

understand better the historical biogeography (Dodson et al., 1995; Randi, 2000). Dodson et al. (1995)

studied the phylogeographic structure in mtDNA of Southeast Asia freshwater fish, Hemibagrus

nemurus, and concluded that the Pleistocene sea-level changes on the Sunda Shelf region did not

promote gene flow among populations on different islands.

Phylogeography study is also useful to resolve the taxonomy disputes within species by

constructing the phylogenetic trees. It provides a bridge between systematics and population genetics

(Avise et al., 1987). As an example, Zhi et al. (1996) suggested that the Orang-utan (Pongo pygmaeus)

from Sumatra and Borneo are two distinct subspecies.

In the context of conservation implications, the populations with higher genetic diversity and

the most ancestral haplotype could be determined by phylogeographic study (Surget-Groba et al.,

2002). Warren et al. (2001), in their study on Bomean Orang-utan populations, suggested that all

populations are showing significant genetic diversity. They proposed that all populations in Borneo

should be protected in each geographic region to conserve their genetic diversity. Identification of

taxonomic status, Evolutionary Significant Unit (ESU) and Management Unit (MU) in the

2

phylogeographic study could also be extremely useful in strategy planning of conservation works

(Moritz, 2002).

1.1.2 Palaeogeography of Southeast Asia - special focus on the Peninsular Malaysia and Borneo

Southeast Asia, ranged from 15° to 47 °N and 74 ° to 144 ° E, consists of nine countries,

namely Vietnam, Laos, Myanmar, Cambodia, Thailand, Malaysia, Singapore, the Philippine and

Indonesia. Indochina and Peninsular Malaysia are connected to the mainland of Asia. Apart from them

are numerous islands of Indonesia, islands of the Philippines, Borneo, and Papua-New Guinea.

Geologically, Southeast Asia is one of the most intriguing parts of the world.

Paleaogeographical evidence shows that it has very complex historical tectonic movements, which

involved subduction, amalgamations, accretions, rotations and collisions (Metcalfe, 1990; 1991; 1998;

Hall, 1996; 1998; Holloway, 1998).

Palaeogeography together with biogeography can provide insights on the definition of areas of

endemism, evolution patterns and process, molecular clock calibration and the dating of the fossils,

and the influence of climate changes (Holloway. 1998). When discussing on the relationship of

palaeogeography and biogeography, Holloway (1998) admitted that there are geological knowledge

which `set the scene' for the biological concepts, but there is rarely contribution of biogeography to

the geological studies. However, objective comparison of biogeographic information with geological

information, both of which are still incomplete, remains a major challenge despite recent

methodological advances in both sides (Holloway, 1998).

Mainland Southeast Asia comprises four principal terranes, South China, Indochina, Sibumasu

(now Thailand, west coast of Peninsular Malaysia, north Sumatra and the Strait of Malacca) and East

Malaya (now east coast of Malaysia Peninsula and south Sumatra, and extended to the south-west

South China Sea; Metcalfe, 1990; 1991; 1998). It is not until Jurassic and Cretaceous period when a

number of smaller terranes, including southwest of Borneo, were accreted to the continental core of

3

Southeast Asia to form the `Sundaland' (Metcalfe, 1990; 1991; 1998). The northeastern Borneo is

described as the `younger additions' to the continental core. The subduction and accretion became

increasingly important in northeastern Borneo during Late Tertiary (between 20 - 10 Ma).

Palaeomagnetic evidence shows that there was a counter-clockwise rotation of the island of Borneo

since 30 Ma (Hall, 1996; 1998; Moss and Wilson, 1998). There is a continuous land connection

between Borneo and mainland Southeast Asia throughout much of the Tertiary, which might allow the

migration of the terrestrial biota (Hall, 1998; Moss and Wilson, 1998).

1.1.2.1 Historical Biogeography of Southeast Asia

Although the full extent of Southeast Asia has tropical climate, the biogeographers had

discovered that the flora and faunal distribution in Malay Archipelago is significantly unequal. The

Wallace's Line was proposed by Alfred Russel Wallace, the foremost representative of classical

zoogeography (Mayr, 1944), based on the variation of avifauna in Southeast Asia. The Wallace's Line

starts from east of the Philippines and extends south to separate Borneo from Sulawesi and Bali from

Lombok. Oriental avifauna presents on the west of the line, whereas Australian avifauna presents on

the east of the line. His work had drawn a great attention of naturalists. Later, a few lines were

proposed by several biogeographers based on the distribution of different groups of flora and fauna

(figure 1.1). Mayr (1944) suggested that the Wallace's Line 1863 is not the boundary between the

Indo-Malayan and the Australian Regions, but rather it indicates the edge of Sunda shelf that emerged

during the Pleistocene glaciations. Weber's Line is preferable to Wallace's Line, if a single borderline

between the Oriental and the Australian Regions is to be found (Mayr, 1944).

Instead of drawing a line to indicate species boundary of certain taxa, some zoogeographers

investigated the inter-island relationships in Southeast Asia from various aspects, such as species

richness, endemicity, fossil records, numerical taxonomic methods and non-metric multidimensional

scaling (Holloway, 1968; Groves, 1984, Heaney, 1991; Reis and Garong, 2001). Michaux (1995)

suggested four distribution patterns of fauna in west Wallecea based on the data compilation of the

4

PeSat Khidmat Makiumat Akademik UNIVERSITI MALAYSIA SARAWAK

fauna record. Pattern 3 (in which its generalized track is Borneo-Sumatra-Malaya Peninsula) is almost

exclusively confined to the bird distributions.

It has also long been recognized that the last glaciation during late Pleistocene greatly

influenced the floral and faunal distribution in Southeast Asia. The last glacial maxima had caused the

sea level to decrease for more than 100 meters (Haile, 1969). The vast areas of the Sunda Shelf (Figure

1.1) were exposed and land bridges among the mainland and islands were created (Verstappen, 1975;

Hewitt, 2000). It was predicted that the landmasses were predominated by savannah vegetation

(Verstappen, 1975). Avise and Walker (1998) revealed that Pleistocene is a time of active population

differentiation that led to speciations, provided the environmental conditions are conducive to

fostering the survival. The flooding and re-colonization of Southeast Asian continental shelves,

occurred even before the quarternary, had caused the expansion and contraction of the rain forest.

These processes act as an important factor of speciation and promote the regional biodiversity

(Morley, 1998).

Based on the palaeogeographic reconstruction of the sequence and timing of the connections

between the present Sunda islands (Figure 1.2), Brooks and McLennan (1991) had formulated two

scenarios on the mode of speciation and colonization by terrestrial organisms. The first mode of

allopatric speciation occurs when the islands split off one by one from the continent (Figure 1.2a, stage

4 to stage 1). The vicariant event would lead to the phylogeny depicted in Figure 1.2b. The second

mode of allopatric speciation is more accurately described as sequential dispersal, in which the

animals colonized the unoccupied islands following the succession connection of the islands (Figure

1.2a, stage 1 to stage 4). The expected phylogeny for this mode of colonization is shown in Figure 2c.

However, phylogeographic study of Gymnures (Genus Holomys) suggested that the second mode

(sequential dispersal) is more appropriate to describe the genetic variation of Holomys taxa (Ruedi and

Fumagalli, 1996). It was also predicted that during the glacial periods when sea level was lowered,

`savannah-like' habitat with `gallery forest' dominated the dry land connecting to Malay Peninsula and

Borneo (Verstappen, 1975; Heaney, 1991).

5

Legend

Historical shoreline of Sunda and Sahul shelves

Figure 1.1: Wallace's Line, Huxley's Line, Weber's Line in Sunda Shelf to mark the area of greatest faunal changes in Southeast Asia (Briggs, 1987).

6

Indochina

Glacial

Co co m

.. ýa E ý U,

romeo

Fava

Stage 1

Interglacial B

J

Stage 2

IN pý

-4 04

Interglacial Interglacial

J

Stage 3

i

M

B

J s

Stage 4

Figure 1.2a: Paleaogeographic reconstruction of the succession of island separation during one cycle of glaciation and deglaciation in the Malay Archipelago (Ruedi and Fumagalli, 1996). M- Malaya, S- Sumatra, B- Borneo, J- Java, I- Indochina, arrows pointing to the right represent glaciation while arrows pointing to the left represent interglacial period during deglaciation.

7

Figure 1.2b: Expected phylogenetic relationships of a species of a monophyletic group which speciated through successive vicariance of a widespread ancestor, based on the allopatric-speciation mode of evolution (Ruedi and Fumagalli, 1996). For the abbreviation, see Figure 1.2a.

IMSBJ

Figure 1.2c: Expected phylogenetic relationships for a group which speciated after successive dispersal from the continent (Ruedi and Fumagalli, 1996). For the

abbreviation, see Figure 1,2a.

9

1.1.2.2 Genetic Variation in Relation to the Historical Biogeography in Southeast Asia

Some previous workers suggested that the sea level changes during the late Pleistocene do not

promote substantial gene flow among populations in different islands on the Sunda Shelf. In a study on

genetic variation of a freshwater fish, Hemibagrus nemurus, Dodson et al. (1995) revealed that there is

only very little exchange of genetic groups occurred between the mainland and the Sunda Islands

during recent glaciations. However, they recognized a genetically and morphologically distinct

Sundaic Glade, which colonized the Kapuas River in east Borneo, east Sumatra and southern Peninsula

Malaysia.

Based on the genetic structure on the gymnures (Hylomys), Ruedi and Fumagalli (1996) also

concluded that drops in sea level did not promote migration among distant areas of the Sunda Shelf.

Zhi et al. (1996) suggested that the effective or persistent hybridization between the Orang-utan

(Pongo pymeaus) from Sumatra and Borneo has been rare or absent for 1-2 million years, even when

the two islands were inter-connected during late Pleistocene. Morphology and genetic analyses on the

cave fruit bat (Eonycteris spelaea) in Indonesian Archipelago showed that there is a considerable

divergence among the populations on the large islands of Sunda Shelf (i. e., Sumatra, Java, Kalimantan

and Sulawesi). On the other hand, the populations on Nusa Tenggara Islands (Lombok to Timor),

which are distinct from those from Sunda Islands, form a more cohesive cluster (Hisheh et al., 1998;

Maharadatunkamsi et al., 2003). Latest study on historical biogeography of three Sunda Shelf Murine

rodents (Maxomys surifer, M. whiteheadi and Leopoldamys sabanus) also showing no evidence to

support the hypothesis of broad Pleistocene migrations in Sunda Shelf (Gorog et al., 2004).

However, certain islands on Sunda Shelf had been proved to have closer relationship. Based

on the allozyme electrophoresis, the Asian shrew (Crocidura) populations from Sumatra are associated

with those of Java, Sulawesi, and most other Sunda species, except for Peninsula Malaysia species

(Ruedi, 1996). Genetic evidence also suggested that the Sumatran rhinoceros (Dicerorhinus

sumatrensis) from Peninsula Malaysia and from Sumatra have only shallow divergences, but those in

Borneo are different subspecies (Morales et al., 1996). The study on the little spiderhunter

10

(Arachnothera longirostra) conducted by Rahman (2000) suggested that populations from Peninsula

Malaysia and Borneo were showing a substantial gene flow, whereas populations from Thailand were

emerged from different refugia.

There are also several studies on genetics variation among populations within island. In the

study on Sulawesi grasshopper (Chitaura), Walton and Butlin (1997) found that the levels of genetic

divergence among various parts of the island is very high, which they could not explain by isolation

only in the last one or two million years ago (Pleistocene epoch). Another study conducted by Warren

et al. (2001) concluded that the four populations of Bornean Orang-utan subspecies (Pongo pygmaeus

pygmaeus) within Borneo island are genetically isolated populations which are in the process of

divergence.

1.2 Avian Mitochondrial DNA

Desjardins and Morais (1990), for the first time, published the avian mtDNA complete

genome mapping, using chicken (Gallus gallus) as the modal. The size of the mtDNA is about 16700

nucleotide (nt) (Desjardins and Morais, 1990; Härlid et al., 1998; Ruokonen and Kvist, 2002;

Ruokonen et al., 2004), with a few exceptions of 17611 nt in penguin (Slack et al., 2003), 17622 nt in

Oriental white stork (Genbank accession number: AB026193) and 18068 nt in falcon (Genbank

accession number: N0000878). The avian mitochondrial genome contains the same genes as all other

vertebrates, but the order of those genes is unique (Figure 1.3; Desjardins and Morais, 1990; Quinn,

1997). However, Mindell et al. (1998) proposed an alternative gene arrangement in the avian

mitochondrial genome (Figure 1.4), which is found in Picidae, Cuculidae, suboscines of Passeriformes

and Falconiformes.

11