i

,Morphometric Analysis of Fawn Roundleaf Bat, Hipposideros cervinus (Gould, 1854) (Chiroptera: Hipposideridae) from Several Populations in Malaysian Borneo

Josephine anak Juary

This project is submitted in partial fulfilment of the requirements for a Bachelor of Science with Honours

(Animal Resource Science and Management)

Faculty of Resource Science and Technology Universiti Malaysia Sarawak

2011

-

Acknowledgement

•

I would like to thank my supervisor, Miss Siti Nurlydia Sazali for her guidance and comments

on my final year project. She was kind enough to share her knowledge and experience with me

and inspired me to the best for this project. I wish to express my gratitude to Prof. Dr. Mohd

Tajuddin Abdullah, whose grant made the specimens available for this study. A special thank

to Mr. Jayaraj Vijaya Kumaran for willing to provide me with some information regarding the

study site and also for his comments and suggestions. I would also like to thank Department of

Zoology for the facilities and equipments provided for the completion of my final year project.

I am also grateful for the assistances from Department of Zoology staffs, especially Mr. Huzal

Irwan Husin, Mr. Nasron Ahmad and Mr. Isa Sait. Not forgotten, to all my course mates, for

their support, ideas and cooperation throughout the years of my study. Last but not least, I

would like to thank my parents for their supports and motivations.

I

------

•

Table of Contents

Acknowledgement I

Declaration II

Tables of Contents III

List of Abbreviations IV

List of Tables and Figures V

Abstract I

1.0 Introduction 2

2.0 Literature Review 5 2.1 Taxonomy 5 2.2 The Discovery of Echolocation in Bats 5 2.3 Distribution ofHipposideros cervinus in Sarawak 6 2.4 Morphometric Studies 7

3.0 Materials and Methods 10 3.1 Study Sites 10

3.1.1 Kubah National Park 10 3.1.2 Bako National Park 11 3.1.3 Busuk Cave, Tubau 11 3.1.4 Niah National Park 12 3.1.5 Poring Hot Springs 12

3.2 Laboratory Method 13 3.3 Data CoUection 13 3.4 Data Analysis 15

4.0 Results 17

5.0 Discussion 23 5.1 Morphological Variation 23 5.2 Morphometric Analysis 26

6.0 Conclusion and Recommendation 28

7.0 References 29

35Appendices

III

•

•

List of Abbreviations

External measurement D3MCL - third digit metacarpal length D3PI L - third digit first phalanx length D3P2L - third digit second phalanx length D4MCL fourth digit metacarpal length D5MCL fifth digit metacarpal length EL - ear length FA - forearm length PES - pes length TB - tibia length TVL - Tail to ventral length

Skull measurement BL bulla length CW - cranial width DBC - distance between cochleae DL - dental length GBPL greatest basial pit length GSL greatest skull length lOW - interorbital width MW - mastoid width PL - palatal length PPL - post palatal length ZW zygomatic width

Dental measurement C1BW - canine tooth basal width CICIB breadth across both canines outside surface CIM3L canine molar length or maxillary tooth row length M2L second molar tooth crown length M2W - second molar tooth crown width M3M3B - breadth across both third molar teeth outside surface

IV

'. .

List of Tables and Figures

•

Content Page

Table 1 Descriptive statistics for the studied populations. 17

Table 2 Eigenvalues for Discriminant Function Analysis. 20

Table 3 Wilks' Lambda for Discriminant Function Analysis. 20

Table 4 Standardized Canonical Discriminant Function Coefficient. 21

Figure 1 Distribution ofH. ce rvin us. 4

Figure 2 Map of the study sites. 10

Figure 3 Twenty-seven morphological characters measured in Hipposideros. 14

Figure 4 Cluster Analysis ofH. cervinus. 19

Figure 5 Canonical Variate Analysis ofH. cervinus. 22

V

Morphometric Analysis of Fawn RoundleafBat,Hipposideros cervinus (Gould, 1854) (Chiroptera: Hipposideridae) from Several Populations in Malaysian Borneo

Josephine anak Juary

Animal Resource Science and Management Department ofZoology

Faculty of Resource Science and Technology Universiti Malaysia Sarawak

Abstract

The study of morphometric analysis was done on five different populations of Hipposideros cefIJinus in Malaysian Borneo. These populations were from Bako National Park, Kubah National Park, Busuk Cave, Tubau, Niah National Park and Poring Hot Springs. A total of 27 characteristics were measured which included external, cranial and dental measurement. The measurements were recorded and analysed using Cluster Analysis (CA) and Discriminant Function Analysis (DF A) using SPSS Version 18.00. For the DFA, the highest character loading for Function 1 and Function 2 were dental length (DL) and ear length (EL) with standardized canonical discriminant function coefficient value of 0.931 and 0.999 respectively. These two characters are identified as the best predictors to distinguish the five populations based on their morphological measurement.

Key words: H cervinus, morphometric, Cluster Analysis (CA), Discriminant Function Analysis (DFA)

Abstrak

Kajian tentang anal isis morfometrik telah dilalalkan ke atas lima populasi Hipposideros cervinus yang berbeza di Malaysia Borneo. Populasi ini adalah daripada Taman Negara Bako. Taman Negara Kubah. Gua Busuk. Tubau. Taman Negara Niah dan Poring Hot Springs. Sebanyak 27 ciri-ciri telah diukur termasuk moifologi luaran. tengkorak and gigi. Ukuran direkod dan dianalisa menggunakan Cluster Analvsis (CA) dan Discriminat Function Analvsis rDFA) menggunakan SPSS Version 18.00. Dua cM yang paling tinggi diperolehi daripada Fimgst 1 dan Fungsi 2 adalah dental length (DL) dan ear length (EL) dengan nilai standardized canonical discriminant timet ion coefficient masing-masing 0.931 dan 0.999. Kedua-dua ciri ini telah dikenalpasti sebagai prediktor yang terbaik untuk membezakan kelima-lima populasi tersebut berdasarkan ukuran moifologikal.

gata kunci: H cefIJinlls. moifometrik. Cluster Analysis (CAl. Discriminant Function Analysis (DFA)

•

tl r

1.0 Introduction

..

The tropical rainforest of Malaysia is a home for many animal species. In class Mammalia,

the second largest order is Chiroptera which consist of 18 families worldwide (Nowak, 1999)

and 966 species estimated (Altringham, 1996). In Borneo, there are 8 families of Chiroptera

(Payne et aI., 1985) and can be classified into two suborders; Megachiroptera and

Microchiroptera. There are seven families of Microchiroptera namely Emballonuridae,

Megadennatidae, Nycteridae, Rhinolophidae, Hipposideridae, Vespertilionidae and

Molossidae (Payne et al., 1985). Family Pteropodidae is the only order in suborder

Megachiroptera. Chiropteran special feature that is not found in other mammals is their wings

and the ability to fly. The word "chiroptera" which literally means "wing-handed" (Yasuma

and Andau, 1999) is used as general name for all bats.

Most species of bats are nocturnal and they can be found in almost all types of climate because

they are able to adapt to their environmental condition. They also have many different feeding

and foraging strategies including frugivores, nectarivores, carnivores, piscivores,

sanguinivores as well as aeriel and gleaning insectivores (Thomas and Speakman, 2006).

Generally, frugivores and nectarivores are Megachiroptera meanwhile carnivores, piscivores,

sanguinivores and insectivores are Microchiroptera. A minority of Microchiroptera have

evolved to become frugivores and nectarivores (Altringham, 1996).

Microchiropterans are bats weighed between 5g to 20g, except for certain species (Neuweiler,

2000). They are believed to have evolved from an early insectivore because their teeth reflect

this ancestry (Altringham, 1996). Microchiropterans have a highly developed echolocation

system. They have tragus inside their pmna (external ear), which is absent in

2

I

•

Macrochiropterans (Altringham, 1996). The combination of flight and echolocation makes

microhiropteran skillful nocturnal hunters of insects (Neuweiler, 2000).

Among all families of Microchiroptera, Rhinolophidae and Hipposideridae are the high duty

echo locators (Bogdanowicz et aI., 1999). Due to the similarities in some of the morphological

features, some authorities consider Hipposideridae (named as subfamily Hipposiderinae) as a

subfamily of the Rhinolophidae (Wilson and Reeder, 1993; Martin et al.. 2001). Others (Payne

et al. 1985; Flannery, 1995; Eisenberg and Redford, 1999; Nowak, 1999; Yasuma and Andau,

1999) classified Hipposideridae and Rhinolophidae as two different families. Some of the

features that can be used to distinguish Hipposideridae from Rhinolophidae are the form of the

noseleaf, the foot structure, the absence of the lower small premolar, and the structure of the

shoulder and hip girdles (Nowak, 1999).

According to Neuweiler (1999), the hipposiderids can be found in tropical and subtropical

regions of the Old World. H. cervinus is one of the species in family Hipposideridae which is

common in Borneo. By referring to Nowak (1999), H. cervinus has the second highest

distribution in genus Hipposideros after H. diadema. The body colouration of H. cervinus

varies from grey-brown or yellowish-brown to red-brown or orange and the noseleafis greyish

pink with two lateral leaflets meanwhile the ear is broad and triangular (Payne et al. 1985).

They usually roost in caves, sometimes in very large colonies (up to 30000) and they feed in

the forest understory (Payne et al., 1985).

H. CerVill11S plays an important role in ecology and capable of hunting insects in darkness due

to their ability to echolocate. A study conducted by Jalaweh (2004) found that H. cervinlls

3

I

feed on vanous types of insects from orders Orthoptera, Hemiptera, Coleoptera and

Hymenoptera. Some of these insects might be important agricultural pests. Since insectivorous

bats lack the first upper incisor and the first upper and lower premolar, they only have a total

of 3 8 teeth (Neuweiler, 1999). Insectivorous bats are important as biological control for pests,

therefore there are beneficial to human. A bat is able to conswne huge nwnber of insects, as

much as a quarter of its own body weight (Richardson, 1985).

Figure 1 Distribution of H. cervinus (source: IUCN, 2010).

Despite their abundance and wide distribution in Malaysian Borneo, there has been no current

and specific studies conducted on this species in terms of their morphological variations

between different populations. This study aims to assess and review the population variation

of H. cervintls by using their morphometrical characters.

4

Pusal Khidmat Mak'umat Akadem:k, UNIVERSITI MALAYSIA SARAWAK

2.0 Literature Review

2.1 Taxonomy

Family Hipposideridae (roundleaf bats) vary in size from quite small to moderately large

(Payne et al., 1985). This bat is further classified into 9 genera namely Hipposideros, Asellia,

Anthops, Aselliscus, Rhinonycteris, Triaenops, Cloeo tis , Coelops and Paracoelops (Nowak,

1999). Genera Anthops, Rhinonycteris, Cloeotis and Paracoelops only have single species

meanwhile genera Asselia, Aselliscus, Triaenops and Coelops only has two species. One of the

species in genus Coelops can be found in Borneo which is C. robinsoni.

In genus Hipposideros there are 57 species altogether (Nowak, 1999) but only 10 are recorded

in Borneo (Payne et al., 1985). Genus Hipposideros was first described by Hill (1963) and

formerly known as one of the Hipposideros galeritus subspecies which was 11. g. cervinus. 11.

cervinus is similar to 11. galeritus, but differs in possessing a marked antitragal projection on

the ear, and in usually possessing four tail vertebrae (Flannery, 1995). The Bornean species is

further classified into subspecies named 11. c. labuanensis (Payne et al., 1985; Corbet and Hill,

1992).

2.2 The Discovery of Echolocation in Bats

Some species of bats are highly dependent on echolocation, especially insectivorous bats.

Echolocation is the analysis by an animal of the echoes of its own emitted sound waves, by

which it builds a sound-picture of its immediate environment (Altringham, 1996). In the late

18th, Spallanzani demonstrated that bats depend on their ears to finds their way rather than

their eyes (Neuweiler, 2000). He observed that the bats were able to avoid from hitting the bell

I

5

that he hung when they were put in a dark room but not when their ear canal was closed. His

experiments ended without any explanation.

In 1938, Griffin revealed that the world's first ultrasound microphone transformed silent bats

into noisy animals that emit short ultrasound signals, with sounds frequency above 20kHz

using their mouth (Neuweiler, 2000). In 1943, a zoologist, Dijkgrafnoticed the "Tick/aut", an

audible component produced by Myotis emarginatus (Neuweiler, 2000). When he prevented

the bats from producing the sounds, they became disoriented. Observation by Dijkgrag (1943)

gave an explanation to Spallazani's experiment that is, bats release high-frequency sounds

using their mouth or nose and the echoes from the objects around them enable them to

recognize the dark environment (Neuweiler, 2000).

2.3 Distribution ofHipposideros cervinus in Sarawak.

A few studies on diversity of bats or small mammals (Hall, 1996; Hall et al., 2002; Mohd­

Azlan et al., 2005; Wilson et ai., 2006; Anrawali et ai., 2007) in Sarawak have shown that H.

cervinus has a considerable number and widely distributed in the state. Hall et ai. (2002)

recorded 10 individuals from Niah National Park. Mohd-Azlan et al. (2005) conducted their

study at a limestone forest area with 14 individuals ofH. cervinus were recorded. Wilson et al.

(2006) captured 32 individuals in limestone caves and two individuals in a clearing adjacent to

a rainforest patch. In a study conducted by Anrawali et al. (2007), 91 individuals were

collected from various types of habitats including beach, heath, mangrove, mixed dipterocarp

and riverine forest. Another study conducted by Mohd-Azlan et al. (2008) recorded 164

individuals ofH. cervinlls, the highest among 26 species recorded.

6

-

I

2.4 Morphometric Studies

Morphometric analysis is one of the alternatives that have been used to classifY species or

population as well as to detennine the relationship between species or population (Weimin and

Robbins, 2004; Sazali et al., 2008a, 2008b). Morphometric analysis is important because

sometimes misidentification of the specimens in the field occurs due to the overlapping

measurement of the external morphological characters among closely related species (Sazali et

a/., 2006a, 2006b; Wong, 2007, Hasan and Abdullah, 2008).

In addition, morphometric analysis has been used in describing new species (Weimin and

Robbins, 2004; Nicolas et al., 2008). One of the examples of morphometric analysis that has

been widely used is Discriminant Function Analysis (DF A). As mentioned by Biiyiikoztiirk

and Cokluk-Bokeoglu (2008), discriminant analysis is a multivariate statistical method that

provides a model or to forecast the member of a particular group. Another example of

morphometric analysis is Cluster Analysis which classified individuals based on their

similarities.

As a species is separated into different population, they tend to diverge from each other in

order to adapt to their habitat and environmental condition. Morphological divergence can

occur between local populations due to strong differential natural selection (Maryanto et aI.,

2005). Both the internal and external mOlphologies undergo modification to enable species or

J1opulations to adjust to the environment. The shape of certain morphologies has been

associated in detennining an animal's ecology, especially for bats, (Findley, 1993; Birch,

1997). The combination of wing morphology and echolocation is important in the study of

I

7

- -

foraging behaviour of bats (Jacobs, 1999; Fenton and Bogdanowicz, 2002; Stoffberg and

Jacobs, 2004).

Reis et al. (2002) claimed that patterns of variation in morphological characters are essential

for describing boundaries of evolutionary units in nature, besides genetic characters.

Furthermore, morphological variation is a result of genetic and environmental components to

individual development that possibly explain genetic and ecophenotypic variation (Wehausen

and Ramey II, 2000). Guttierrez and Molinari (2008) mentioned that the tight dependence of

member of subgenus Phyllodia on forest may explain the high levels of morphometric and

molecular variability of the subgenus.

Small differences in skull morphology can affect the size ofdietary items consumed whereby

a larger skull offers a wider range of prey sizes in insectivorous bats (Jacobs and Barclay,

2009). Phillips (2000) claimed that dentition and dental are still important to modem bat

systematists. The dental features are very important when considering the food, in terms of

hardness, softness and brittleness (Phillips, 2000), thus affects the selection of food items and

diet preferences.

The morphologically diverse skull of bats affects the bite force (Dumont and Herrel, 2003). A

study conducted by Tingga (2010) on seven populations of Aethalops aequalis in Sabah and

Sarawak showed that these populations can be discriminated by the lower jaw length which is

related to the food sources. Dental characters also has been used to distinguish fossils of

kangaroo rats (Dipodomys), suggesting its reliability in distinguishing the heteromyid taxa and

perhaps elucidate higher level taxonomic relationship (Carrasco, 2000). The study of

8

;;;;--"';~~""-----'-'""'''"'''

I

relationship between morphology and ecological behaviour is termed as ecomorphology

(Findley, 1993). Morphology and ecology are also correlated with the coexistence ofdifferent

species at a single site (see Campbell et ai., 2007).

In insectivorous bats, morphology has a significant effect on the echolocation call besides the

diet. As a result of the negative correlation between size and dominant call frequency, the

bigger the size of drum membranes and string, the lower the frequency produced

(Bogdanowicz et ai., 1999). This might explain why insectivorous bats are relatively smaller

compared to frugivorous bats. Zhao et al. (2003) found that the diet composition of

insectivorous bats is related to the ear length and call frequency because bats can eat more of

their prey iftheir calling frequency is out ofthe range oftympanate prey.

In paleontology, morphological variation and distance are the only indicators of species- and

genus-level taxonomic distinction and clear phylogenetic information normally absent for

distinct species or species cluster (Stafford and Szalay, 2000). Lemos and Cerqueira (2002)

used morphologic and morphometric data to split the white-eared Didelphis into 3 populations

and concluded that the analysed populations merit taxonomic recognition at the species level.

I

9

3.0 Materials and Methods

3.1 Study Sites

Figure 2 Map of the study sites (source: Google Earth, 2010).

3.1.1 Kubah National Park

Kubah National Park is situated about 22km from Kuching City and this park includes three

mountains; Mount Selang, Mount Serapi and Mount Senduk . Gazetted in 1989, this park

covers an area about 22 sq . km (Hazebroek and Abang Kashim, 2000) and small compared to

other national parks in Sarawak. Kubah National Park is consists of mixed dipterocarp forest ,

liveline forest, montane forest and heath forest (Mohd-Azlan et aI., 2008). The liverine forest

is found in the valley of the Rayu River (Hazebroek and Abang Kashim, 2000). Kubah

National Park is located near the sea with latitude 01 °36'N and longitude 1100 09'E. Extensive

10

mangrove forests stretch along the coastline outside the park boundary (Hazebroek and Abang

Kashim, 2000).

3.1.2 Bako National Park

Bako National Park was established in 1957 with an area of about 27 sq. km. It is located 30

km away from Kuching City at latitude 01°42'59.5"N and longitude l10026'39.3"E. This

oldest national park of Sarawak is an isolated island with rocky coastlines (Hazebroek and

Abang Kashim, 2000). Therefore, there is a probability that the population of H. cervinus

inhabited this island is morphologically different from those in the mainland of Sarawak.

According to Hazebroek and Abang Kashim (2000), Bako National Park consists of heath

forest, mangrove forest, mixed dipterocarp forest, riverine forest, beach forest, grassland and

cliff vegetation. In addition, this national park also has seasonal swamp forest during the

monsoon season. Anrawali et al. (2007) recorded eight additional species to this park which

make the total number of insectivorous bat knOVv'Il to occur here was at least 28 species.

3.1.3 Busuk Cave, Tubau

Busuk Cave is situated at Tubau, Bintulu with latitude 03°05'30.27"N, and longitude 113°

42'46.33"E. The area was surrounded by Acacia mangium plantation and also logging camp

during the sampling time (Jayaraj, 2009). Busuk Cave is a sandstone cave inhabited by a high

number of H. cervinus with both orange and brown form exists there. Other bat species that

can be found there are including E. spelaea and also some Rhinolophus sp. deep in the darker

areas ofthe cave. The cave also has a small river inside with some blind cat fishes.

11

3.1.4 Niah National Park

Niah National Park is located in Miri at latitude 03°49'24.3"N and longitude 113°45'42.7"E.

The types of vegetation that can be found in Niah National Park are mixed dipterocarp forest,

riverine forest, limestone forest and secondary forest (Hazebroek and Abang Kashim, 2000).

The limestone contains a number of caves with the largest and the most well known being

Niah Cave (Hall et ai., 2002). The park is surrounded by extensive areas that have been

cleared and planted with oil palm (Hall et al., 2002). Niah's Great Cave is a home to 300,000

bats which roost in the darker parts of the cave. These bats co-exist with the birds in the cave

and they hunt insects outside the cave.

3.1.5 Poring Hot Springs

Poring Hot Springs lies at the Mount Kinabalu Park's southeastern comer and is the reserve's

lowest area at 3000-4000 feet (900-1200m) (Yates and Domico, 1992), about 136 km from

Kota Kinabalu with latitude 06°04'40.84"N and longitude 116°42'32.18"E. The type of

vegetation that can be found in Poring is lowland dipterocarp forests, rich with fruit trees,

lianas and spiny rattan palms (Phillips, 1998) with raffiesia as primary attraction (Yates and

Domico, 1992). One of the main attractions at Poring is the animals such as flying lemur, red

leafmonkey, orangutan, montane birds and numerous butterflies (Phillips, 1998).

12

3.2 Laboratory Method

The specimens used were taken from voucher specimens preserved in 95% ethanol, deposited

in the UNIMAS Zoological Museum. Adult specimens were identified according to Kunz and

Kurta (1988). A total of 50 male specimens were chosen for this study. The skulls extractions

were done by peeling off the skin using a forceps starting from the muzzles. The brain and the

muscle tissue on the skulls were removed until the skulls were clean. The skulls were left to

dry before being measured. The head of the specimens were filled with cotton and stitched

with thread.

3.3 Data Collection

A total of 27 characters (Figure 3); third digit metacarpal length (D3MCL), third digit first

phalanx length (D3PIL), third digit second phalanx length (D3P2L), fourth digit metacarpal

length (D4MCL), fifth digit metacarpal length (D5MCL), ear length (EL), forearm (FA), pes

length (PES), tibia length (TB), tail to ventral length (TVL), bulla length (BL), cranial length

(CW), distance between cochleae (DBC), dental length (DL), greatest basial pit length

(GBPL), greatest skull length (GSL), interorbital width (lOW), mastoid width (MW), palatal

length (PL), post palatal length (PPL), zygomatic width (ZW), canine tooth basal width

(CIBW), breath across both canines outside surface (CICIB), canine molar length or

maxillary tooth row length (CIM3L), second molar tooth crown length (M2L), second molar

tooth crown width (M2W), breadth across both third molar teeth outside surface (M3M3B)

Were measured using digital caliper (calibrated to O.Olmm) following Sazali et at. (2008a,

2008b) and recorded.

13

T

'\ \~ , '

\,_ , _L"'_

\

\0;:------- ,'Sf- -------;11

Figure 3 Twenty-seven morphological characters measured in Hipposideros (drawings not to scale). (adapted from Sazali et at, 2008a, 2008b).

14

3.4 Data Analysis

The data obtained were analysed with Cluster Analysis (CA) using average linkage of

Euclidean distance and Discriminant Function Analysis (DF A) using stepwise procedure, in

Statistical Package for Social Sciences (SPSS) Version IS.OO (SPSS Inc., 2010) to find the

significant characters for discriminating the five populations.

The purpose of using stepwise procedure is to select the variable that has a statistically

significant different across the groups. DF A focus on finding linear combinations of variables

or discriminate variate (such as morphological characters) that best divide populations based

on multivariate observations (Rencher, 1995; Hair et ai., 2006). The group membership is

predicted based on a variate of the independent variables selected for their discriminatory

power in the fOlm oflinear equation (Hair et al., 2006).

where

Zjk =discriminant Z score ofdiscriminant function j for object k a = intercept

Wi = discriminant weight for independent variable i = independent variable i for object kXik

The uniqueness of DF A is that more than one discriminant function may be present (Hair et

qt., 2006). It is used to study the morphometric variation between closely related species and

also species of different population because it is recommended for studying the morphometric

variation within one taxa (see Kitchener et al., 1993; 1995; Sazali et al., 2006a; 200Sa;

200Sb).

15