Download the publication - Tropenbos International

39 downloads 657 Views 3MB Size Report
daerah tersebut adalah hampir sama dengan antara 2.1 dan 2.9 kelompok per km2. Perkiraan ...... sex of the singer, and starting and ending time of all song bouts we heard. We heard ...... monitoring yang tepat terhadap hewan pindahan itu.
Chapter 1. Forest and primates, a general introduction to the conservation of endemic primates in the Sundaic region Introduction The Sundaic region, also known as Sundaland, Malesia, or the Indo-Malayan region, and defined as Thai-Malay peninsula south of the isthmus of Kra, Sumatra, Borneo, Java and Bali, has been identified as one of the hottest biodiversity hotspots on earth (Meyers et al., 2000). The area ranks second in number of endemic plants, fifth in number of endemic vertebrates (excluding fishes), tenth for endemic plants / area ratio and endemic vertebrate / area ratio, and ranks seventh in having the least amount of primary vegetation remaining as percentage of the original extend. With the Philippines and Madagascar, the Sundaic region is the only area to appear in the top ten of all the above listed factors (Meyers et al., 2000). During interpluvial periods the region was originally covered by and large in rain forest of different types, depending on, among other things, altitude and soil type (Park, 1994). Interpluvial periods correspond with the 'glacial' periods in temperate regions and during such times so much of the northern and southern latitudes was covered in ice that sea levels were reduced by as much as 100 m. Along the south eastern edge of the Sunda Shelf, i.e. eastern Java and parts of Bali, drier forest types could be found including deciduous forest. Especially over the last decades, but much earlier in the lowlands of Java, due to rapid deforestation, large areas, are now converted into agricultural land, plantations, and, increasingly 'wastelands', i.e. virtually unoccupied land covered in e.g., alang-alang Imperata grassland and other depauperate vegetation types (e.g., Myers, 1989). The once large continuous areas of forest of the Sundaic region are home to a large variety of non-human primate species (hereafter 'primates'1). Depending on the taxonomy followed 26-28 species can be found in the region. This includes four of the eleven families commonly recognised (Corbet & Hill, 1992; Mittermeier & Konstant, 1996/1997): Loridae (one species), Tarsidae (one species), Cercopithecidae (17-18 species, including the bear macaque Macaca arctoides which occurs marginal in northern part of the Thai-Malay peninsula), Hylobatidae (six species), Hominidae (one or two species depending on whether the two orang-utan taxa from Sumatra and Borneo are given the species rank, see below). From a broad geographic perspective the region is important as it harbours a disproportionately large number of primate species and primate endemics (Mittermeier & Konstant, 1996/1997). With the exception of a few, most notably the long-tailed macaque M. fascicularis, most species are confined to natural forest. Some species are able to survive in certain man-made habitats, most notably certain forest plantations, but most of them will not survive for any length of time in these habitats.

1

Biologically and taxonomically humans are included in the order primates, for practical reasons however, in the present thesis, 'primates' unless specified otherwise normally denotes 'non-human primates'.

In this thesis I will present data on the ecology and conservation of the endemic primates of Java and Borneo. Combined, these two islands, and their smaller neighbouring islands, comprise more than half of the land surface of the Sundaic region. In this first chapter I will briefly introduce the islands of Java and Borneo (geology, climate, vegetation, human population etc.), the endemic primate species occurring on them and their conservation status. The main causes of the decline of the populations of the endemic primates on Java and Borneo are presented, after which a brief overview of primate conservation studies that have been conducted on the islands will be given. The chapter ends with a concise summary of the history of the present study, its aims, and an outline of the thesis. Background on the study islands Java is an island of about 130,000 km2 and politically includes the island of Madura (5,620 km2) which lies just north of the east Javan province. It is administratively divided in six areas, viz. West Java (at the end of 2000 this was split in two smaller provinces Banten and West Java), DKI Jakarta (the nation's capital), Central Java, DI Yogyakarta, and East Java (which includes Madura). As Indonesia’s political and industrial centre, it is one of the most densely populated areas in the World. The very fertile soils which lend themselves to terracing for irrigated rice, sustain about 115 million people, at an average population density of 862 people km-2 (Whitten et al., 1996). Geologically Java is dominated by its backbone of (active) volcanoes, running over the central axis of the island. Eleven volcanoes reach over 3000 m and with 3676 m Mt Semeru is Java's tallest mountain. Rivers are relatively short and run mostly from the central axis of the island straight to either the Java Sea or the Indian Ocean. Java is largely deforested and most of the remaining forest fragments cover (parts of) the numerous volcanoes; human pressure on the remaining forests is very high. Less than 10% of the original forest remains: 54% of the mountain forest, 19% of the hill forest, and only 2% of the lowland forest (Smiet 1992). The last mentioned forest type is now almost exclusively found scattered along the southern coast and in the easternmost part of the island. Once the island was probably completely covered by tropical forest (MacKinnon et al., 1982), but its destruction already commenced in the 15th century (Whitten et al., 1996). An estimated total area of 10 million ha of natural forest was present in the 17th century. Some hundred years ago four million were left, which was reduced to about one million in the first half of the 20th century. During the past 50 years no further significant change in forest cover has taken place (Smiet, 1990). At present deforestation has slowed down, but fragmentation and forest degradation continues (Smiet, 1992). Forest has been replaced by cities and villages, agricultural land, estate crop plantations (coffee Coffea sp., quinine Cinchona calisaya, sugar cane Saccharum sp.), forest plantations (teak Tectona grandis, pine Pinus sp., rubber Hevea brasiliensis), leaving the natural forest areas as habitat islands. Less than 25% of the remaining forest on Java is included in the protected area network (RePPProTT, 1990). The climate on Java differs greatly from the west to the east. The eastern part of Java and the north coast have a pronounced dry season, while in the western half it is weak and nowhere marked. In general, the wettest vegetation types (mixed lowland and hill rain forest and everwet montane forest) only occur in areas with at least 30 rainy days during the driest four consecutive months (van Steenis & SchippersLammertse, 1965), and hence is mostly found in the west and central part of Java.

Rain forest is also found throughout the otherwise seasonally dry east in the wet 'islands' which arise as a result of stowage on the southern and south-eastern slopes of the higher mountains (van Steenis, 1972). In the drier areas rain-forest is replaced by moist forest and deciduous forest. The island of Bali (5560 km2), situated east of Java, is politically a separate entity, but bio-geographically it is included in the Javan faunal province. Bali has a population of some three million people at an average population density of 520 per km2. Most humans live in the fertile southern part of the island (Whitten et al., 1996). Less than one fifth of the island remains under forest cover, most of which is situated in the central mountain range and in the northern part of the island (MacKinnon et al., 1982). Large-scale deforestation on Bali is more recent than on Java, and around the turn of the 20th century most of the northern half of the island was still covered with forest. The loss of forest during the 20th century has been largely due to the introduction of coffee, clove Syzygium aromaticum and coconut Cocos nucifera plantations, and use of fuel-wood (see maps in Whitten et al., 1996: 335). The climate of Bali is comparable to the eastern part of Java, with a long dry season along the north coast and a negligible dry season on some of the higher volcanoes (RePPProT, 1990). Borneo is much larger than Java, in fact with a size of 746,305 km2 it is the third largest island in the world (after Greenland and New Guinea). Administratively it is divided into the four Indonesian provinces of West, Central, East and South Kalimantan, the two autonomous Malaysian states Sabah and Sarawak and the Brunei Sultanate. The Indonesian part of Borneo covers some three quarters of the total land area of Borneo. Borneo mainly consists of relatively low lying areas and over half of the island lies below 150 m a.s.l. In the centre of the island lies a chain of higher hills and mountains, running from south-west to the north-east. Borneo's highest mountain is Mt. Kinabalu in Sabah, which is, with its 4101 m, the highest peak between the Himalayas and the mountains of New Guinea. Other than Java, Borneo is dissected by a large number of great rivers; the Kapuas river (1143 km in length) to the west, the Barito River (900 km) to the south, and the Mahakam River (775 km) to the east. High levels of weathering and leaching are characteristic of many Bornean soils (Burnham, 1984), and the soils on Borneo are generally much less fertile than the volcanic rich soils of Java. Borneo has a much smaller human population than Java (some 12.5 million in 1990: MacKinnon et al., 1996), and an average population density of less than 17 people km-2. All major cities are located near the coast, and population densities in large part of Borneo's interior are less than one person per km2. Within large parts of the island the infrastructure is poorly developed and boats are the main mode of transportation. Settlements are also concentrated along waterways. Borneo supports the largest expanse of lowland evergreen rain forest in the Sundaic region, with some 60% of the land surface still under natural forest (MacKinnon et al., 1996; Collins 1991: this figure may be as low as 45% due to deforestation over the last decades, E. Meijaard, in litt.). The forests are characterised by a high diversity of dipterocarps, the most important timber species in the region (Whitmore, 1984). Timber is a major source of revenue for the Malaysian states and for Kalimantan; oil-rich Brunei has less need to exploit its forest for timber. Large scale exploitation of forest for timber began at the end of the 1960s; in 1967 all Indonesian forests were declared property of the state. Some 90% of all forest

(excluding conservation areas) in Sarawak is under concession (MacKinnon et al., 1996) whereas the total are of forest under concession in Kalimantan is actually larger than the total area of remaining forest (Rijksen & Meijaard, 1999). Besides for timber production, every year vast areas are cleared for agriculture, plantations, human settlements and transmigration. Lowland forests in particular are directly threatened by these practises due to their accessibility and their higher soil fertility than higher altitude forests. The last few decades forest fires have taken an immense toll on the remaining forest areas. During the 1982-1983 fires an area of some 50,000 km2 was affected (Goldammer et al., 1999), and, although figures vary widely, the 1997 fires affected an area significantly larger than this. Less than 10% of the forest on Borneo are formally protected as conservation forest, and most of this is concentrated in the mountains (Sujatnika et al., 1995; MacKinnon et al., 1996). Most parts of Borneo have few months with rainfall less than 100 mm. Most of the hilly inland areas receive between 2000 and 4000 mm per year. West and Central Borneo are the wettest parts of the island, whereas certain parts of east Borneo have a longer dry season with several months receiving less than 100 mm of rain. However, nowhere is the dry season as pronounced as it is in eastern Java or Bali. The endemic primates of Java and Borneo On Java and Borneo a large proportion of the non-human diurnal primates are endemic, viz. three out of four and five out of 13 respectively. Six of these eight species are leaf monkeys (Cercopithecidae, subfamily Colobinae), whereas the other two are gibbons (Hylobatidae) (Oates et al., 1994; Geissmann, 1995) (see Table 1.1). In the other diurnal primate families present on Java and Borneo, viz. the macaques (Cercopithecidae, subfamily Cercopithecinae) and orang-utan (Pongonidae), endemism is absent (although consistent differences between the orang-utan from Borneo and Sumatra in mtDNA [e.g., Karesh et al., 1997], karyotype [Seuanez et al., 1979], habitus [e.g., Markham 1980] and morphology [e.g., MacKinnon 1973] suggest that the two are perhaps best treated as distinct species). Without exception, all of the endemic primates of Java and Borneo are found only in forested areas. For most species this originally consisted of everwet forest only (including riverine, swamp, and montane forest), apart from the ebony leaf monkey which is also found in deciduous forest. A number of species, in particular some of the leaf monkeys, can also be found in forest plantations; but mostly only if more natural forest is present nearby. Since large areas on Java and Borneo are deforested or are under severe threat of being deforested in the near future, and because of increasing human pressure on populations, most if not all endemic species are threatened with extinction. Half of them are included in the IUCN (1996) list of threatened species (see also Chapter 13 for a more detailed assessment of IUCN threat criteria). Two species of leaf monkey, i.e. red leaf monkey Presbytis rubicunda and Bornean leaf monkey P. hosei are listed as Lower Risk (least concern), whereas the Bornean gibbon is listed as Lower Risk (near threatened). White-fronted leaf monkey P. frontata is listed as Data Deficient, i.e. there is inadequate information to make a direct, or indirect, assessment of its risk of extinction based on distribution and / or population status (Table 1.1).

Table 1.1 Endemic primates on Borneo and Java and their IUCN threat status Sub(family) Species Family Cercopithecidae, subfamily Colobinae grizzled leaf monkey Presbytis comata Bornean leaf monkey Presbytis hosei white-fronted leaf monkey Presbytis frontata red leaf monkey Presbytis rubicunda ebony leaf monkey Trachypithecus auratus proboscis monkey Nasalis larvatus Family Hylobatidae Javan gibbon Hylobates moloch Bornean gibbon Hylobates muelleri

IUCN status

Islands

Endangered Lower Risk Data Deficient Lower Risk Vulnerable Vulnerable

Java Borneo Borneo Borneo Java, Bali, Lombok Borneo

Critically Endangered Lower Risk

Java Borneo

Note that fuscous leaf monkey Presbytis fredericae is not included in this listing, as it is considered synonymous with P. comata (chapter 5). Based on Eudey, 1996/1997, Geissmann 1993, Oates & Davies, 1994.

Main causes of decline Over the last few decades the interest in non-human primates in the Sundaic region has increased substantially, and researchers from many disciplines and from various countries have focused their attention on this diverse order. Ironically enough these years have also seen serious declines in primate populations. For an increasing number of species these declines threaten their survival. The major threats to wild populations of primates in the Sundaic region fall into three broad categories: habitat destruction (including total loss and fragmentation), hunting, and capture for local trade. It is important to understand the threats an endangered species is subjected to in order to make recommendations that could positively influence its survival. This is all the more important as the effects of different threats are cumulative. 1. Habitat destruction The main threat to the endemic primates of the Sundaic region is habitat destruction (e.g., MacKinnon, 1987; Eudey, 1996/1997). This includes not only the total loss of forest and its replacement by forest plantations, pasture, or other forms of cultivated land, but also the degradation of the forest. The continuous fragmentation of primary forests and the intensification of land use in the areas between the remaining forest patches result in isolation of the populations trapped in these forest remnants. Many species of primates are completely arboreal and will not cross open land of any width,

which makes them especially susceptible to the effects of forest fragmentation. In the long run, gaps between populations may soon become unbridgeable due to further habitat loss and fragmentation may result in loss of variability due to genetic drift and inbreeding depression. In practise, however, small population will often never reach this stage of the extinction vortex as they are wiped out by hunters, are captured for the pet-trade, or further habitat destruction will results in the death of the last remaining individuals. Thus, in all likelihood, the reduction and the fragmentation of the forests result in populations too small and too widely separated to persist in the long term. Another immediate threat to many of the endemic primates is encroachment by humans along the edges of the forested areas. As the infrastructure is being improved the accessibility is increasing and primate and man are getting more and more in conflict. This usually turns out to be unfavourable for the former species. 2. Hunting The greater accessibility may also increase hunting activities. More on Borneo than on Java and Bali, primates are hunted for food and sport, as crop pest and for medicine. Many rural people depend on wildlife meat for their protein and primates are frequently eaten. Hunting is deeply ingrained into almost all cultures on Borneo (Bennett et al., 1994). With little traditional controls and the almost universal availability of shotguns and cartridges (more so in the Malaysian States than in Kalimantan), the effect on primate populations is devastating. Although largely protected by law, in practise the only safeguard for most species is inaccessibility. With the spread of logging roads, and improved river transport no areas are anymore safe (Bennett et al., 1994). With respect to the success of the conservation of primates, it is relevant to note that human attitudes towards primates differ greatly between religious groups, and that the distribution of religious groups differs between Borneo and Java (and Bali). For Hindus primates are often considered sacred and in certain areas and at certain times offerings are brought to primates (Wheatly, 1999). Primates are not considered sacred by Muslims, but religious restriction permits the consumption of primates. This is in contrast to Christianity where few dietary restrictions prevail. The human population on Java is predominately Muslim and only a small proportion is Christian or of another religion. Concentrations of Hindus on Java are only found in a few remote highland areas. Bali however, is predominately Hindu. The coastal regions of Borneo are mostly inhabited by people with a Malayan origin who have adopted Islam as their main religion. Most of Borneo's interior is inhabited by people of the Dayak and Punan tribes, many of which have been converted to Christianity, although animistic beliefs are still widespread (Cleary & Eaton, 1992). Over the last hundred years there has been a heavy migration from the interior towards the coastal areas, generally as a result of better health facilities, better education and better living conditions (Sirait et al., 1994) bringing people from different cultural and religious backgrounds in closer contact. At a different scale transmigration (in the present context mostly involving people from Java, Sumatra and Sulawesi migrating to (rural) Borneo) has done the same. In general, hunting of primates is rare or absent in most areas on Java and Bali (although in certain areas long-tailed macaques are hunted as crop pest, and sometimes for sport), whereas it is widespread on Borneo, especially in the interior. On Borneo all primates that constitute more than a mouthful of meat (involving all

species with the exception of the nocturnal slow loris and the tarsier) are frequently eaten (Caldecott, 1992). 3. Trade In the Sundaic region, the human attitude towards keeping primates as pets differ greatly from those in the western world and seems to differ little between religious groups. Capturing of primates for pets is widespread throughout western Indonesia, with trade routes mostly running from the 'outer islands' (Sumatra, Borneo, and Sulawesi) to Java, and within Java, from east to west (Nursaid & Astuti, 1996; Nursaid, 1998; R. Nursaid, pers. comm.). The trade in primates is a profitable business and although many species protected by law (both in Indonesia and Malaysia) they are frequently offered for sale at bird markets. Few quantified data are available, but it must be feared that trade in primates has increased, especially after the economic depression of the late 1990's (R. Sözer, pers. comm.; R. Nursaid, pers. comm.). Zoos and safari parks, just as some 'animal lovers', see primates, and preferably the rarer ones, as status symbols and important assets for their collections. It is beyond doubt that zoos and birdparks play an important role in education as well as in conservation of many animal species, but prudence is called for keeping some of the rarest species. Field studies on the endemic primates of Borneo and Java Most long-term primatological studies in Indonesia, and indeed in most of South-east Asia have been concentrated at a limited number of field stations. These stations are mostly situated in relatively pristine habitat, in areas with limited or no hunting pressure, and almost without exception in the lowlands (Table 1.2). At these field stations different aspects of the synecology of tropical rain forests and their inhabitants have been studied, including primates. Borneo seems to follow this pattern, but on Java few long-term studies have been conducted and many primatological studies have been conducted in isolation. From a conservation perspective the data collected at field stations (habitat preferences, densities, group sizes, and other population parameters) are often the only ones available. Examples of studies on the endemic primates of Borneo and Java are listed in Table 1.2.

Table 1.2. Selected study sites on Borneo and Java where long-term studies (> one year or several shorter studies) on endemic primates have been conducted. Country State, Province Malaysia Sarawak

Study site

Altitudinal Study species range

Principal researchers

Samunsam

0-50

E.L. Bennett, A. Sebastian

N. larvatus

Sabah

Danum valley Sepilok Kinabatang --

200-300 50-100 0-100

P. hosei, H. muelleri P. rubicunda N. larvatus --

A.D. (Grieser)-Johns A.G. Davies R. Boonratana --

Gn Palung Tanjung Puting Barito Ulu -Kutai

0-100 0-50 150-350

P. rubicunda N. larvatus, H. muelleri x H. agilis -H. muelleri

West Java

Ujung Kulon Telaga Patengan Pangandaran

0-100 H. moloch 1600-1800 P. comata 0-50 T. auratus

Central Java

Dieng

300-800

M. Leighton, N. Salafski C.P. Yeager D. Chivers, R. Mather -D. Leighton, J.C. Mitani, P.S. Rodman, A Suzuki M. Kappeler, Rinjani Y. Ruhiyat, I.M. Wedana E. Brotisworo, K. Kool, E. Megantara, K. Watanabe V. Nijman

Cepu -Bali Barat --

0-100

East Java Bali Lombok

Brunei Indonesia West Kalimantan Central Kalimantan South Kalimantan East Kalimantan

100-300

0-50

P. comata, H. moloch, T. auratus T. auratus -T. auratus --

Djuwantoko -T. Voght --

Altitudinal ranges are approximate and refer to the main study areas; often higher parts are in the vicinity.

Of the endemics on Borneo the gibbons are relatively well-studied, with long-term studies conducted in Kutai National Park (e.g., Mitani 1984, 1985ab; Leighton, 1987; Rodman, 1978, 1988), Danum valley (Johns, 1992; Grieser-Johns, 1997), and Barito Ulu (Chivers, 1992, including studies on the hybrid zone between H. muelleri and H. agilis: Mather, 1992). Of the colobines, only the proboscis monkey has been studied in detail at a number of localities throughout its range (e.g., Yeager, 1989, 1990, 1991, 1993; Bennett, 1988; Bennett & Sebastian, 1986; Boonrata, 2000). Red leaf monkey has been studied in detail in Sabah (Davies, 1987, 1991; Davies & Baillie, 1988) and to a lesser extent Central Kalimantan (Supriatna et al., 1986). Data on the other colobines is scant and often collected during short studies. Hardly any published studies are available on white-fronted leaf monkey (Blouch, 1997) or Bornean leaf monkey. Of the Javan endemics the Javan gibbon has been studied in detail in Ujung Kulon only (Kappeler, 1981, 1984 abc; Rinaldi, 1999), although some base-line surveys have been conducted in other parts of the island (Kappeler, 1981, 1984c; Asquith, 1995; Asquith et al., 1995; Nijman, 1995). The grizzled leaf monkey has been studied mostly in montane habitats (Ruhiyat, 1983, 1991; Sujatnika, 1992; Wedana, 1993), whereas this species is probably more common at lower elevations (Nijman, 1997b; chapter 6; Whitten et al., 1996). The ebony leaf monkey has a wider niche breadth than the other endemics and occurs in a large variety of forest types. This species has been studied mostly in the more open forest types, including teak plantations (Brotoisworo, 1983; Brotoisworo & Dirgayusa, 1991; Kool, 1993; Kool & Croft, 1992; Djuwantoko, 1991) and deciduous forest (e.g., Kartikasari, 1986; T. Voght, pers. comm.). No long-term studies on this species have been conducted in the rain forest, although it can be found in all wet forest types on Java. Most studies are selectively conducted in those areas where the study species are present at relatively high densities. This may introduce a bias for assessing the conservation status of primates. For example for estimation of total population sizes of primates often rely heavily on published density estimates and hence over-

estimation of true population sizes is likely to occur. Only recently have studies begun to evaluate the effects of habitat disturbance. However most factors of disturbance (selective logging, hunting, and fire) have been studied in isolation (e.g., Suzuki, 1984; Caldecott, 1992; Johns, 1985, 1992). In reality disturbances do not occur in isolation but seem to be tightly linked. As an example, it is not uncommon for a forest area to be selectively logged (during which the rules for proper forest management may or may not be followed), during and after which it is frequently visited by hunters. Parts of the concessions are subsequently invaded by opportunistic farmers and settlers, who take out some additional timber and 'non-timber products', including animals. During a extended dry season the forest is set alight, after which the remaining large trees are felled, and the last animals hunted out. Hence, for primates to persist in under the current management regime, it is necessary to assess the conservation status of these primates based on data collected over a variety of habitats in different stadia of re- and degeneration, facing different threats (human pressure, encroachment, logging, hunting etc.). Aims of the study There is a lack of base line knowledge concerning the ecology of most, if not all, endemic primates in Indonesia. The precise geographical distribution of many species has not been documented. The types of habitats preferred and the densities at which individual species occur in different land use types remains largely unknown. The likely impacts of current factors such as hunting, capturing, habitat alteration and habitat fragmentation are unknown for many of the primates concerned. The islands of Java and Borneo are excellent locations to study the effects of human interference on the survival and conservation of primates. On both islands similar endemic primates are found but the pressures facing these species are quite different. Java represents an area where little forest remains, where the pressure on the remaining populations of primates dates back several centuries, and where many people are no longer dependent on the forest, whereas Borneo represents an area in transition. Although still largely covered in forest, rapid changes in land-use and changing human attitudes will greatly alter the pressures that wildlife populations are facing. The history of deforestation on Java will most likely repeat itself on the other Sundaic islands, e.g., Sumatra and Borneo and possibly other parts of South-east Asia. These areas have a much higher number of primate species. The findings and conclusions of the present study will therefore aim at presenting a framework for the conservation of South-east Asian primates. In order to gain greater understanding in the conservation status of the endemic primates of Java and Borneo, this study set out to collect relevant ecological data and to document the pressures facing the different species. Specific aims of the research are: (i). To assess the geographical distribution of individual species on Java and Borneo (Chapters 6, 7, 9 and 11). (ii). To develop, test and evaluate census methods by which primate populations can be assessed and monitored (Chapters 2, 3 and 4). (iii) To determine the type and magnitude of the threats facing the individual species and habitats on the islands (Chapters 8, 9, 10 and 11).

(iv) Using data collected under (i), (ii) and (iii), to re-assess the conservation status of the endemic primates of Java and Borneo using the IUCN threat criteria (Chapter 12). and subsequently (v). To identify key areas for conservation based on densities of particular primate species, the co-existence of a disproportional large subset of primate species and management feasibility (Chapters 8 and 11). (vi). To discuss the results of the present study into greater perspective and formulate further research priorities (Chapter 13). Setting of the project The Zoological Museum of the University of Amsterdam has several decades of practical experience with conservation related studies in the tropics, including the Sundaic and Wallacean regions. During the decades prior to Indonesia gaining independence most studies were focused on descriptive taxonomy, although part the studies of among others Prof. Dr L.F. de Beaufort, director of the museum during 1922-1949, would nowadays be classified as biodiversity conservation research. Later Prof. Dr. K.H. Voous, curator of birds from 1940-1964, worked on the ornithology of the region and published some influential studies on the birds of Java and Sumatra (Voous, 1950; van Marle & Voous, 1988). In the years after Voous' superannuation the number of projects increased due to the activities of the former heads of the department of birds, Dr J. Wattel, and mammals, Dr P.J.H. van Bree. For vertebrates studies include: seabirds (de Korte, 1989, 1991; de Korte & Silvius, 1994; Argeloo, 1993; Argeloo & Dekker, 1996), megapodes (Dekker, 1990ab; Argeloo, 1992; Jones et al., 1995), pheasants (Sözer, 1997; Sözer et al., 2000, in press; Nijman & Sözer, 1997), birds of prey (van Balen et al., 2000; Sözer et al., 1998), woodpeckers (Lammertink, 1998, 2001), bats (Bergmans & Rozendaal, 1982, 1988; Bergmans & van Bree, 1986), and marine mammals (de Iongh et al., 1997; Kreb, 1999). The present project was initiated in 1996 and stems from an agreement between the Zoological Museum Amsterdam and the Zoological Museum Bogor with ongoing projects concerning woodpeckers, birds of prey, pheasants, primates, dolphins, and sunbears. Outline of this thesis: After the general introduction of primate conservation studies in the Sundaic region, in the following twelve chapters the results of field studies on Javan and Bornean primates which were carried out between 1994 and 2001 will be described, interpreted, and their implications discussed. Section I (Chapter 1-5) presents some background information on survey methodology, behaviour, and morphology. In Chapter 2 a comparison is made between three commonly used methods for estimating densities and biomass of rain forest vertebrates. Chapter 3 deals with the effects that behavioural changes in rain forest vertebrates due to habitat disturbance have on estimation of densities. Both chapters use data from the Bornean gibbon and Javan gibbon, but the results and conclusions are likely to be applicable to other animal groups as well. Chapter 4 deals with the calling behaviour of Javan gibbons. It presents data on the frequency of calling in both sexes, and explores the use of bio-acoustics in conservation studies.

Chapter 5 describes the geographical variation in pelage of grizzled leaf monkeys, one of the three endemic primates of Java. Contrary to previous studies it argues that a number of morphological and behavioural characteristics of individuals of the western populations of this species are not diagnosably different from those in the eastern part of the species' range. It is therefore argued that the grizzled leaf monkey on Java comprises only one species; for conservation purposes the populations on Java should be treated as one single unit. The following two sections deal with the geographical distribution, conservation status and conservation of the endemic primates of Java and Borneo; section II deals with the Javan species, whereas Section III deals with those from Borneo. Section II (Chapter 6-9) begins with addressing the geographical distribution of the two endemic colobines on Java. Chapter 6 deals with the grizzled leaf monkey and chapter 7 deals with the ebony leaf monkey. Chapter 8 presents data on the conservation status and distribution of the endemic primates in the Dieng mountains. Section III begins with presenting data on the distribution and conservation of proboscis monkeys on Borneo (chapter 9), and chapter 10 tells the story of the local extinction of this species from the Pulau Kaget Nature Reserve. Both chapters demonstrate the in-effectiveness of species conservation on Kalimantan. Chapter 11 attempts to assess the patterns of primate diversity on Borneo, and the implications of these patterns for the selection of priority sites for conservation. Section IV starts with a re-assessment of the conservation status of the endemic primates of Java and Borneo based on the present IUCN threat criteria (Chapter 12), incorporating data from the previous three sections. Finally an overall discussion and an integration of the different themes is presented (chapter 13). This chapter also provides some suggestions for further research on primates and their forest in Java and Borneo. Most chapters have been published in scientific journals, with a number of different co-authors. In order to increase readability as much as possible a single style is adopted. Nomenclature and english names for the different species follows Geissmann (1993) for the gibbons, and Oates & Davies (1994) for the colobines (see Table 1.1); these may differ from the ones used in the original publication.

Density and Biomass Estimates of Gibbons in Bornean Rainforest: a Comparison of Techniques

CHAPTER 2 DENSITY AND BIOMASS ESTIMATES OF GIBBONS (HYLOBATES MUELLERI) IN BORNEAN RAINFOREST: A COMPARISON OF TECHNIQUES with Steph B.J. Menken ABSTRACT Censuses were conducted of the Bornean gibbon in two forest areas (Kayan Mentarang National Park [TNKM] and Sungai Wain Protection Forest [HLSW], East Kalimantan, Borneo, in 1996 and 1999-2000, respectively) using three different techniques. Gibbons live in stable and cohesive groups in permanent territories and sing regularly, making it relatively easy to locate them. Firstly, range mapping 3.8 km2 (TNKM) and 5.0 km2 (HLSW) was executed to locate all groups and to map their ranges. Secondly, repeated line transects were run on permanent transects and along ridges. Encounter rates on ridges in both areas were higher than on the permanent transects, reflecting gibbons' preference for higher ground during morning hours. Thirdly, fixed point counts were executed, which make use of the frequent calling of gibbons; the locations from where gibbons were vocalizing were mapped during the early morning (06.00-09.00 hrs) from listening points at summits and ridges. Densities were calculated for two sets of data based on the distance (0.7 and 1.0 km) from where gibbons could be mapped. Overall, density estimates in both areas were relatively similar with between 2.1 and 2.9 groups km-2. The lowest density estimate for TNKM, obtained by the fixed point counts using a radius of 1.0 km (2.1±0.1 groups km-2 or 6.9±2.2 individuals km-2, corresponding with a biomass of 30.7±31.0 kg km-2) was some 30% lower than the highest estimate which was obtained by the line transect technique (2.9±0.2 groups km-2 or 9.9±3.3 individuals km-2, corresponding with 43.2±40.0 kg km-2). The lowest density estimate for HLSW however, obtained by the line transect technique (2.4±0.4 groups km-2 or 7.9±5.8 individuals km-2, corresponding with a biomass of 35.1±89.4 kg km-2), was up to 17% lower than the highest estimate obtained by the fixed point counts technique using a radius of 0.7 km (2.7±0.1 groups km-2 or 9.5±3.5 individuals km-2, corresponding with a biomass of 41.0±31.4 kg km-2). The interaction between site and census technique explained a large proportion of the variation in density, larger than census technique did alone. This data suggests that care must be taken when interpreting density estimates from different areas obtained by different techniques.

13

Forest (and) Primates

RINGKASAN Perkiraan Kepadatan Populasi serta Biomasa Kelawat (Hylobates muelleri) di Hutan Hujan Borneo: Suatu Perbandingan Metoda (bersama Steph B.J. Menken): Sensussensus tentang Kelawat telah dilaksanakan di dua kawasan hutan (Taman Nasional Kayan Mentarang [TNKM] dan Hutan Lindung Sungai Wain [HLSW], Kalimantan Timur, Borneo, masing-masing pada 1996 dan 1999-2000) dengan menggunakan tiga metode yang berbeda. Owa hidup dalam rombongan yang tetap dan kohesif dalam teritori yang permanen dan secara teratur menyanyi, yang mempermudah untuk mencari mereka. Pertama, pemetaan rentang sejauh 3.8 km2 (TNKM) dan 5.0 km2 (HLSW) telah dilakukan untuk menemukan semua kelompok serta untuk memetakan masing-masing daerah jelajah. Kedua, metoda transek garis berulangulang telah dilaksanakan di transek-transek yang permanen dan di sepanjang punggungan bukit. Frekwensi pertemuan di punggungan bukit di kedua daerah tersebut lebih tinggi daripada di transek-transek yang permanen, mencerminkan preferensi Kelawat terhadap dataran yang lebih tinggi selama waktu pagi. Ketiga, metoda titik hitung yang tetap telah dilaksanakan, yang menggunakan frekuensi bersuara dari Kelawat; lokasi-lokasi yang dipakai Kelawat untuk bersuara telah dipetakan selama waktu pagi (jam 06.00-jam 09.00) dari tempat-tempat dengarnya di puncak-puncak dan punggungan-punggungan bukit. Kepadatan populasi telah dikalkulasikan untuk dua set data berdasarkan jarak maksimal (0.7 dan 1.0 km) yang dapat dipakai untuk memetakan Kelawat. Rata-rata, kepadatan populasi di kedua daerah tersebut adalah hampir sama dengan antara 2.1 dan 2.9 kelompok per km2. Perkiraan kepadatan populasi yang paling rendah untuk TNKM, yang didapatkan dari perhitungan titik-titik tetap dengan menggunakan radius 1.0 km (2.1± 0.1 kelompok km-2 atau 6.9± 2.2 individu km-2, setara dengan biomasa 30.7± 31.0 kg km-2) adalah 30% lebih rendah daripada perkiraan yang tertinggi yang didapatkan dengan metode transek garis (2.9± 0.2 kelompok km-2 atau 9.9±3.3 individu km-2, setara dengan 43.2± 40.0 kg km-2). Akan tetapi, perkiraan kepadatan populasi terendah untuk HLSW, yang diperoleh dengan metoda transek garis (2,4±0,4 kelompok km-2 atau 7,9±5,8 individu per km-2, berhubungan dengan biomasa dari 35,1±89,4 kg km-2), yaitu hingga 17% lebih rendah daripada perkiraan kepadatan populasi tertinggi dengan metoda hitung titik tetap dengan radius 0,7 km (2,7±0,1 kelompok km-2 atau 9,5±3,5 individu km-2, setara dengan biomasa 41,0±31,4 kg km-2). Hubungan antara lokasi dan metode sensus menerangkan bagian besar dari variasi kepadatan, daripada metoda sensus itu sendiri. Data ini memberi kesan bahwa kehati-hatian harus diambil pada saat kita menginterpretasikan perkiraan kepadatan dari berbagai daerah yang berbeda yang didapat dari metoda yang berbeda pula.

14

Density and Biomass Estimates of Gibbons in Bornean Rainforest: a Comparison of Techniques

INTRODUCTION Gibbons, or lesser apes (genus Hylobates), are widely distributed throughout the Indo-Chinese and Sundaic regions, from Assam and Bangladesh in the north-west, across southern China and Vietnam, through the Thai-Malay peninsula to Sumatra, Java, and Borneo. Apart from the siamang Symphalangus syndactylus, which occurs sympatrically with white-handed gibbon H. lar and agile gibbon H. agilis, in parts of their range, gibbons are largely parapatric with varying degrees of hybridization and reproductive isolation (Gittens & Raemaekers, 1980; Geissmann, 1995). On Borneo two species occur: the agile gibbon and the Bornean gibbon, H. muelleri (see Figure 2.1). The agile gibbon occurs in the southernmost part of Thailand, Peninsular Malaysia, central and south Sumatra, and in the south-western corner of Borneo, between the Kapuas and Barito rivers. The Bornean gibbon is endemic to Borneo and can be found in the remaining part of the island. In the contact zone at the headwaters of the Barito river, an apparently stable, hybrid population occurs (Mather, 1992).

Figure 2.1 Bornean gibbon Hylobates muelleri in Sungai Wain protection forest (photo G.M. Frederiksson).

15

Forest (and) Primates

Bornean gibbons are completely arboreal and are confined to closed canopy forest in the lowlands and hills up to c. 1500 m a.s.l. (Payne et al., 1985; V. Nijman & E. Meijaard, unpubl. data). The species is largely frugivorous, eating mostly ripe, sugar-rich, juicy fruits (Rodman, 1978, 1988; Leighton, 1987). Bornean gibbons are territorial and live in monogamous family groups consisting typically of an adult pair and up to four offspring. Pairs perform elaborate duet songs that are thought to form and maintain the pair bond and to establish and maintain the territory (Raemaekers & Raemaekers, 1985; Mitani 1984; Leighton, 1987). The population may furthermore contain a number of floating (sub-adult) males and females, who call rarely. Bornean gibbon is classified as Lower Risk (near threatened) according to IUCN threat criteria (Eudey 1996/1997), and the species is protected throughout its range. Although the total area of forest on Borneo is still large compared to other parts of the distribution range of gibbon, every year vast areas are cleared for timber production, transmigration or agriculture and are increasingly lost due to arson (e.g., Rijksen & Meijaard, 1999). Especially lowland forests are directly threatened by these practices due to their accessibility and higher soil fertility than higher altitude forests. Throughout Borneo, accessibility has greatly increased over the last few decades: out-board motors and logging roads make all but the most remote areas accessible for exploitation. Only a small proportion of the lowland forest on Borneo is formally protected as conservation forest (MacKinnon et al., 1996), but especially in the Indonesian part of the island, many conservation areas are protected on paper only (Rijksen & Meijaard, 1999; Meijaard & Nijman, 2000). Numbers of Bornean gibbons are declining overall because of habitat disturbance or habitat alteration, and populations in some areas have been greatly reduced or even eliminated by hunting. Local gibbon populations are easily exterminated by hunting because of the species' loud songs, which attract attention, monogamy, which easily disrupts breeding, and strong sedentary behavior which renders both evasion of hunters and rapid recolonization of depopulated areas more difficult (Bennett et al., 1987a). Range mapping of all known primate groups in a given area is generally considered to provide the most accurate approximation of true density for rain forest primates (see e.g., NRC, 1981; Skorupa, 1987). However, forest primates are difficult to census accurately. Because of the three-dimensional structure of the South-east Asian evergreen closed canopy forests, where trees can reach heights of 60 m and more, arboreal primates are difficult to locate. Range mapping is, furthermore, time consuming and labor intensive, and only applicable in relative small accessible areas. The technique is furthermore not suitable for rugged or mountainous terrain, where access can only be obtained via ridges. Although range mapping is considered more accurate, the most commonly used technique of estimating primate population density is the (repeated) line transect (NRC, 1981; Whitesides et al., 1988; Buckland et al., 1993). This technique depends on the detection of animals (or sometimes merely animal signs such as nests) on one or both sides of a survey path. Ideally, transects should be placed randomly or through a stratified random technique and should follow a straight line. 16

Density and Biomass Estimates of Gibbons in Bornean Rainforest: a Comparison of Techniques

In practice, however, transects often follow geographic features as crests, ridges and spurs (Blouch, 1997) or logging roads (Johns, 1985; Grieser-Johns, 1997). The technique has been employed for survey work (Davies & Payne, 1982; Bennett & Dahaban, 1995; Nijman & van Balen, 1998), comparative studies (Glanz, 1982; Johns & Skorupa, 1987; Yanuar et al., 1995; Johnson & Overdorff, 1999), and for estimation of population parameters in areas where other methods (mark-recapture, complete counts, home range or territory mapping) are not feasible (Green, 1978). Although line transects have been widely used for censusing gibbons (Marsh & Wilson, 1981; Davies & Payne, 1982; Johns, 1985; Haimoff et al., 1986; Bennett & Dahaban, 1995; Blouch 1997), Brockelman & Srikosamatara (1993) considered the technique not particularly suitable for this taxon. Gibbons live in small family groups which makes them difficult to detect and they can behave unpredictably (e.g., flee, hide, approach) when detecting humans. Brockelman & Ali (1987) and Brockelman & Srikosamatara (1993) discussed the possibilities of estimating gibbon densities by fixed point counts making use of gibbon’s great calls. Fixed point counts have the advantage of allowing density estimation over relative large areas in a short time span. A number of studies have been conducted with the specific aim of comparing different census techniques (Green, 1978; Whitesides et al., 1988; Defler & Pintor 1985; Mitani et al., 2000; Fashings & Cords, 2000; Brugiere & Fleury 2000), but few studies were conducted in South-east Asian forest and none involving gibbons have been published. Here we compare three different techniques, namely range mapping, repeat line transects, and fixed point counts, to estimate densities of Bornean gibbon at two lowland primary forest sites. METHODS Study Area Data were collected in the Kayan Mentarang National Park (Taman Nasional Kayan Mentarang [TNKM]) in Oct.-Dec. 1996 and Sungai Wain protection forest (Hutan Lindung Sungai Wain [HLSW]) in Dec. 1999-Feb. 2000. Both field studies lasted ten weeks. Figure 2.2 depicts the location of the study areas. Gazetted in 1980 as a strict nature reserve (cagar alam), Kayan Mentarang became a national park in 1997. With adjacent (proposed) reserves, TNKM totals some 20,000 km2. The study site proper was the Nggeng Bio river valley, in the surroundings of the Lalut Birai field station. The Nggeng Bio river is a tributary of the Bahau river which in this part of TNKM marks the eastern boundary of the reserve. The study area consists of rather steep hills intersected by many small streams, with the study conducted between c. 350 and 750 m a.s.l. The natural vegetation type in the area is lowland dipterocarp rain forest. Despite the existence of many rivers, true riverine terrace forest is rare or even absent. In many areas the riverbanks consist of steep, rocky gorges, with riverine forest only present in the more gently sloping areas. The Nggeng Bio river valley has been a restricted forest 17

Forest (and) Primates

(tana ulen) of the nearby village of Long Alango for at least the last 75 years. Since hunting, fishing, cultivation, and collection of forest products have been mostly prohibited, the valley is still covered with mature, tall primary forest. Besides gibbons, other primates regularly recorded in the study area are Bornean leaf monkey Presbytis hosei, long-tailed macaque Macaca fascicularis, and pig-tailed macaque M. nemestrina (Nijman, 1997).

Figure 2.2

Location of Kayan Mentarang National Park and Sungai Wain protection forest on the island of Borneo. The squares show the locations of the study sites.

Part of the Sungai Wain reserve was gazetted as a closed forest (hutan tutupan) in 1934 by the Sultan of Kutai. Since 1947, the forest received protection as a water catchment area for the oil industry in the Balikpapan region, and in 1983 it was gazetted as a protection forest (hutan lindung). As a research site for the Ministry of Forestry-Tropenbos program, HLSW became an orang-utan Pongo pygmeus, introduction forest in 1992. In HLSW, the study site proper was in the Bugis river valley, in the surroundings of the Camp Djamaludin field station. The area consists of undulating terrain with the study conducted between c. 50 and 127 m a.s.l. HLSW covers a variety of forest types, including fresh water swamp, riverine forest, moist lowland dipterocarp forest, and dry hill dipterocarp forest. In the east it is bordered by some unprotected mangrove forest. The study was conducted in the moist 18

Density and Biomass Estimates of Gibbons in Bornean Rainforest: a Comparison of Techniques

lowland dipterocarp forest and dry hill dipterocarp forest, with small parts in riverine forest. At present HLSW is the last remaining area covered with mature undisturbed primary rain forest in the south-eastern coastal region of East Kalimantan. In 1998, forest fires affected some half of the 100 km2 large reserve, but the central core remained untouched (Frederiksson & de Kam, 1999). Other primates regularly recorded in the study area are red leaf monkey Presbytis rubicunda, white-fronted leaf monkey Presbytis frontata, and pig-tailed macaque M. nemestrina. At the time of the study no zoological research had been conducted in TNKM other than a few short reconnaissance surveys (Yeager, 1991; Foead, 1995; van Balen, 1997) and no animals were habituated. In HLSW, some 20 introduced orangutans were present (G.M. Frederiksson, pers. comm.), one of which was occasionally seen in the study area. No primatological studies have been conducted in the study area and no animals were habituated. In summary, both study sites are covered in tall mature forest, have been effectively protected since the beginning of the last century, no primatological studies have been conducted to date, and none of the primate species were habituated. This makes these areas one of the few in Borneo suitable for the study of primates in undisturbed forests. DATA ACQUISITION AND DATA ANALYSIS Range Mapping In TNKM a number of non-overlapping permanent transects were present; these transects were originally laid out to monitor the phenology of the forest. In HLSW a network of transects is present running north-south and east-west, and spaced some 500 m apart. In both areas, apart from the permanent transects a number of small trails allows access to much of the entire area. All records of primates were plotted on a map. In both study sites, apart from the first author, a number of researchers or field assistants were present, collecting additional data. In this way we collected data on the precise locations and group sizes of gibbons within an area of 3.8 km2 (TNKM; excluding the field station itself and its direct surroundings) and 5.0 km2 (HLSW), disregarding additional area due to slopes. Groups that were occasionally seen, but had more than an estimated three-fourths of their range outside the sample area, were omitted. Density estimates were obtained by dividing the total number of groups or the total number of individuals found by the census areas. Line Transects At both study sites, three transects were selected, which were between two and three kilometers in length. These transects were walked by V. Nijman; in TNKM occasionally a second observer was included. Data were collected in both directions, but always after a stop of at least 45 min, and always during periods of good weather. Since gibbons become less active in the afternoon (Leighton, 1987; V. Nijman, pers. observ.) only censuses that were completed prior to noon were included for analysis. An average walking speed of c. 1.5 km h-1 was maintained. A 19

Forest (and) Primates

total of 142.5 km (TNKM) and 172.8 km (HLSW) were thus covered before noon. Densities of gibbons were estimated using the effective distance method of Whitesides et al., (1988). The density of groups km-2 is given by: D = n · A-1 = n · (L · 2 · (Ed + 1/2 S))-1

(equation 2.1)

where D = density (groups km-2), n = number of groups seen, A = census area (in km2), L = length censused (km), Ed = Effective distance (km, estimated in m), and S = mean group spread (km, estimated in m). The effective distance is defined as the distance on each side of the transect at which the number of sightings at greater perpendicular distances equals the number ‘missed’ at nearer distances (Whitesides et al., 1988). It is determined by using a histogram of perpendicular distances and calculated as (Nt / Nf) · Fd, where Nt = total number of sightings, Nf = the number of sightings below the fall-off distance, and Fd = fall-off distance, defined as the maximum reliable distance beyond which the number of sightings is reduced by 50% or more (Brugiere & Fleury, 2000). The group spread is the diameter of the circle of equivalent area to that occupied, on average, by a group of primates. Group spread proved difficult to estimate and in order not to violate the underlying assumptions of the line transect method (such as deviating from the transect line, remaining at one position for a longer period of time and walking backwards to obtain a better view: Buckland et al., 1993; Krebs 1989), it was considered best to obtain group spread estimates from the range mapping technique. The variance among the means of the three transects was used as a measure of error in order to estimate confidence limits. When the density is known in groups km-2, the density of individuals is then calculated using the mean group size, as observed along the transect lines. For estimation of mean group size, only 'complete counts' (counts that were made when there was confidence that all members of the group had been actually observed) were included. Standard errors of the mean (s.e.) for individual densities were calculated following Raj (1968): s.e.(ID) = GD2 · (s.e.GS) + GS2 · (s.e.GD) + (s.e.GS) · (s.e.GD).

(equation 2.2)

Where ID = individual density (individuals km-2), GD = group density (groups km-2) and GS = mean group size. Brugiere & Fleury (2000) expressed the need to explore the influence of topography on bias in density estimation when using line transects. Therefore, in order to test for possible biases of censusing along ridges, spurs, and crests only, data from each transect line was converted into an encounter rate (average number of groups encountered within a band of 50 m per km surveyed between 06.00 and 12.00 hrs), and compared with encounter rate for one (HLSW) and two (TNKM) trails following main ridges.

20

Density and Biomass Estimates of Gibbons in Bornean Rainforest: a Comparison of Techniques

Fixed Point Counts The density of gibbons was calculated by mapping localities from where gibbons were vocalizing from four (TNKM) and three (HLSW) listening posts mostly positioned at summits or ridges. These listening posts had a hearing angle of at least 240°. The bearing and distance of gibbons singing between 06.00 and 09.00 hrs were plotted on a map (scale 1:25,000 [TNKM] and 1:17,000 [HLSW]). This time window coincided with the time of greatest singing activity in Bornean gibbon in Kutai National Park as reported by Mitani (1984). Mated female Bornean gibbons usually confine their singing behavior to duet song bouts only. The most prominent song contributions of female gibbons consist of a loud stereotyped phrase, the ‘great call’. This great call includes an acceleration type climax with fast bubbling notes. Great calls may be audible further than two kilometers (V. Nijman, pers. observ.). Adult males do not produce great calls, but often produce solo song bouts. During duetting the male and the female combine their song contributions to produce complex but relatively stereotyped vocal interactions (Geissmann, 1995). For surveying, only songs that included a great call were used and it is assumed that this indicates a family group (cf. Leighton, 1987). During calling, gibbons move only limited distances. Songs that map more than c. 500 m apart are considered to be given by separate groups (cf. Brockelman & Ali, 1987). Different groups calling simultaneously are distinguishable if one has knowledge of song organization; those singing at different times can be distinguished with a combination of directional, distance information and occasionally by individual characteristics. Since weather has been found to affect singing frequency in most if not all species of gibbon studied (e.g., Brockelman & Ali, 1987; Brockelman & Srikosamatara, 1993; this study), and since estimating distances is more difficult when it is raining or windy (V. Nijman, pers. observ.), censuses were only conducted during periods of suitable i.e., still and dry, weather. In a given population of gibbons on a given day there are also non-calling groups. The proportion of groups calling on an individual day (p) varies between gibbon species and between populations within gibbon species (Brockelman & Ali, 1987). For three (TNKM) and four (HLSW) groups, situated nearest to the field stations, the proportion of them calling between 06.00 and 09.00 hrs (p6-9) was estimated by remaining within hearing distance of a focal group for a period between five and 14 days. Only song bouts that included great calls produced by the female were included. The fixed point count technique requires knowledge of song organization and is probably suitable for experienced observers only (cf. Brockelman & Srikosamatara, 1993). Therefore, in TNKM, prior to the fixed point counts, a three week training period allowed improvement of distance estimation. In both study sites, during the line transect surveys and during non-census walks, vocal bouts were noted, and their distance and bearing were estimated at different times. Using triangulation from different locations along the transect and with the aid of

21

Forest (and) Primates

topographic maps it was possible to check the distance estimates and subsequently to improve skills. The census area was obtained by plotting the locations of vocal bouts on a map. From this it was concluded that, taking into account the topography of the area, gibbon song bouts could accurately be recorded within a radius (r) of both 0.7 and 1.0 km. A radius of 1.0 km covers and area twice that of 0.7 km. The density of gibbons was calculated by D = n · p6-9-1 · A-1 = n · p6-9-1 · (φ · π r2)-1

(equation 2.3)

where D = density (mated pairs km-2), n = the average number of groups heard calling from the listening posts on a given day, p6-9 = proportion of groups calling between 06.00 and 09.00 hrs, A = the census area, φ = the proportion of a circle from where gibbons could be heard (between 240° and 360°), and r = radius from where gibbons could be mapped (either 0.7 or 1.0 km). The variance among the means from the listening stations allowed an estimate of confidence limits. Fixed point counts sensu stricto do not provide information on group sizes, and hence no estimates on the density in number of individuals can be made using this technique. Data on group sizes were obtained by combining all accurate counts of gibbon groups at other times during the study in the wide surroundings of the listening points. Standard errors of the mean for individual densities were calculated following equation (2.2). Biomass Estimates Gibbon biomass was calculated based on group densities for each of the above census techniques. Geissmann (1993) tabulated weights of wild-shot animals of different gibbon taxa on Borneo. Weights of animals from the north-eastern part of Borneo (H. m. funereus) do not differ significantly from those in the south-eastern part (H.m. muelleri) (t-test, n.s.), and data are pooled in order to provide an average weight of Bornean gibbons in east Borneo. An adult female, on average, weighs 5.25 kg (s.e.=0.18, n=16) and an adult male 5.57 kg (s.e.=0.17, n=12); weights for two sub-adult females were 3.29 kg and 4.20 kg, respectively (Geissmann 1993: 347). Neonate weights of Bornean gibbons are some 0.40 kg (Geissmann & Orgeldinger, 1995). Few data are available on weights of the ‘average’ immature, but for calculation of biomass this was, arbitrarily, taken as halfway between birth weight and mature weight, viz., 2.91 kg (cf. T. Geissmann, in litt.). Groups always contained an adult male and an adult female (with a combined weight set at 10.82 kg), and a varying number of immatures (between none to four). Since each group contains an adult pair variation in group sizes reflects the variation in number of immatures in a group, and standard errors of group mass were calculated using the variation in number of immatures multiplied by the average weight of an immature. Standard errors of total biomass were then calculated using standard errors of both group mass and group density following equation (2.2). 22

Density and Biomass Estimates of Gibbons in Bornean Rainforest: a Comparison of Techniques

Statistical analysis were performed using Minitap 8.2, and throughout means are reported plus and minus one standard error of the mean (± s.e.). Significance was assumed when p0.05). In all three study areas, and for both species, the response was in the same direction, i.e. gibbons tended to behave in such a way as to reduce the likelihood of being detected. Increase in freezing, hiding and silently moving away, makes it more difficult to locate or detect groups of primates and will lead to a decrease of groups 37

Forest (and) Primates

detected. It is furthermore likely that group sizes will be under-estimated as it becomes more difficult to detect all individuals in a group. 2. Change in activity patterns Primates are most often detected when engaged in conspicuous activities such as vocalising, travelling or feeding, either due to visual or auditory cues of the animals themselves or their surroundings (moving of branches, falling fruit etc.). They are less easily detected when resting. Since time spent travelling and feeding is lowered in disturbed forests, this means that fewer groups will be detected in disturbed habitats. This will also include groups at the transect line, violating one of the critical assumptions of the method. Considered in isolation, the observed change in activity patterns will lead to an under-estimation of true densities in disturbed habitats.

Figure 3.1 1. 2.

Percentage of canopy use by two gibbon species in disturbed and undisturbed habitats

undisturbed situation consists of hill forest in the Sungai Tekam, West Malaysia, whereas the disturbed situation consists of the same area after >50% of the trees were lost due to selective logging (after Johns, 1986) see footnote 3, Table 3.1.

3. Use of canopy levels Gibbons prefer tall trees for certain activities. Emergent trees and the upper canopy are disproportionately used (favored) for singing and travelling (cf. Kappeler, 1984; Johns, 1986). In disturbed forests, due to the loss of many large trees, generally activities have shifted from the upper to the middle canopy (Figure 3.1: H. lar: χ2= 89.4, df=3, p1 650 >4 925 1085 1150 >2 820 7 775 >3 750-800 >4 750-800 7 (+1 neonate) 850 3 765 15 (+4 neonates) 780 10-12 (+1 neonate) 815 1300 1 865 8-10 800 >5 570 600 500 1

25/09/’94 11-14, 16/06/’95 12/06/’95 13/06/’95 14/06/’95 14-15/06/’95 29/06/’95 P. comata

05/07/’94 11/06/’95 13/06/’95

2 1 3-4 1 1 2 5-6 1 1 1

14/06/’95 1 1 15/06/’95 T. auratus

25/09/’94 12/06/’95 14/06/’95 15/06/’95

1

16/06/’95 1 1 1. 2.

Altitude

Locality Linggo Linggo Linggo Linggo (centr.) 1 Linggo Linggo Linggo (centr.) Linggo (centr.) Linggo (centr.) Mt. Lumping Mt. Rogojembangan Kroyakan Mt. Prahu Mt. Prahu2 Linggo Linggo Linggo Linggo Linggo (centr.) Mt. Lumping Mt. Lumping Linggo Linggo Linggo Linggo Linggo (centr.) Linggo Linggo Linggo Mt. Lumping Linggo (centr.) Linggo (centr.) Linggo (north) 1 Linggo (north) Kroyakan

Linggo (centr.) refers to the area between Linggo and Mt. Lumping, Linggo (north) refers to the lowland forests north of Linggo. See figure 8.1. unconfirmed, see text.

More direct observations were made in the western part of the study area. Groups of 2-13 individuals were seen on seven occasions. The smaller group sizes were probably a reflection of the difficulties in observing the complete group rather than 120

A Faunal Survey of the Dieng Mountains, Central Java

the size of the group itself. On three occasions we had the impression that we could observe the whole troop, with group sizes of at least seven individuals, nine plus one neonate, and eleven plus two neonates. Two groups were detected by means of their characteristic vocalisations. In total at least six different groups inhabited the study area. On six occasions we observed a group of grizzled leaf monkeys within 50 m from the transect line. Assuming an average group size of seven individuals (cf. Supriatna et al., 1994) the population density was estimated at 28 individuals km-2. We observed the species between altitudes of 650 and 2565 m. Local inhabitants and officers of the forestry department reported it to be present north of Linggo (forests descends to c. 300 m) and also in the forests above Kroyakan (altitude c. 400 m). Grizzled leaf monkeys were recorded in primary forest and in secondary forest, both in edges and the interior. They were found in lowland forests, in forests on steep slopes and on hills and in upper montane forests. The ebony leaf monkey was the species observed most frequently. It was reported to be rather common on both on Mt. Prahu and the central part of the study area. In the western part we sighted groups on nine occasions and twice located a a group or individual by their characteristic vocalisation. In the forests above Kroyakan a group was also identified by means of vocalisation. Records come from the north-western part north of Linggo, to Mt. Lumping in the south and Kroyakan in the north-east, between altitudes of 500 and 1300 m (Table 8.1). At least eight groups of 3-19 individuals of ebony leaf monkey are present in the study area. Along the transect line groups were observed five times. Assuming an average group size of seven individuals (V. Nijman, unp. data) the density may be 23 individuals km-2. The habitat in which ebony leaf monkeys were observed included primary and secondary forest, both on the edges and in the interior. Population estimates Although preliminary, we present some information of the number of individuals of the two most endangered primate species --the Javan gibbon and the grizzled leaf monkey-- possibly present within the area. The Javan gibbon was observed between altitudes between c. 300 and 1300 m. It is likely to be distributed throughout the western half of the study area up to altitudes of c. 1600 m. This is considered to be the species' upper limit (Kappeler, 1984a) although it has been reported from higher altitudes (e.g., 2400 m on Mt. Pangrango: Doctors van Leeuwen, 1926; 1900 m on Mt. Tangkuban Perahu: R. Sözer, pers. comm. 1992). The most widely used method for estimating population sizes of Javan gibbons is by extrapolation based on the forest area inhabited by gibbons and their population density (e.g., Kappeler, 1984a, Kool, 1992; Supriatna et al., 1994; Asquith et al., 1996). Population densities vary with altitude (Kappeler, 1984a, Supriatna et al., 1994) and we followed the assumptions of Supriatna et al. (1994) which are comparable with those of Kappeler (1984a) and Asquith et al. (1996). The assumptions of Supriatna et al. (1994) are based on data derived from studies of the white handed (H. lar) and pileated gibbon (H. pileatus). They considered the edge effect (defined as the size of habitat on the 121

Forest (and) Primates

forest periphery not occupied by gibbons) for Javan gibbons to be one km. Thus, suitable habitat for gibbons was calculated by subtracting the first kilometre of habitat from the available habitat. This area was multiplied by the gibbon population density for that altitudinal vegetation zone (Table 8.2). A population of over 500 individuals may be present in the area. Less is known about densities and altitudinal distribution of grizzled leaf monkeys than for the Javan gibbon. Reported densities vary from 4-5 individuals km-2 in Halimun National Park (Maitar in Supriatna et al., 1994) to 35 individuals km-2 in Patenggang (Ruhiyat, 1983). Our density estimate of 28 individuals km-2 is similar to the 25 individuals km-2 found by Sujatnika (1992; Sujatnika, pers. comm.) in Mts. Gede Pangrango National Park. According to Supriatna et al. (1994) grizzled leaf monkey do not inhabit the core of a forested area. Although tentative, if we, for reason of comparison only, follow the assumptions of Supriatna et al. (1994), i.e. grizzled leaf monkeys range in a band of 2 km around the perimeter of the area with a density of five individuals km-2, the available habitat totals c. 140-155 km2 and the total number of grizzled leaf monkeys present in the Dieng mountains may be c. 700-800 animals. Table 8.2

Estimates of the numbers of Javan gibbon present at the different vegetation zones on the Dieng mountains and total number of gibbons present in the area, calculated following the assumptions of Supriatna et al. (1994).

Altitudinal vegetation zone (m)

Size↑ (in km2)

Edge (in km2)

Core (in km2)

Density↑↑ (ind. km-2)

Estimated number

Lowland (0-500m) Hill (500-1000m) Lower montane (1000-1500m) Total

13-15 90-100 17-20

6-7 20-23 6-8

7-8 70-77 11-12

1-3 7 2

7-24 490-539 22-24 519-577



Size of natural forest area after RePPProT (1990), scale 1: 250,000: edge is defined as the first kilometre of habitat on the forest periphery not occupied by gibbons, while core is calculated by subtraction of the area of the edge from the forest size (after Supriatna et al., 1994). ↑↑ Densities after Supriatna et al. (1994).

DISCUSSION Even though we realise that some of our data are rather limited and based on few observations, we think that they allow comparison with other studies. Supriatna et al. (1994) concluded that there was an inconsistent and incomplete dataset available to estimate the sizes of wild gibbon and leaf monkey populations. In our attempts to estimate the number of Javan gibbon and grizzled leaf monkey present in the Dieng mountains, we largely followed the assumptions of Supriatna et al. (1994). Some of these assumptions however, are considered not to be valid in our study area, e.g., gibbons were heard frequently less than 100 m from the main road from Kayen to 122

A Faunal Survey of the Dieng Mountains, Central Java

Peninggaran, and they were reported to be present in the lowland forest south-west of the road from Kroyakan to Doro, questioning the assumption that gibbons do not inhabit the first kilometre of the forest periphery. If we use the assumptions and densities estimated by Kappeler (1984a) and Asquith et al. (1995) to estimate the number of gibbons in the Dieng mountains, the results are similar to our own: 249955 and 523-577, respectively (see Table 8.3). Table 8.3

Estimates of the numbers of Javan gibbon present in the different vegetation zones on the Dieng mountains using the assumptions of Kappeler (1984) and Asquith et al. (1995).

Altitudinal vegetation zone (m) After Kappeler (1984) Lowland (0-500m) Hill (500-1000m) Lower montane (1000-1500m) Total After Asquith et al. (1984) Lowland (0-1000m) Montane (1000-2000m; western part only) Total ↑ ↑↑

Size↑ (in km2)

Density↑↑ (ind. km-2)

Estimated number

13-15 90-100 17-20 120-135

1-13 2-7 1-3

52-195 180-700 17-60 249-955

77-85 11-12 88-97

6.5 2

501-553 22-24 523-577

Size of natural forest area after RePPProT (1990), scale 1: 250,000. Note that Asquith et al. (1994) took the edge effect into acount but that Kappeler (1984) did not.

Furthermore, we do not agree with the assumption of Supriatna et al. (1994) that grizzled leaf monkeys are not found in the core of a forested area, and a density of 5 individuals km-2 seems rather conservative because our own estimate and those reported from other studies (e.g., Ruhiyat, 1983; Sujatnika, 1992) are considerably higher (see chapter 12). However, pending further research on the population density and habitat preferences of grizzled leaf monkeys, any further adjustment seems unwarranted. Despite the number of limitations and perhaps low level of accuracy we have presented our population estimates as it is believed that (i) even these rough estimates are useful in order to get some idea on population sizes and (ii) the estimates allow comparison with other areas and other studies which have been conducted so far. The only area, other than the Dieng mountains, where both grizzled leaf monkey and Javan gibbon are present in greater or similar numbers is Halimun National Park in West Java (360 km2 of forest between 500 and 1929 m, of which 80 per cent is above 1200 m; Kool, 1992). Most patches of strict lowland forest in Halimun National Park are discontinuous and are often situated outside the reserve boundaries (Whitten et al., 1996), so the area's protected status might create the illusion that populations of Javan gibbons and grizzled leaf monkeys are relatively secure, this is not necessarily the case. Population estimates of Javan gibbon for this 123

Forest (and) Primates

area range from 852-1320 animals (Kool, 1992), 870 animals (Asquith et al., 1995) to 908 animals (Supriatna et al., 1994). Up to 720 grizzled leaf monkeys may be present (Supriatna et al., 1994), although Maitar (in Supriatna et al., 1994) reports very low densities of this species in the area and Kool (1992) was not able to obtain a density estimate because of the low number of sightings. During our surveys in the Dieng mountains we were repeatedly confronted with the fact that no one had recognised the high biological value of this area before. Therefore, and as our surveys were limited in time and did not cover the area as a whole, we highly recommend the area for those interested in studying Javan fauna and flora in more detail. In our survey for instance we observed or found traces of species like large flying fox Pteropus vampyrus, binturong Arctictus binturong, leopard Panthera pardus, pigs Sus scrofa and/or S. verrucosus, barking deer Muntiacus muntjak, and black giant squirrel Ratufa bicolor. The is still much to be discovered (see for a more complete listing of the mammals of the Dieng mountains Nijman & Setiawan, 2001). Scientific research would not only provide new information and better insight, it would also serve as a constant reminder of the conservation importance of the Dieng mountains. Contrary to MacKinnon (1987), we strongly believe that there is room for a further expansion of the reserve system on Java. In Central Java, currently scarsely any large terrestrial area has an adequate protected status, although several have been proposed as nature reserves or wildlife sanctuaries (MacKinnon et al., 1982). As conversion of the last remaining natural forest areas on Java is an ongoing process, resulting in an ever increasing fragmentation, it is of utmost importance to raise one or preferable more areas in the central part of the island to a higher conservation status. The Dieng mountains support relatively high numbers of two of Indonesia's most threatened primate species, the Javan gibbon and the grizzled leaf monkey, making it a top priority area for primate conservation. The area is amongst the most diverse left on Java, notably it is one of the few unprotected forests that covers the whole range from lowland to upper montane. It harbours 67 per cent of all bird species endemic to Java and Bali (20 of 30 species), and 70 per cent of the Javan Restricted Range forest birds (23 of 33 species; see Box 8.1). The creation of a reserve in the Dieng mountains would be of international importance for global biodiversity conservation. The area is probably sufficiently large to ensure the maintenance of viable populations of unique flora and fauna. Following Nijman & Sözer (1996) we suggest that the proposed wildlife sanctuary (MacKinnon et al., 1982) be extended to the north-west so as to include the lowland and lower montane forest of Linggo. In the central area, the boundary should follow the line between the summits of Mt. Rogojembangan and Mt. Kendalisodo. From Mt. Kendalisodo the boundary should run westwards -excluding the village of Petungkriyana- up to the Sengkarang River. From there the boundary should follow the eastern and southern slopes of Mt. Lumping. We recommended that the provincial or national authorities eassessed the feasibility of creating a nature reserve or national park in the Dieng mountains. If the reserve is to become a 124

A Faunal Survey of the Dieng Mountains, Central Java

national park consideration should be given whether or not to include the Dieng plateau, with the three established lake nature reserves. The results of the present survey, and the proposals to extend the proposed wildlife sanctuary into the lowland zone, have been well received by the Directorate General of Forest Protection and Nature Conservation (PHPA), and currently gazettment is being processed. Table 8.4 Endemic and Restricted Range (R.R.) bird species recorded in the Dieng mountains. Endemism is defined as being restricted to the Javan faunal region i.e. the islands of Java and Bali, while a Restricted Range species is a species which has a breeding range fewer then 50,000 km2 (Bibby et al., 1992; Sujatnika et al., 1995). Nomenclature and sequence follow Andrew (1992). English name Javan Hawk-eagle Chestnut-bellied Patridge Pink-headed Fruit-dove Yellow-throated Hanging-parrot Javan Kingfisher Brown-throated Barbet Black-banded Barbet Orange-fronted Barbet Sunda Minivet Orange-spotted Bulbul Sunda Streaked Bulbul Sunda Blue Robin White-breasted Babbler White-bibbed Babbler Crescent-chested Babbler Grey-cheeked Tit-babbler Javan Fulvetta Javan Tesia White-rumped Warbler Rufous-tailed Fantail White-bellied Fantail Pygmy Tit White-flanked Sunbird Violet-tailed Sunbird Grey-throated Darkeye Sunda Serin

Scientific name Spizaetus bartelsi Arborophila javanica Ptilinopus porphyreus Loriculus pusillus Halcyon cyanoventris Megalaima corvina Megalaima javensis Megalaima armillaris Pericrocotus miniatus Pycnonotus bimaculatus Hypsipetes viriscens Cinclidium diana Stachyris grammiceps Stachyris thoracica Stachyris melanothorax Macronous flavicollis Alcippe pyrrhoptera Tesia superciliaris Seicercus grammiceps Rhipidura phoenicura Rhipidura euryura Psaltria exilis Aethopyga eximia Aethopyga mystacalis Lophozosterops javanicus Serinus estherae

Endemic • • • • • • •

• • • • • • • • • • • •

R.R. • • • • • • • • • • • • • • • • • • • • • • •

ACKNOWLEDGEMENTS We thank the Indonesian Institute for Science (LIPI) for sponsoring our research and the Directorate General of Forest Protection and Nature Conservation (PHPA) for allowing us to conduct our fieldwork. BirdLife International Indonesia Programme is thanked for their support and especially the co-operation of Paul Jepson and Sujatnika is acknowledged. Sujatnika helped with preparing the map. Resit Sözer is 125

Forest (and) Primates

thanked for his work during parts of the 1994 surveys, as are Iwan Setiawan and Andi Prima Setiadi. Drs Boeadi (Museum Zoologi Bogor) helped with identification of prey remains. Dr H. Albrecht, Dr P.J.H. van Bree, Dr J. Chapman and three reviewers commented on earlier drafts.

126

Distribution and Conservation of the Proboscis Monkey

CHAPTER 9 DISTRIBUTION AND CONSERVATION OF THE PROBOSCIS MONKEY NASALIS LARVATUS IN KALIMANTAN, INDONESIA. with Erik Meijaard, Biological Conservation 92: 15-24, 2000 ABSTRACT The proboscis monkey Nasalis larvatus is endemic to the island of Borneo. A review of the species' distribution reveals that it occurs throughout Kalimantan, the Indonesian part of Borneo, from the coastal areas to the headwaters of probably all major rivers. Proboscis monkeys are more widely distributed than has been thought previously, and were never confined to the coastal and downstream areas of rivers on the island of Borneo (including what is now Sabah, Sarawak, Brunei and Kalimantan) as has been assumed in some primate literature. Proboscis monkey habitat, i.e. riverine and coastal forest, is the most threatened of all vegetation types in Borneo, owing to conversion into agricultural land and logging. Another threat to their survival is hunting. The combination of these threats has reduced populations of N. larvatus in Sabah, Sarawak and East Kalimantan, and based on this it is expected that other populations elsewhere in Borneo are likewise threatened. Our study shows the present low efficiency of conservation programmes in Kalimantan, which adds to the problem of protecting N. larvatus. For the survival of the species the populations in Kalimantan are still of great significance, as they are considerably larger than those in Sabah, Sarawak, and Brunei. We therefore recommend the protection of some of the largest populations in order to ensure the long-term survival of the species. RINGKASAN Penyebaran dan konservasi Bekantan Nasalis larvatus di Kalimantan, Indonesia (bersama Erik Meijaard, Biological Conservation 92: 15-24, 2000): Bekantan Nasalis larvatus merupakan spesies endemik Pulau Borneo. Tinjauan ulang terhadap distribusi spesies ini menunjukkan bahwa spesies ini terdapat di seluruh Kalimantan, Borneo bagian Indonesia, dari daerah pesisir sampai hulu sungainya dari kemungkinan besar semua sungai-sungai besar. Bekantan ternyata tersebar lebih luas dibanding dengan yang diperkirakan sebelumnya, dan tidak hanya terbatas di pesisir dan kawasan hilir sungai-sungai di Pulau Borneo (termasuk Sabah, Sarawak, Brunei dan Kalimantan) seperti diperkirakan di beberapa literatur primata. Habitat bekantan, yaitu hutan di sepanjang aliran sungai dan pantai, merupakan tipe vegetasi yang paling terancam di Borneo, karena peralihan menjadi lahan pertanian dan penebangan kayu. Ancaman lainnya adalah perburuan. Gabungan ancaman-ancaman 127

Forest (and) Primates

ini telah menurunkan populasi N. larvalis di Sabah, Sarawak dan Kalimantan Timur, dan berdasarkan hal tersebut diperkirakan bahwa populasi-populasi lainnya di Borneo juga terancam. Penelitian kami memperlihatkan rendahnya efisiensi program-program konservasi di Kalimantan, yang menambah permasalahan dalam perlindungan Bekantan. Untuk keberlangsungan hidup spesies ini populasi-populasi di Kalimantan masihlah penting sekali, di mana populasi-populasi itu sangat lebih besar dibandingkan dengan di Sabah, Sarawak, dan Brunei. Karena itu kami merekomendasikan perlindungan beberapa populasi yang terbesar untuk menjamin keberlangsungan hidup spesies ini. INTRODUCTION The proboscis monkey Nasalis larvatus (van Wurmb, 1787) is a large, sexually dimorphic Colobine, endemic to the island of Borneo. It is closely associated with waterways, returning to the water's edge in the evening, and rarely ranging far from rivers, generally 250,000, with c. 25,000 protected inside reserves. Yeager & Blondal (1992) considered this last figure too high and made an adjustment to < 5,000 animals inside protected reserves. N. larvatus is protected by law throughout its range, and is listed on Appendix I of the CITES convention. The aim of this paper is twofold: (1) to provide information on the distribution of N. larvatus throughout Kalimantan in historic and present times, and (2) to provide information on threats facing the species.

128

Distribution and Conservation of the Proboscis Monkey

METHODS Information was obtained by direct observation and by interviewing local people on the status of N. larvatus: by the author EM from 1994-1997 in the course of a Kalimantan-wide orang-utan Pongo pygmaeus survey, and by VN in the framework of a WWF-Indonesia survey in 1996, which concentrated on Colobine monkeys. For the orang-utan survey a total of 78 field-checks involving 208 field days were made in Kalimantan. In addition 69 days were spent in towns and villages for official visits and interviews. During that period a total of almost 35,000 km were travelled by various means of transportation, including transects on foot. The surveys covered all major river systems of West, Central, and East Kalimantan, all main towns in Kalimantan, and mountainous areas in Central and East Kalimantan. The WWF survey was conducted between September and December 1996, totalling 78 field days, and covered the north-eastern part of East Kalimantan. Because of the limited means of transportation in Kalimantan, survey routes were mostly dictated by the course of rivers and roads and were not randomly chosen. Additionally, 32 days were spent in Sabah, Sarawak, and Brunei. Figure 9.1 shows the survey routes.

Figure 9.1

Area covered during the 1994 - 1997 survey. Thick lines indicate the survey routes.

129

Forest (and) Primates

Interviews were conducted in a semi-structured fashion. Questions were asked in Indonesian, which is also understood in the Malaysian states and Brunei. Interviews always started as an informal conversation, and if the informant knew about general wildlife subjects, the interviews became more specific. As the orang-utan was the main focus of the larger part of the survey, questions initially addressed that particular species, after which information about other wildlife species was collected, including proboscis monkeys. Questions mostly concerned absence or presence of the species and threats to its survival. The anecdotal information used in this survey provides only subjective data, and information on absence of proboscis monkeys was not recorded. We only include locations from where the species was reported by at least two independent sources (c.f. Salter & MacKenzie, 1985). Additional information was obtained from literature, and from biologists and conservationists working in Kalimantan. The Environmental Impact Assessment reports of Kalimantan's logging concessions also provided information on the present distribution of N. larvatus. A total of 115 reports were consulted. All N. larvatus records were given a latitude/longitude coordinate and entered in a computerized Geographic Information System (GIS), using PC ArcInfo and PC ArcView software. Other data layers contain information on the 1993 forest cover (World Conservation Monitoring Centre (WCMC) data base), and topography. Habitat was classified according to personal observation, data from literature, and via personal communication. The distance from the N. larvatus locations to the coast was calculated by proximity analysis in the spatial query builder of PC ArcView. For analysis we only included recent data. This was arbitrarily taken as records (published) after 1980. We have listed all information on the occurrence of N. larvatus, regardless of the numbers in which they may occur. For the 15 of what are probably the largest populations we give an indication of population sizes based on personal observation, personal communication, and information in the literature, taking into account the approximate extend of suitable habitat. Population sizes are classified as: 1: < 100; 2: 100-1000; and 3: >1000. RESULTS Distribution Groups of N. larvatus were observed at 30 locations during the survey, and an additional 123 records were derived from literature and interviews (Fig. 9.2, Appendix 9.1). The species is scattered throughout Borneo from the mangroves and small islands in the coastal deltas, along virtually all major rivers to numerous inland sites. Eight percent (12/153) of the presence reports were > 200 km directly inland from the coast; 18 % (n = 28) between 100 and 200 km; 16 % (n = 25) between 50 and 100 km; and 58 % (n = 88) < 50 km from the coast. Distances to the coast following the course of the rivers are even larger, often > 300 km, and sometimes as much as 750 km. Our data suggest that N. larvatus does not occur at high altitudes. Over 90 % (n = 138) of the presence records were from altitudes < 200 m a.s.l., and 130

Distribution and Conservation of the Proboscis Monkey

the highest reports are from c. 350 m a.s.l. The occurrence of inland groups have sometimes been thought to consist of wandering males only (C. Yeager, pers. comm.). It is therefore important to note that our inland sightings included groups of N. larvatus comprising juveniles and females with dependent young.

Figure 9.2

The 153 recent records of proboscis monkey Nasalis larvatus on Borneo as mentioned in the text. The numbers 1-16 refer to priority areas for N. larvatus mentioned in the text and Table 9.1. Protected areas in Borneo are cross-hatched.

Within Kalimantan large populations have been reported from Danau Sentarum Wildlife Reserve (No. 3 in Fig. 9.2 and Table 9.1; c. 600, Sebastian, 1994), and from Gunung Palung National Park (No. 1; “several hundreds”, Yeager & Blondal, 1992), Tanjung Puting National Park (No. 5; c. 2000, Yeager & Blondal, 1992), and the Mahakam delta (No. 16; 600 - 900, Alikodra et al., 1992). Our data indicate that within Kalimantan a number of other areas remain where there are relatively large populations of N. larvatus. Notably, this includes the delta of the Sungai Sesayap 131

Forest (and) Primates

(Sungai = River, hereafter abbreviated as ‘S’), S. Sembakung and S. Sebuku (No. 15), the Mahakam lakes area (No. 10), the fresh water and peat swamp forest areas of Central Kalimantan (No. 6), and the Kendawangan area (No. 2). Table 9.1

Priority areas for the protection of proboscis monkeys Nasalis larvatus in Kalimantan, Indonesia.

Noa

Priority Areab

Statusc

Numbersd

Habitat e

Main threats

Reference f

1 2 3 4 5

G. Palung Kendawangan Danau Sentarum Sambas Paloh Tanjung Putting

NP NR WR UNP NP

2 2-3 2-3 2 3

F, P, Ri. LoD F, P F, P, Ri Ni, Ma, LoD, P F, P, LoD

2, 3 4, 5 1, 6, 15 1, 7, 8 1, 3

6

Central Kalimantan NR/UNP Rivers Lower S. Barito UNP P.Laut UNP Balikpapan Bay UNP S. Mahakam and UNP Lakes S. Kedang Kepala NR/UNP Kutai NP S. Kayan UNP Sangkulirang UNP S. Sesayap, S. UNP Sebuku, S. Sembakung Mahakam delta UNP

2-3

F, P, Ri

Illegal logging Illegal logging Fire, hunting, logging Logging Goldmining, logging, hunting Swamp reclamation

2 ? 1-2 2

Ri Ri, LoD, F Ri, Ma, F Ri, F, P

1 9 1 1, 5, 8, 14

1-2 1 1 1-2 3

Ri, F, P Ri, F, P, LoD Ri, F, Ma, Ni Ri, F, Ma, Ni Ri, F, Ma, Ni

Logging Logging Logging Logging, disturbance, hunting Logging, fire Logging, fire Logging, hunting Logging Shrimp farming, logging

2

Ma, Ni

Shrimp farming, logging

17

7 8 9 10 11 12 13 14 15 16 (a) (b) (c) (d) (e) (f)

1

1 1, 10, 11, 12 1, 8 1, 13 1, 5, 16

cf. Figure 9.2. G.= Gunung (Mountain); P. = Pulau (Island); S. = Sungai (River). NP = National Park; NR = Nature reserve; WR = Wildlife Reserve; UNP = Unprotected. Indication of population sizes: 11000 individuals. See text for details. Mangrove forest (M), Freshwater swamp (F), Peat swamp (P), Riverine (Ri), Lowland Dipterocarp (LoD), and Nipah palm (Ni). 1 personal observation; 2 MacKinnon & Warsito, 1982; 3 Yeager & Blondal, 1992; 4 Noor & Hanafia, 1994; 5. K.Jeanes, pers. comm.; 6 Sebastian, 1994; 7 McCarthy, 1997; 8 Silvius et al., 1987; W. Smits, pers. comm.; 10 Suzuki, 1984; 11 MacKinnon et al., 1994; 12 Rodman, 1978; 13 Wibowo pers. comm.; 14 R. Sözer, pers. comm.; 15 R. Dennis pers. comm.; 16 Momberg et al., 1998; 17 Alikodra et al., 1992.

Threats The only direct threat to the survival of N. larvatus recorded during this study was hunting. Hunting was reported at six locations during the three year surveys. In the northern Mahakam lakes, in East Kalimantan, N. larvatus is increasingly hunted to serve as bait for monitor lizards Varanus salvator, whose skins are highly valued. 132

Distribution and Conservation of the Proboscis Monkey

Further inland the species is hunted for food (R. Sözer, pers. comm). In the Danau Sentarum Wildlife Reserve, West Kalimantan, at least two proboscis monkeys were shot by army or police hunters in late 1996 (R. Dennis, pers. comm.). Iban people from the area north of the Reserve will opportunistically shoot the monkeys while on turtle hunting expeditions during the dry season, much to the annoyance of local Muslim fishermen (pers. obs.). Three instances of hunting along the S. Kayan and S. Pangean (where the species is rare) were reported by two Dayak hunters in 1996. Finally, it was reported by inhabitants of Tanjung Selor and its surroundings, East Kalimantan, that since the early 1990s the population of proboscis monkeys had declined dramatically due to hunting by non-Muslim Dayaks. Another threat to N. larvatus' survival is habitat destruction. Table 9.1 shows that in the 16 priority areas for protection, the main factors contributing to habitat destruction are logging (both legal and illegal), forest fire, gold mining, swamp reclamation and shrimp farming. Unfortunately, the proboscis monkeyís specialised habitat coincides with the areas on Borneo that are the first to be colonised, farmed, industrialised, and least protected by man. Table 9.2 shows that of the most important N. larvatus habitat types, only c. 39% has survived, of which 15% is protected. Hereby we need to note that the data in Table 9.2 are based on MacKinnon & Artha’s 1981 study, as no more recent data are publicly available. After almost two decades of a continuous logging regime (for reviews see Sunderlin & Resosudarmo 1996, and Rijksen & Meijaard 1999) considerably less habitat remains. Table 9.2

Proboscis monkey Nasalis larvatus habitat (km2) in Kalimantan (after MacKinnon et al., 1996)

Vegetation type Freshwater swamps Peat swamps Mangrove forest Wet lowland forest (alluvium) Moist lowland (alluvium) Dry lowland forest (alluvium) a.

Original area 38,950 44,030 15,600 22,010 870 210

Total remaining ( c. 500 km2) provide refuge for nine or more species, in three more than ten species have been recorded, and only one (Kutai National Park) is home to the maximum number of eleven sympatric species (Table 11.3). An analysis of the endemic species leads to a similar reserve listing. Few reserves harbour population of more than four sympatric endemics and in reserves that do, often one or two are very rare (e.g., Danum Valley and Gunung Palung). The only reserve that has populations of all five endemics within its boundaries is again Kutai National Park. The real tragedy is, however, that of the five most species-rich reserves, three have been almost completely destroyed by a combination of illegal and legal logging, mining, encroachment, and the forest fires of 1982-1983 and 1997-1998 (MacKinnon et al., 1994, Rijksen & Meijaard, 1999, Suhartoyo & Toma, 1999, Jepson et al., 2001). Additionally, law enforcement with respect to logging, extraction of non-timber products, hunting, capturing of wildlife etc. is virtually absent; primates are not exempt. Table 11.3

List of large reserves that rank(ed) among the highest in primate species richness↑

Reserve Kutai Muara Kaman Gunung Palung Bukit Suharto Danum Valley

Size (km2) 2000 625 900 710 438 (2476)↑↑

Number of Number of endemics species (13 is 100%) (5 is 100%) 11 5 10 4 10 3 9 4 9 4

Notes Forest largely, if not completely, destroyed Forest largely, if not completely, destroyed P. frontata rare and restricted to the eastern part Forest largely, if not completely, destroyed N. larvatus rare



Within the reserves many species show a restricted distribution, and may even be more common outside the reserve proper. ↑↑ The larger figure represents the Ulu Segama commercial forest reserve of which Danum Valley forms a part.

In theory, all species can be included in the protected area network if in addition to (part of) the central-eastern part of East Kalimantan, parts of West and / or Central Kalimantan south of the Kapuas River (to include the agile gibbon Hylobates agilis) and the low-lying regions of Sarawak (to include the banded leaf monkey Presbytis femoralis) are added.

162

Patterns of Primate Diversity on Borneo

Few reserves have been proposed in the region most rich in primates, those of significance include the proposed south-eastern extension Kayan Mentarang National Park (Ulu Kayan Mutlak 3500 km2) and Sangkulirang Peninsula (Mangkilat-Kelompok Kapur Sangkulirang 2000 km2). However, unfortunately, in both of these areas densities of primates are generally low, in part due to the efforts of the Kenyah Dayaks and especially the (formerly) nomadic Penan and Punan. With their formidable hunting skills these tribes are able to assure that all but the smallest primates remain rare.

DISCUSSION The number of locality record holds little information on the actual abundance of the majority of species. Proboscis monkeys are less common than the number of records (n=153) indicate (see Meijaard & Nijman, 2000), but their striking appearance and their occurrence near waterways makes them particularly easy to see and identify. In contrast, the two nocturnal species are possibly more common then the data suggest but they are rarely recorded because of their activity pattern and silent behaviour. The large number of records for the orang-utan (n=273) largely derives from the fact that one of us (E.M.) has been conducting surveys to document the distribution of this very species (see Rijksen & Meijaard, 1999); survey efforts were concentrated in those areas where the species was expected to occur. In fact, the orang-utan is among the rarest primate in Borneo. The banded leaf monkey (n=26) is probably genuinely rare on Borneo for at least two reasons. First, it has the most restricted distribution of any primate, and second, it is confined to tropical wet-evergreen forest below 1000 m a.s.l in an area that is largely deforested. Likewise, long-tailed macaques (n=219) are genuinely common as they occur all over the island, in a large range of habitat types and are quite tolerable to habitat disturbance. Red leaf monkey (n=190), on the contrary, have a much more restricted range of habitats in which they occur, but they are found almost over the entire island. Both species are also easily identified and quite vocal which also facilitates identification. The present study reconfirms the importance of the tropical wet evergreen forest below 500 m a.s.l., and especially near rivers as one of the most important habitats for wildlife. For primates, including many of the endemics, this habitat type is of prime importance. Peat swamp forest comes out as an habitat of importance as well, although less so for most of the endemics. Of the two most threatened species (the orang-utan and the proboscis monkey) considerable populations do occur in the peat swamp forests (Meijaard & Nijman, 2000; Meijaard, 1997). Only small amounts of this forest type have so far been included in the protected area network, despite the fact that some considerably large area still remain. Restricted parts of northern Borneo have been identified as centers of endemism and as particularly species rich for a number of taxa including vascular plants, insects, birds, and mammals. Whether these patterns are a true reflection of 163

Forest (and) Primates

the distribution of these taxa or that it reflects a strong research bias remains to be seen. However, quite illustrative are recent findings for birds. Until recently, a great number of species was considered to be confined to Mt Kinabalu or the northern mountains (Smythies, 1981; MacKinnon & Phillipps, 1993). Field work in mountain regions in Kalimantan and in the interior, however, revealed that the 'northern mountain endemics' are neither confined to the north nor to the heighest mountains (Wilkinson et al., 1991; Van Balen 1997, 1999a, 2000) and in fact some species have been recorded as far as the Schwaner range (Rice, 1989), Mt Niut (Prieme & Heeregaard, 1988), and the Meratus mountains (Davidson, 1997). Expert ornithologists can identify most species by sighting or call and there is generally no need to collect specimens. Hence fairly complete species lists can be made with relatively little effort. Many other taxa (including small mammals, insects, and plants) can only be identified when collected and preserved, and so far, intensive collections have been made more in northern Borneo than in any other part of Borneo. Although progress is made in uncovering the distribution patterns of even the most elusive of animals, e.g., the bay cat Catopuma badia (Meijaard, 1997), biologically, large areas of the island remain virtually a terra incognita. For primates, we found a centre of species richness and endemism in the central eastern part of East Kalimantan. To our knowledge this has not previously been recognised as such. This finding derives in part from more detailed distribution maps of especially three colobines viz. proboscis monkey, silvered leaf monkey, and white-fronted leaf monkey, and the orang-utan. The notion that eastern-central East Kalimantan is a centre for primate endemism is only in part related to the taxonomic position taken. Including the orang-utan as an endemic species only emphasises the importance of the area, albeit smaller in size. A change of the taxonomy of the banded leaf monkeys Presbytis femoralis, however, would put more emphasis on northern Sarawak as a centre for primate endemism. Populations of this species on Borneo are generally considered to be congeneric with taxa occurring on the Natuna Islands, southern Malay Peninsula and eastern Sumatra (Brandon-Jones, 1984). However, the populations on Borneo do differ in their pelage characteristics from populations from other islands (most pronounced in P. f. cruciger from northern Sarawak), and in fact the banded leaf monkey may comprise several species (V. Nijman, unp. data; cf. Eudey, 1987). The fact that central-eastern East Kalimantan is a hotspot for primate diversity illustrates the invalidity of high diversity overlap for different taxa, and points to the importance of identifying complementary sites (cf. Pressy et al., 1993, Williams et al., 1997; Howard et al., 1998). It is expected that for a number of other taxa, especially with a high level of frugivory, this area will turn out to be of equal importance. From a conservation perspective our findings are gloomy as well as sobering. The regions identified as of prime importance are among the least protected in Borneo. Large parts of the east-central hotspot have suffered from the large forest fires both in 1982-1983 and 1997-1998 (Goldammer et al., 1999). Forest that has 164

Patterns of Primate Diversity on Borneo

burned twice is of little importance to primate conservation (pers. observ.). In theory, forest areas that have burned once can regenerate and, given the fact that even in the worst affected areas some small patches of 'good' forest with some primate groups remain, these areas can be of potential importance for primate conservation. However, unless one takes an unrealistically positive view, in practice, most if not all of these forests will burn again in the next El Niño Southern Occilation Event; the large amount of dead wood makes them extremely fire-prone. The remaining patches of unburned forest are few and far in between. Unless these areas are actively protected during the next dry years it is feared that these will be seriously affected by forest fires as well. Given the amount of active forest protection that was given during the previous forest fires and their succes rate, it has to be feared that most forests will disappear within the next ten to twenty years. The remaining unprotected forest is all earmarked for conversion or selective logging; actually, the total area of consession is considerably larger that the remaining forest (Rijksen & Meijaard, 1999). Selective logging will make the forest more fire-prone. Conservation of primates in the region's protected area network offers little more hope for optimism. For example, if not for the signs indicating entering Kutai National Park or Bukit Suharto Greater Forestry Park it would be hard to note one enters a conservation area. Like Muara Kaman Nature Reserve, both areas are almost completely deforested (pers. observ.). Combined these areas originally included almost 3500 km2 of prime primate habitat but at present they are of hardly any value for primate conservation. ACKNOWLEDGEMENTS The surveys were conducted in co-operation with the Directorate General for Forest Protection and Nature Conservation (PKA, formerly PHPA), the Ministry of Forestry and Estates Crops (MOFEC formerly MOF) and under sponsorship of the Indonesian Institute for Sciences (LIPI). Steph B.J. Menken (IBED, University of Amsterdam) is thanked for comments on previous versions of this paper.

165

Forest (and) Primates

166

Re-Assessment of IUCN Conservation Status of the Endemic Primates

CHAPTER 12 RE-ASSESSMENT OF IUCN CONSERVATION STATUS OF THE ENDEMIC PRIMATES OF JAVA AND BORNEO ABSTRACT A re-assessment of the IUCN status of the endemic primates of Java and Borneo was made on the basis of a study from 1994-2001 and additional data from the literature. Five (or six) of the nine (or ten) endemics have their current status changed. Fuscous leaf monkey Presbytis frederica is currently classed as Data Deficient, but the species is here synonymised with grizzled leaf monkey P. comata. This species is currently classed as Endangered and this status remains unchanged in the reassessment. Bornean leaf monkey P. hosei and Bornean gibbon Hylobates muelleri are currently both classed as Lower Risk but on the basis of a sharp reduction of available habitat aggravated by hunting, both species are more appropriately classed as Vulnerable. The white-fronted leaf monkey Presbytis frontata is currently classed as data Deficient, but on the basis of low population densities over the greater part of its range, and the sharp reduction of its lowland forest habitat it is classed as Vulnerable. The proboscis monkey Nasalis larvatus is currently classed as Vulnerable, but given a sharp reduction in available habitat and inadequate protection inside reserves, it is more appropriately classed as Endangered. Finally, the Javan gibbon Hylobates moloch is currently classed as Critically Endangered, but given a less dramatic decrease of remaining forest on Java as suggested, and the finding of significant populations in Central Java, the species is more appropriately classed as Endangered. RINGKASAN Peninjauan kembali mengenai status konservasi menurut IUCN dari primata endemik di Jawa dan Kalimantan: Peninjauan kembali kami buat mengenai status IUCN dari primata endemik di Jawa dan Kalimantan berdasarkan hasil sebuah penelitianlapangan dan survai pada tahun 1994-2001 ditambah data pustaka. Lima (atau enam) dari sembilan (atau sepuluh) jenis kera telah berubah status konservasinya saat ini. Rekrakan Presbytis fredericae saat ini diklasifikasi sebagai Kurang Data, tetapi jenis ini dianggap kami sejenis dengan Surili P. comata. Jenis ini dianggap Genting saat ini, dan status ini tidak berubah pada tinjauan kembali kami. Baik Bangat P. hosei maupun Kelawat Hylobates muelleri saat ini diklasifikasi Berisiko Rendah, tetapi kedua jenis ini lebih tetap diklasifikasi sebagai Rentan, berdasarkan penyusutan drastis dari habitat yang ada, yang diperburuk oleh pemburuan. Lutung dahi putih Presbytis frontata saat ini diklasifikasi sebagai 167

Forest (and) Primates

Kurang Data. Tetapi statusnya kami anggap sebagai Rentan, berdasarkan kepadatan rendah di bagian terbesar wilayah penyebarannya. Bekantan Nasalis larvatus adalah satu jenis saat ini diklasifikasi Rentan tapi terbatas pada kantung-kantung terakhir dari hutna tropis dataran di pingir sungai dan hutan bakau, lebih tetap diklasifikasi bagai Genting. Terakhirnya, Owa Jawa H. moloch adalah satu jenis diklasifikasi Kritis tetapi kerena masih ada beberapa populasi yang cukup besar di Jawa Barat dan Jawa Tengah, lebih tetap diklasifikasi bagai Genting. INTRODUCTION In the following section a re-assessment of the IUCN conservation status of the endemic primates of Java and Borneo is presented. It is based on data presented in the previous chapters of this thesis, additional and as yet unpublished results from the study, and published data. Firstly, for each species its legislative status is given, i.e. whether or not it is protected by Indonesian and / or Malaysian law. When the species is included on Appendix I or II of the Convention on International Trade in Endangered Species (CITES) this is indicated. Secondly, its present IUCN listing (based on Eudey, 1996/1997) is given with the criteria according to which it is included in that particular category. This is followed by the suggested IUCN listing and its criteria. Finally, the justification of the changes (if any) are presented. RESULTS AND DISCUSSION Grizzled leaf monkey Presbytis comata Legislative status: Protected by Indonesian Law as Presbytis aygula (Surat Keputusan Menteri Pertanian No 247/Kpts/Um/4/1979, Undang-undang No. 5 / 1990, Surat Keputusan Mentri Kehutanan No 301/KPT-II/1991 and No. 882/KPTII/1992). Included on Appendix II of the Convention on International Trade in Endangered Species (CITES). Present IUCN status: Endangered based on criteria: A1(c): A reduction of at least 50% over the last three generations (