Thesis GF.Indb

UvA-DARE (Digital Academic Repository)

Effects of El Niño and large-scale forest fires on the ecology and conservation of Malayan sun bears (Helarctos malayanus) in East Kalimantan, Indonesian Borneo

Fredriksson, G.M.

Publication date 2012 Document Version Final published version

Link to publication

Citation for published version (APA): Fredriksson, G. M. (2012). Effects of El Niño and large-scale forest fires on the ecology and conservation of Malayan sun bears (Helarctos malayanus) in East Kalimantan, Indonesian Borneo.

General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.

UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

Download date:04 Oct 2021 Effects of El Niño and forest ﬁ of sun bears,res and conservation on the ecology Indonesian Borneo Effects of El Niño and large-scale forest ﬁ res on the ecology and conservation of Malayan sun bears (Helarctos malayanus) in East Kalimantan, Indonesian Borneo G. Fredriksson

Gabriella Fredriksson Effects of El Niño and large-scale forest fires on the ecology and conservation of Malayan sun bears (Helarctos malayanus) in East Kalimantan, Indonesian Borneo Effects of El Niño and large-scale forest fires on the ecology and conservation of Malayan sun bears (Helarctos malayanus) in East Kalimantan, Indonesian Borneo

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctor

aan de Universiteit van Amsterdam

op gezag van de Rector Magnificus

prof. dr. D.C. van den Boom

ten overstaan van een door het college voor promoties

ingestelde commissie,

in het openbaar te verdedigen in de Agnietenkapel

op donderdag 15 november 2012 om 12:00 uur

door

Gabriella Margit Fredriksson

Geboren te Amsterdam, Nederland Promotiecommissie

Promotor: Prof. Dr. S.B.J. Menken

Co-Promotores: Dr. D.L. Garshelis Prof. Dr. J.E. Swenson

Overige leden: Prof. Dr. P. van Tienderen Dr. H.D. Rijksen Prof. Dr. A.M. Cleef Prof. Dr. J.H.D. Wolf Prof. Dr. H. de Iongh

Faculteit der Natuurwetenschappen, Wiskunde en Informatica

Front cover illustration: In the foreground a sun bear is holding a wild durian fruit (Durio dulcis), which only fruits during the mast episodes. On the left is a strangling ﬁg (Ficus spp.) a species that fruits at irregular intervals, one of the most important fruit species in the sun bear diet during the prolonged inter-mast period. To the right is a tree with clawmarks left behind after a bear climbed the tree. In the center a sun bear is feeding on a termite mound. Fires, set by humans, are looming in the background. Back cover illustration: On the right a sun bear is climbing a tree to investigate a stingless bee’s nest: only the characteristic nest entrance tube can be seen at the top of the tree. Small pitcher plants at the base of the tree hold rainwater, occasionally used by sun bears to quench their thirst. In the lower left corner a large reticulated python can be seen, one of the few natural predators of sun bears. Behind the snake, coconut trees planted near the forest edge by farmers have been covered with metal sheeting to protect them from sun bears climbing up and eating the growth shoot. In the back farmers have set ﬁres to clear land.

This publication (cover and interior) is printed on recycled paper, Revive®. With ink from renewable resources and alcohol free fountain solution. Recycled, EU Flower, FSC, ISO 14001.

More info: http://www.haveka.nl/greening This thesis is dedicated to sun bears, in particular my bear friends Ganja, Ucil, Schizo and si Kecil may enough forest be conserved for this wonderful species to persist till eternity!

And to Ingmar… may he become inspired to safe forests and wildlife in the future! Contents

Chapter 1: General introduction 1

Chapter 2: Frugivory in sun bears (Helarctos malayanus) is linked to El Niño-related fluctuations in fruiting phenology, East Kalimantan, Indonesia 17 with S. A. Wich and Trisno Biological Journal of the Linnean Society (2006), 89(3): 489-508

Chapter 3: Movement and activity patterns of female sun bears in East Kalimantan, Indonesian Borneo: implications for conservation 53 with D. L. Garshelis

Chapter 4: Impacts of El Niño-related drought and forest fires on sun bear fruit resources in lowland dipterocarp forest of East Borneo 81 with L. S. Danielsen and J. Swenson Biodiversity and Conservation (2006), 15(4): 1271-1301

Chapter 5: Human-sun bear conflicts in East Kalimantan, Indonesian Borneo 103 Ursus (2005), 16: 130-137 viii Contents

Chapter 6: Predation on sun bears by reticulated python in East Kalimantan, Indonesian Borneo 119 The Raffles Bulletin of Zoology (2005), 53(1): 165-168

Chapter 7: Exploring the use of sign surveys for monitoring relative

abundance of sun bears: a case study in burned and unburned forests in East Kalimantan, Indonesian Borneo 131 with D.L. Garshelis, W.J. Sastramidjaja, S.B.J. Menken

Chapter 8: Synthesis and conclusions 163

Acknowledgements 179 Summary 185 Samenvatting 189 Ringkasan 195 Curriculum Vitae and List of Publications 201 Chapter 1

General introduction General introduction 3

General introduction

Among the world’s tropical regions, Southeast Asia is of particular conservation concern because it has the highest rate of habitat loss (Sodhi et al. 2004, 2010, Miettinen et al. 2011). Most of Southeast Asia is considered a biodiversity hotspot because it is home to an exceptionally high number of endemic species that are threatened by the loss of >70% of original habitats (Myers et al. 2000). Southeast Asia is also highlighted as an area where past and present human-driven land-use changes are expected to cause extinctions across a wide range of taxa (Brook et al. 2003, Cardillo et al. 2006, Lee and Jetz 2008). Furthermore, environmental apathy, corruption, poor natural resource governance, poverty, and limited conservation fund- ing constitute difficult obstacles for conservationists in Southeast Asia (Sodhi et al. 2004, 2010, Sodhi and Brook 2006, Posa et al. 2008). Indonesia, the country with the largest expanse of remaining contiguous forest in Southeast Asia, has put its forests and wildlife under serious pressure in recent decades through a combination of unsustainable logging practices (Curran et al. 2004, Fuller et al. 2004, Meijaard et al. 2005), high rates of deforestation for plantation development (Hansen et al. 2009, Miettinen et al. 2011), and a succession of devastating forest fire events (Lennertz and Pfanzer 1984, Malingreau 1985, Siegert et al. 2001). Kalimantan, the Indonesian part of Borneo that covers 73% of the island, has experienced two extreme forest fire events in modern times, the first in 1982-83 and the second in 1997-98. Both these fire events coincided with severe and prolonged droughts associated with severe El Niño-Southern Oscillation (ENSO) phenomena (Siegert et al. 2001). The fire event in 1982-83 affected an estimated 3.5 million ha (Mha) of forest in Kalimantan alone (Guhardja et al. 2000), whereas the disastrous fire event of 1997-98 affected circa 5 Mha of land in the province of East Kalimantan (Indonesian Borneo) alone, including 2.6 Mha of forest (Hoffmann et al. 1999, Siegert et al. 2001). Between 1997 and 2006, a total of 16.2 Mha of Borneo’s landmass1 (21% of total land surface area) have been affected by fires (Langner and Siegert 2009) (Fig. 1). These fires also entered several national parks and other protected areas (Curran et al. 2004, Langner and Siegert 2009), affecting their biodiversity value (Kinnaird and O’Brien 1998). At present the island is troubled by frequent smaller fire events, now occurring even during ‘normal’ dry seasons. This increased fire frequency is caused by a combination of droughts, forest degradation leading to increased susceptibility to fires, and the growing use of fire as a cheap tool for land clearing by plantation companies and farmers (Hoffman et al. 1999).This increase

1 For size comparison, the total land surface area of the Netherlands is 3.35 Mha. 4 Chapter 1

of ﬁres has primarily been connected with land speculation and development projects (Siegert et al. 2001, Page et al. 2002, Goldammer 2007). These recurring ﬁre events are not only degrad- ing vast areas of rainforest, including their biodiversity, but are also causing massive carbon dioxide emissions, which contribute to global warming (Page et al. 2002, Achard et al. 2004).

Figure 1: Map of Borneo depicting ﬁre affected areas (1997-2006) adapted from Langner and Siegert (2009).

Although forest fires have been recorded historically in western Indonesia (Goldam- mer and Seibert 1989), their incidence was rare, with intervals of hundreds to thousands of years (Cochrane et al. 1999). As a result, these tropical forests are maladapted to fire events (Slik et al. 2010), with fires having a devastating effect on tree survival (Leighton and Wirawan 1989, Woods 1989, Slik et al. 2002, van Nieuwstadt 2002). The combined effects of extreme drought and fire lead to more than 70% mortality of trees (>10 cm DBH) in once-burned areas two years after the fires in 1997-98 (van Nieuwstadt 2002). Animal populations are also affected by these forest fires, either directly through fire General introduction 5

injury or indirectly through changes in food availability or habitat structure (Whelan 1995). Mammals provide vital ecological services in tropical rainforests and are key in the structuring of biological communities through their roles in, amongst others, seed dispersal and predation, pollination, folivory, and frugivory, and as top predators (Terborgh 1988, 1992). Apart from a handful of papers on mortality in the immediate aftermath of fires (e.g. Kinnaird and O’Brien 1998, Peres 1999), and a small number of qualitative reports following the 1982–1983 ENSO event (Berenstain 1986, Leighton and Wirawan 1986, Azuma 1988, Doi 1988), information on the effects of fires on large vertebrates remains scarce and anecdotal (Oka 1999, O’Brien et al. 2003). Barlow and Peres (2006) examined the effects of both single and recurrent wildfires on fruit production and large vertebrate abundance in a central Amazonian terra firme forest, showing that many large frugivores and other vertebrate species declined in response to single fires, and most species dependent on primary forest were extirpated from twice-burnt forest (Barlow and Peres 2006). The only study on large vertebrate responses to forest fires from In- donesia by O’Brien et al. (2003) reported on the negative effects of forest fires on demography and persistence of siamang (Symphalangus syndactylus) groups in Sumatra. Their study indicated that it was doubtful whether groups of this large primate would be able to persist in burned- over areas for more than two generations. Besides their sensitivity to fires, forests in Southeast Asia are characterized by irregular, supra-annual, mast fruiting [i.e. synchronous variable production of large fruit crops (‘masts’) amid years of smaller crops] events (Appanah 1985, Ashton et al. 1988). These events are possibly triggered by environmental factors occurring during ENSO events (Ashton et al. 1988), although the influence of ENSO events on mast fruiting varies throughout Southeast Asia (Wich and van Schaik 2000): mast fruiting events and ENSO phenomena are most strongly correlated on the eastern side of the Malesian landmasses, specifically eastern Borneo (Ashton et al. 1988). The canopy and emergent trees of these forests are dominated by trees of the Dipterocarpaceae, distinguished by a unique reproductive pattern (Medway 1972, Appanah 1985, Curran et al. 1999), causing them to flower and fruit at intervals of 2-10 years with little or no reproductive activity in between (Medway 1972, Ashton et al. 1988, Curran and Leigthon 2000). A large proportion of individual trees covering varied taxonomic groups (Ap- panah 1985, Sakai et al. 1999) and flowering syndromes (Momose et al. 1998), flower and fruit synchronously with these dipterocarps. During these mast fruiting events, an overabundance of fruit is produced for a short period of time but these are often followed by periods of low fruit production. These large fluctuations in fruit availability (both temporal and spatial) can affect primary consumers in a variety of ways and in order to compensate for low fruit avail- 6 Chapter 1

ability, mammals have developed a variety of behavioural adaptations, ranging from local dietary shifts to long-distance migrations (e.g. Leighton and Leighton 1983, Curran and Leighton 2000). During these prolonged intermast periods a number of ‘keystone’ species, especially figs (Ficus spp.) still produce fruit (van Schaik et al. 1993). Janzen (1974) hypothesized that prolonged food scarcity in these dipterocarp forests can limit populations of frugivores and granivores during non-mast years, and on rare occasions food scarcity during these intermast periods may lead to starvation (Curran and Leighton 2000, Wong et al. 2005). Malayan sun bears (Helarctos malayanus), the smallest of the extant bear species, are currently the largest member of the order Carnivora on Borneo. Bornean sun bears are considered a separate subspecies (H. m. eurispylus), considerably smaller than their Sumatran and mainland counterparts, H. m. malayanus (Pocock 1941, Meijaard 2004). Sun bears occur throughout Southeast Asia, with the westernmost part of their distribution starting in eastern India, throughout Burma, possibly southern China (Yunnan), Laos, Cambodia, Vietnam, Thailand, Malaysia, Indonesia (Sumatra, Kalimantan), and Brunei (IUCN/BSG 2006) (Fig. 2). The species now occurs patchily through much of its former range, and has been extirpated from many areas, especially in mainland Southeast Asia. Sun bear fossils from the Pleistocene have been found much further north into China and on the island of Java (Erdbrink 1953), but sun bears did not occur there in more recent historical times (Meijaard 2004). When this study commenced in 1997, no field research had yet been undertaken on sun bears anywhere in their distribution area. Sun bears were listed as Data Deficient in the IUCN Red List and only rather crude distribution maps were available (Servheen 1999). Reli- able estimates of sun bear population numbers were lacking (and still are). In the available literature at the time, sun bears were described as nocturnal frugivores, augmenting their diet with a variety of insects, small mammals, and honey (Lekagul and McNeely 1977, Kunkun 1985, Payne et al. 1985). Sun bears rely on tropical forest habitat, with two ecologically distinct categories of tropical forest occurring within their range (Fredriksson et al. 2008). These two types of forests are distinguished by differences in climate, phenology, and floristic composition. Tropical evergreen rainforest is the sun bear’s main habitat in Borneo, Sumatra, and peninsular Malaysia. This aseasonal habitat receives high annual rainfall that is relatively evenly distributed throughout the year. Tropical evergreen rainforest, includes a wide diversity of forest types used by sun bears, including lowland dipterocarp, peat swamp, freshwater swamp, limestone/ karst hills, hill dipterocarp, and lower montane forest. In contrast, sun bears in mainland South- east Asia inhabit seasonal ecosystems with a long dry season (3–7 months), during which rainfall is <100 mm per month. Seasonal forest types are usually interspersed in a mosaic that General introduction 7

Figure 2: Sun bear distribution range (IUCN/BSG, Japan Bear Distribution Mapping Workshop, 2006). 8 Chapter 1

includes semi-evergreen, mixed deciduous, dry dipterocarp (<1,000 m elevation), and montane evergreen forest (>1,000 m). The range of sun bears overlaps that of Asiatic black bears (Ursus thibetanus) in this seasonal forest mosaic. Sun bears occur from near sea level to over 2,100 m elevation, but appear to be most common in lower-elevation forests (Steinmetz et al. 2011). This study commenced a few months before the onset of the catastrophic forest fires of 1997-98, in a small lowland reserve (10,000 ha), the Sungai Wain Protection Forest (SWPF) in East Kalimantan. At that time, this forest was still connected to a number of other forest areas to the west and north beyond the boundaries of the reserve, and although encroached by illegal settlers on the eastern border, the SWPF still covered some 8,000 ha of primary coastal lowland forest. A great opportunity for close behavioural observations of these other- wise extremely cryptic bears presented itself in January 1998, when we received the first of three young confiscated sun bears, with a request from the Conservation Department (Balai Konservasi Sumber Daya Alam) to rehabilitate and release them in the reserve. In March 1998 forest fires entered the reserve, towards the end of the long ENSO drought, initially from a neighbouring state-owned logging concession, but subsequently also from surrounding agricultural lands. Despite two months of extensive fire extinction efforts, approximately 50% of the reserve was affected by forest fires (Fredriksson 2002) leaving an unburned central core of some 4,000 ha of primary forest. Subsequent to these fires, a number of studies commenced in the SWPF to investigate the effects of fires on plant communities (van Nieuwstadt 2002, Slik et al. 2002, Slik and Eichhorn 2003, Slik 2004), butterflies (Cleary 2003), and birds (Slik and van Balen 2006). This study on sun bears spanned five continuous years in the field. Between August 1997 and July 2002 fieldwork was primarily carried out in SWPF. Subsequently sun bear sign surveys were carried out in different forest types throughout Kalimantan, covering the main forest types found on the island. Additional sign surveys of sun bears in burned and unburned forest were carried out in 2005 and 2010 in SWPF. Due to the lack of reference material, a broad approach was taken to gain insight into the ecology of this species, ranging from phenology plots to monitor production of flowers and fruit over several years, behavioural observations of sun bears, collection of fecal samples to look at dietary composition and seasonal changes herein, seed germination trials to investigate seed dispersal efficiency, trapping of wild sun bear for radio-tagging to gain insight into ranging and activity patterns, sampling of insects in burned and unburned forest, interviews with forest- edge communities to gain insight into sun bear activity in agricultural areas, as well as using sign transects to monitor sun bear usage of burned and unburned areas over time. General introduction 9

Outline of this thesis

Chapter 1 General introduction Chapter 2 describes frugivory in sun bears and how this is linked to periodic mast fruiting events in Kalimantan, monitored through monthly phenology censuses. We describe the importance and occurrence of various tree families in the diet of the sun bears over a period of 55 months. This chapter highlights the importance of the need for long-term monitoring in these aseasonal forests where supra-annual, rather than annual, changes in phenology affect sun bear ecology. Although sun bears were feeding primarily on fruits during the start of this study coinciding with a mast fruiting event, invertebrates dominated the diet during the subsequent prolonged intermast period, which lasted up to four years at this study site. Chapter 3 describes the movement and activity patterns of sun bears in unburned forest through telemetry studies. Sun bears within the reserve were found to be primarily diurnal, as opposed to sun bears living at the forest edge, which displayed more nocturnal activity. Female sun bears ranged over relatively small areas (approx. 5 km2) over a period of 12 months. The ranges of several female bears that we monitored were strictly confined within the small reserve. Sun bears seem to shun areas with human traffic, with bears avoiding the main trail leading to the research station. Sun bears also appeared to avoid entering the burned forest areas for several years after the fire event. This avoidance of areas with human disturbance in combination with relatively small ranges implies that even small reserves like SWPF are relevant for sun bear conservation, although the long-term survival of sun bear populations in such smaller protected areas will depend on the matrix of forest habitat and connectivity to other forest areas outside these reserves. Subsequent to the large fire event in 1997-98, the impacts of forest fires on sun bear fruit resources and invertebrates were studied in more detail. In Chapter 4 we describe the effects of drought and forest fires on density and diversity of fruit trees important in the diet of sun bears. The forests of Borneo are maladapted to withstand the impacts of prolonged drought and fires, and sun bear fruit trees experienced 80% mortality in areas affected by the fires. Drought also affected sun bear fruit trees in unburned forest areas, with elevated mortality rates. Species diversity of sun bear fruit taxa also declined by 44% in burned forest areas impacting future species composition in these once-burned areas. In Chapter 5 we look at the interface between humans and sun bears at the forest edge. Elevated levels of human-sun bear conflict were reported after the forest fires in 10 Chapter 1

1997-98. We discuss methods on how to reduce some of these conflicts and how human-bear conflict will affect conservation of this species. Chapter 6 describes the predation of an adult female sun bear, which we had radio- collared, by a python (Python reticulatus) during the prolonged intermast period. I discuss the physical condition of sun bears caught for telemetry studies at this study site compared to that of other sun bears caught for research purposes simultaneously in Sabah (Malaysian Borneo). I discuss how the reduced physical condition of the sun bear predated upon by the python might be linked to a combination of lack of fruit resources and increased competition in unburned forest by bears displaced after the fires. In Chapter 7 we present data from sign transects where we explore the usage of sun bear sign (i.e. claw marks, broken termite nests, ripped logs) as an index of relative abundance to monitor the usage of burned and unburned forest areas by sun bears over more than a decade. Sun bear sign densities/hectare were close to zero in burned forest areas for several years after the fire event, although sign densities increased slowly by the 7th years after the fires and were close to 65% of sign densities encountered in adjacent primary forest, 12 years after the fire event. These once-burned forest areas still have important regeneration and conservation potential as long as source populations are found nearby. Chapter 8 Synthesis and concluding remarks

References Achard, F., H. D. Eva, P. Mayaux, H.-J. r. Stibig, and A. Belward. (2004). Improved estimates of net carbon emissions from land cover change in the tropics for the 1990s. Global Biogeochem. Cycles 18:GB2008. Appanah, S. (1985). General flowering in the climax rain forests of South-east Asia. Journal of Tropical Ecol- ogy 1:225-240. Ashton, P.S., Givnish, T.J., and Appanah, S. (1988). Staggered Flowering in the Dipterocarpaceae - New In- sights into Floral Induction and the Evolution of Mast Fruiting in the Aseasonal Tropics. American Naturalist 132, 44-66. Azuma, S. (1988). Distribution and abundance of primates after the forest fire in the lowland forest of East Kalimantan 1983-1986. Pages 94-116 in H. Tagawa and N. Wirawan, editors. A research on the process of earlier recovering of tropical rain forest. Barlow, J. and C. Peres. (2006). Effects of Single and Recurrent Wildfires on Fruit Production and Large Ver- tebrate Abundance in a Central Amazonian Forest. Biodiversity and Conservation 15:985-1012. General introduction 11

Berenstain, L. (1986). Responses of Long-tailed macaques to drought and fire in eastern Borneo: A preliminary report. Biotropica 18:257-262. Brook, B. W., N. S. Sodhi, and P. K. L. Ng. (2003). Catastrophic extinctions follow deforestation in Singapore. Nature 424:420-423. Cardillo, M., A. Purvis, W. Sechrest, J. L. Gittleman, J. Bielby, and G. M. Mace. (2004). Human population density and extinction risk in the world’s carnivores. PLoS Biology 2:0909-0914. Cochrane, M. A., A. Alencar, M. D. Schulze, C. M. Souza, D. C. Nepstad, P. Lefebvre, and E. A. Davidson. (1999). Positive feedbacks in the fire dynamic of closed canopy tropical forests. Science 284:1834-1836. Curran, L. M., I. Caniago, G. D. Paoli, D. Astianti, M. Kusenti, C. E. Nirarita, and H. Haeruman. (1999). Impact of El Nino and logging on canopy tree recruitment in Borneo. Science 286:2184-2188. Curran, L. M. and M. Leighton. (2000). Vertebrate responses to spatiotemporal variation in seed production of mast-fruiting dipterocarpaceae. Ecological Monographs 70:101-128. Curran, L.M., Trigg, S.N., McDonald, A.K., Astianti, D., Hardiono, Y.M., Siregar, P., Caniago, I., and Kasischke, E. (2004). Lowland forest loss in protected areas of Indonesian Borneo. Science 303, 1000-1003. Doi, T. (1988). Present status of the large mammals in the Kutai National Park, after a large scale fire in East Kalimantan, Indonesia. Pages 82-93 in H. Tagawa and N. Wirawan, editors. A research on the process of earlier recovery of tropical rain forest. Erdbrink, D. P. (1953). A review of fossil and recent bears of the old world. With remarks on their phylogeny based upon their dentition. PhD. University of Utrecht, Drukkerij Jan De Lange, Deventer. Fredriksson, G.M. (2002). Extinguishing the 1998 forest fires and subsequent coal fires in the Sungai Wain Protection Forest, East Kalimantan, Indonesia. In: Communities in flames: proceedings of an inter-

national conference on community involvement in fire management (Bangkok, Thailand, FAO and FireFight SE Asia). Fredriksson, G., Steinmetz, R., Wong, S. and Garshelis, D.L. (IUCN SSC Bear Specialist Group) (2008). Helarctos malayanus. In: IUCN 2012. IUCN Red List of Threatened Species. Version 2012.1. . Downloaded on 20 July 2012 Fuller, D.O., Jessup, T.C., and Salim, A. (2004). Loss of forest cover in Kalimantan, Indonesia, since the 1997- 1998 El Niño. Conservation Biology 18, 249-254. Goldammer, J. G. and B. Seibert. (1989). Natural rain forest fires in Eastern Borneo during the Pleistocene and Holocene. Naturwissenschaften 76:518-520. Goldammer, J. G. (2007). History of equatorial vegetation fires and fire research in Southeast Asia before the 1997–98 episode: A reconstruction of creeping environmental changes. Mitigation and Adapta- tion Strategies for Global Change 12:13-32. 12 Chapter 1

Guhardja, E., M. Fatawi, M. Sutisna, T. Mori, and S. Ohta, editors. (2000). Rainforest Ecosystems of East Kali- mantan. El Nino, Drought, Fire and Human Impacts. Springer-Verlag, Tokyo. Hansen, M.C., Stehman, S.V., Potapov, P.V., Arunarwati, B., Stolle, F., and Pittman, K. (2009). Quantifying changes in the rates of forest clearing in Indonesia from 1990 to 2005 using remotely sensed data sets. Environmental Research Letters 4, 034001. Hoffmann, A.A., Hinrichs, A., and Siegert, F. (1999). Fire damage in East Kalimantan in 1997/98 related to land use and vegetation classes: Satellite radar inventory results and proposal for further actions. IFFM-SFMP Report No.1a (Samarinda, East Kalimantan, MOFEC, GTZ and KfW). Janzen, D. H. (1974). Tropical blackwater rivers, animals, and mast fruiting by the dipterocarpaceae. Bio- tropica 6:69-103. Kinnaird, M.F., and O’Brien, T.G. (1998). Ecological effects of wildfire on lowland rainforest in Sumatra. Conservation Biology 12, 954-956. Kunkun, J. G. (1985).Short information on Malayan sun bear in Indonesia.Pages 67-69 in First Asiatic Bear Conference, Utsunomiya & Nikko, Tochigi Prefecture. Langner, A. and F. Siegert. (2009). Spatiotemporal fire occurrence in Borneo over a period of 10 years. Global Change Biology 15:48-62. Lee, T. M. and W. Jetz. (2008). Future battlegrounds for conservation under global change. Proceedings of the Royal Society B: Biological Sciences 275:1261-1270. Leighton, M. and D. R. Leighton. (1983). Vertebrate responses to fruiting seasonality within a Bornean rain forest. Pages 181-196 in S. L. Sutton, T. C. Whitemore, and A. C. Chadwick, editors. Tropical Rain Forest Ecology and Management. Blackwell Scientific Publications, Cambridge UK.

Leighton, M. and N. Wirawan. (1986). Catastrophic drought and fire in Borneo Tropical Rain Forest Associ- ated with the 1982-1983 El Nino southern oscillation event. Pages 75-101 in G. T. Prance, editor. Tropical rain forest and world atmosphere. Westview Press, Boulder, Colorado, USA. Lekagul, M. D. B. and B. A. J. A. McNeely. (1977). Mammals of Thailand. Kurusapha Ladprao Press, Thailand. Lennertz, R. and K. F. Panzer. (1984). Preliminary assessment of the drought and forest fire damage in Kalimantan Timur. DFS German Forest Inventory Service for Gesellschaft fur Technische Zusam- menarbeit (GTZ), Germany. Malingreau, J. P., G. Stephens, and L. Fellows. (1985). Remote sensing of forest fires: Kalimantan and North Borneo in 1982-83. Ambio 14:314-321. Medway, L. (1972). Phenology of a tropical rain forest in Malaya. Biological Journal of the Linnean Society 4:117-146. Meijaard, E. (2004). Craniometric differences among Malayan sun bears (Ursus malayanus); evolutionary and taxonomic implications. Raffles Bulletin of Zoology 52, 665-672. General introduction 13

Meijaard, E., Sheil, D., Rosenbaum, B., Iskandar, D., Augeri, D., Setyawati, T., Duckworth, W., Lammertink, M.J., Rachmatika, I., Nasi, R., et al. (2005). Life after logging: reconciling wildlife conservation and production forestry in Indonesian Borneo (Bogor, Indonesia, CIFOR, WCS and UNESCO). Miettinen, J., Shi, C., and Liew, S.C. (2011). Deforestation rates in insular Southeast Asia between 2000 and 2010. Global Change Biology 17, 2261-2270. Myers, N., R. A. Mittermeier, C. G. Mittermeier, G. A. B. da Fonseca, and J. Kent. (2000). Biodiversity hotspots for conservation priorities. Nature 403:853-858. Nieuwstadt, M.G.L. v. (2002). Trial by fire-Postfire development of a dipterocarp forest (PhD thesis Univer- sity of Utrecht, Netherlands), pp. 1-143. O’Brien, T. G., M. F. Kinnaird, A. Nurcahyo, M. Prasetyaningrum, and M. Iqbal. (2003). Fire, demography and the persistence of siamang (Symphalangus syndactylus: Hylobatidae) in a Sumatran rainforest. Animal Conservation 6:115-121. Page, S. E., F. Siegert, J. O. Rieley, H. D. V. Boehm, A. Jaya, and S. Limin. (2002). The amount of carbon released from peat and forest fires in Indonesia during 1997. Nature 420:61-65. Payne, J., C. M. Francis, and K. Phillipps. (1985). A field guide to the mammals of Borneo. World Wildlife Fund Malaysia, Kuala Lumpur, Malaysia. Peres, C. A. (1999). Ground fires as agents of mortality in a Central Amazonian forest. Journal of Tropical Ecology 15:535-541. Pocock, R. I. (1941). The Fauna of British India, including Ceylon and Burma. Mammalia Vol. II. Carnivora (suborders Aeluroidae (part) and Arctoidae).Taylor & Francis, Ltd., London, United Kingdom. Posa, M. R. C., A. C. Diesmos, N. S. Sodhi, and T. M. Brooks. (2008). Hope for Threatened Tropical Biodiver-

sity: Lessons from the Philippines. BioScience 58:231-240. Sakai, S., K. Momose, T. Yumoto, T. Nagamitsu, H. Nagamasu, A. A. Hamid, and T. Nakashizuka. (1999). Plant reproductive phenology over four years including an episode of general flowering in a lowland dipterocarp forest, Sarawak, Malaysia. American Journal of Botany 86:1414-1436. Schaik, C. P. v., J. W. Terborgh, and S. J. Wright. (1993). The phenology of tropical forests: adaptive significance and consequences for primary consumers. Annual Review of Ecology and Systematics 24:353-377. Servheen, C. (1999). Status of Bears of the World. Presentation at the 3rd International Symposium on the Trade of Bears and Bear Parts, Seoul, Korea. Siegert, F., G. Ruecker, A. Hinrichs, and A. A. Hoffmann. (2001). Increased damage from fires in logged forests during droughts caused by El Niño. Nature 414:437-440. Slik, J. W. F. (2004). El Niño droughts and their effects on tree species composition and diversity in tropical rain forests. Oecologia 141:114-120. 14 Chapter 1

Slik, J.W.F., Verburg, R.W., and Kessler, P.J.A. (2002). Effects of ﬁre and selective logging on the tree species composition of lowland dipterocarp forest in East Kalimantan, Indonesia. Biodiversity & Conser- vation 11, 85-98. Slik, J. W. F. and K. A. O. Eichhorn. (2003). Fire survival of lowland tropical rain forest trees in relation to stem diameter and topographic position. Oecologia 137:446-455. Slik, J. and S. Van Balen. (2006). Bird Community Changes in Response to Single and Repeated Fires in a Lowland Tropical Rainforest of Eastern Borneo. Biodiversity and Conservation 15:4425-4451. Slik, J., F. Breman, C. Bernard, M. van Beek, C. Cannon, K. Eichhorn, and K. Sidiyasa. (2010). Fire as a selective force in a Bornean tropical everwet forest. Oecologia 164:841-849. Sodhi, N. S., L. P. Koh, B. W. Brook, and P. K. L. Ng. (2004). Southeast Asian biodiversity: an impending disaster. Trends in Ecology & Evolution 19:654-660. Sodhi, N. and B. Brook. (2006). Southeast Asian biodiversity crisis. Cambridge University Press, Cambridge, UK. Sodhi, N., Posa, M., Lee, T., Bickford, D., Koh, L., and Brook, B. (2010). The state and conservation of South- east Asian biodiversity. Biodiversity and Conservation 19, 317-328. Steinmetz, R., D. L. Garshelis, W. Chutipong, and N. Seuaturien. (2011). The Shared Preference Niche of Sympatric Asiatic Black Bears and Sun Bears in a Tropical Forest Mosaic. PLoS One 6:e14509. Terborgh, J. (1988). The big things that run the world - a sequel to E.O.Wilson. Conservation Biology 2:402-403. Terborgh, J. (1992). Maintenance of diversity in tropical forests. Biotropica 24(2b):283-292. Whelan, R. J. (1995). The ecology of ﬁre. Cambridge University Press, Cambridge.

Wich, S. A. and C. P. van Schaik. (2000). The impact of El Niño on mast fruiting in Sumatra and elsewhere in Malesia. Journal of Tropical Ecology 16:563-577. Wong, S.T., Servheen, C., Ambu, L., and Norhayati, A. (2005). Impacts of fruit production cycles on Malayan sun bears and bearded pigs in lowland tropical forest of Sabah, Malaysian Borneo. Journal of Tropi- cal Ecology 21, 627–639. Woods, P. (1989). Effects of Logging, Drought, and Fire on Structure and Composition of Tropical Forests in Sabah, Malaysia. Biotropica 21:290-298. Chapter 2

Frugivory in sun bears (Helarctos malayanus) is linked to El Niño-related fluctuations in fruiting phenology, East Kalimantan, Indonesia

With S.A. Wich and Trisno

Biological Journal of the Linnean Society (2006), 89(3): 489-508 18 Chapter 2

Abstract

We investigated sun bear (Helarctos malayanus) frugivory and fruiting phenology in a lowland dipterocarp forest in East Kalimantan, Indonesia. Two mast fruiting events, both coinciding with El Niño/Southern Oscillation events, occurred four years apart, resulting in large ﬂuctuations in fruit availability. Sun bear fruit availability decreased from 13 trees ha-1 fruiting month-1 during the mast fruiting to 1.6 during the inter-mast period. Almost 100% of sun bear diet consisted of fruit during mast fruiting period, whereas during inter-mast periods sun bear diet was predominantly insectivorous. The majority of sun bear fruit trees displayed ‘mast-fruiting’ and ‘supra-annual’ fruiting patterns, indicating sporadic productivity. Sun bears fed on 115 fruit species covering 54 genera and 30 families, with Ficus (Moraceae) being the main fallback fruit. The families Moraceae, Burseraceae and Myrtaceae contributed more than 50% to the sun bear fruit diet. Sun bear fruit feeding observations were unevenly distributed over forest types with more observations in high-dry forest type despite fewer fruiting events, possibly a side-effect of high insect abundance which causes bears to use these areas more intensively. We discuss the possible evolutionary pathways of sun bears in relation to the Sundaic environment. Phenological inﬂuences on sun bear frugivory 19

Introduction

Irregular community flowering and fruiting have been recorded throughout Southeast Asian forests (Wood, 1956; Medway, 1972; Appanah, 1985; Ashton et al., 1988; Sakai, 2002). The canopy and emergent trees of these forests are dominated by trees of the Dipterocarpaceae, distinguished by a unique reproductive pattern (Medway, 1972; Appanah, 1985; Curran et al., 1999), causing them to flower and fruit at intervals of 2-10 years with little or no reproductive activity in between (Medway, 1972; Ashton et al., 1988; Curran and Leigthon, 2000). A large proportion of individual trees covering varied taxonomic groups (Appanah, 1985; Sakai et al., 1999) and flowering syndromes (Momose et al., 1998), flower and fruit synchronously with these dipterocarps. The size of the area taking part in such a general flowering/mast fruiting event and intensity can vary substantially, from a river valley to the whole of Peninsular Malaysia (Ashton et al., 1988; Yasuda et al., 1999). It is unclear which mechanism triggers these synchronized flowering events, although a drop in night-time temperature (≥2ºC for 3 or more consecutive days) some 2 months before the onset of flowering, has received substantial support (Ashton et al., 1988; Yasuda et al., 1999, but see Corlett and Frankie, 1998). In Borneo, temperature drops associated with the El Niño/ Southern Oscillation (ENSO) phenomenon have preceded mast events (Ashton et al., 1988), although Yasuda et al. (1999) reported a night-time temperature drop preceding a mast fruiting event during a La Niña episode in Peninsular Malaysia. The influence of ENSO on mast fruiting varies throughout Southeast Asia (Wich and van Schaik, 2000), but is reportedly strongest on the eastern sides of the Malesian land masses, especially in East Kalimantan (Ashton et al., 1988). Mast fruiting events provide an overabundance of fruit for a short period of time and are often followed by periods of low fruit production, probably due to exhaustion of energy reserves (van Schaik 1986; Sork et al., 1993). The temporal and spatial patchiness and high variability in fruit production in Southeast Asia cause these forests to have a much lower overall productivity than tropical forests in Africa and South America (Appanah, 1985). Janzen (1974) hypothesized that prolonged fruit scarcity in these dipterocarp forests limits populations of granivores and frugivores during non-mast years. These primary consumers have adapted in a variety of ways to the lack of fruit. Many mammals show a dietary switch and resort to feeding on items of lower nutritional value (Leighton and Leighton, 1983; Peres, 1994; Knott, 1998), materials with increased handling time (e.g., palm-nuts, Strushaker and Leyland, 1977), plants that contain chemical deterrents (Foster, 1977), or diffusely scattered resources that increase 20 Chapter 2

foraging time (Terborgh, 1983). The synchronized timing of reproduction in relation to periods of fruit abundance (Fogden, 1972; van Schaik and van Noordwijk, 1985; Goldizen et al., 1988), changes in range use (e.g. Joshi et al., 1995), and migration (Caldecott, 1988; Curran and Leighton, 2000) have also been recorded for certain mammal species. Only rarely has food scarcity between mast events been so severe that it leads to documented famine and starvation. Curran and Leighton (2000) observed emaciated and starving bearded pigs at their study site in West Kalimantan during inter-mast periods and Knott (1998) found that orangutans suffered negative energy budgets during the lean fruit period after a mast at the same study site. A variety of plant species that produce edible parts during periods of low fruit availability have been labelled fallback resources (e.g. Conklin-Brittain et al., 1998). Fallback resources are critical for metabolic maintenance when preferred foods are unavailable (Leighton and Leighton, 1983; Terborgh, 1986) and their abundance potentially sets the carrying capacity for primary consumers (van Schaik et al., 1993). The magnitude and duration of the period of fruit scarcity, the availability of fallback foods, and the physiological and behavioural ﬂexibilities of animal species all play an important role in determining how successfully animals adapt to lean periods. Figs have been reported to be the main fallback food item during fruit lows for many frugivorous species in East Kalimantan (Leighton, 1993), as well in Neotropical sites (Terborgh, 1986). The sun bear (Helarctos malayanus), one of the largest mammals of the Bornean rainforest, has mainly been described as a frugivore, augmenting its diet with a variety of insects, small mammals and honey (Lekagul and McNeely, 1977; Kunkun, 1985; Wong et al., 2002). They are effective dispersers of the seeds of several plant species (McConkey and Galetti, 1999; G.M. Fredriksson, unpubl. data). We chose the sun bear as a model large frugivore faced with annual and supra-annual variations in fruit availability in East Kalimantan, where effects of ENSO on mast fruiting events are strong (Ashton et al., 1988; Wich and van Schaik, 2000). This study was carried out at the eastern-most limit of the sun bear distribution range, which covers most of tropical mainland Southeast Asia, from the Eastern tip of India, through Myanmar, Laos, Thailand, Cambodia, Vietnam, and Malaysia to the islands of Sumatra and Borneo (Servheen, 1999). The aims of this study were: i) to study the extent of frugivory in sun bear diet; ii) to investigate the temporal variations in fruit availability in East Kalimantan; iii) to investigate fruiting patterns of fruit species important in the diet of the sun bear and distribution of these over topographical types; and iv) to determine which fruit taxa are important as fallback resources for sun bears during inter-mast periods. Phenological inﬂuences on sun bear frugivory 21

Materials and methods

Study site and climate data The study was carried out in a lowland dipterocarp forest, the Sungai Wain Protection Forest, East Kalimantan, Indonesian Borneo (1º 05’ S and 116º 49’ E) (Fig. 1). The reserve covers a watercatchment area of circa 10,000 ha. Approximately 50% of the reserve was affected by forest ﬁres in March-April 1998 (Fredriksson, 2002), during one of the most severe ENSO related droughts ever recorded (McPhaden, 1999), leaving an unburned central core of some 4000 ha of primary forest. Data for this study were collected in the unburned primary forest. The topography of the reserve consists of gentle to sometimes steep hills, and is intersected by many small rivers. Elevations range from 30 to 150 metres a.s.l. The most common families (diameter at breast height, dbh, >10 cm), are Euphorbiaceae, Dipterocarpaceae, Sapotaceae, and Myrtaceae. The relative dominance of Dipterocarpaceae increases substantially in larger size classes. Together, the 25 most common species form 40% of total stem density (van Nieuwstadt, 2002). Palmae (also called Family Arecaceae) are common in the subcanopy and understory, and Zingiberaceae, Marantaceae, Araceae, and Pandanaceae are widespread in the understory. Daily rainfall, minimum-maximum temperature and humidity data were collected between January 1998 and October 2002, at ground level in the primary forest. Rainfall patterns were calculated on a yearly basis, using Mohr’s Index (Whitmore, 1985). At the start of the study rainfall was below average due to the ENSO event that was in progress (van Nieu- wstadt 2002). Although rainfall patterns in the region are not consistent on an annual basis, a distinction could be made between a ‘wet’ season (Nov-April) and a ‘dry’ season (May-Oct) (t-test, t=4.96, df=6, p=0.002). Out of 58 months that rainfall data were collected 12 months had rainfall below 100 mm, classiﬁed as dry months (Walsh and Newbery, 1999). Nine of these occurred between May-September, the remaining 3 dry months were during the 1998 ENSO (Jan-March 1998). Rainfall type falls within the category of ‘slightly seasonal’ according to Mohr’s index, with average annual rainfall 2968 ± 510 mm y-1. Temperature remained stable throughout the study period with average maximum of 29.7 ± 0.7ºC and minimum 23.2 ± 0.4ºC, with an annual average of 26.5 ± 0.2 ºC, equal to the average temperature of 26.5ºC reported for the region historically (Berlage, 1949). 22 Chapter 2

Figure 1. Map of the Sungai Wain study site in East Kalimantan, showing the unburnt forest core where data were collected.

Classification of topographical types Five topographical types were discerned: swamp, alluvial, slope, high-flat, and ridge (Fredriksson and Nijman, 2004). Swamp type was assigned to areas that showed signs of regular inundation (pneumatophore roots) or were permanently inundated, and had a high density of rattans, small trees and saplings, and a large numbers of climbers. This classification of swamp forest refers to swampy patches within the drier forest matrix and has no relation to large areas of peat or fresh-water swamp forest as found in e.g. south-central Borneo. Alluvial type was allocated to flat, non-inundated areas close to rivers characterized by large trees. High-flat type was assigned to flat areas at higher elevated sites away from rivers, usually with large trees. Slope type was assigned when an area had a relatively steep inclination (>12°). Ridge type was allocated to narrow hill tops with steep sides, or crests of longer hill chains. The relative avail- Phenological influences on sun bear frugivory 23

ability of these forest types was determined by means of transects (total length 18.5 km) laid out in an east-west and north-south direction where every 25 m the topographical type was recorded. Slope topographical type covered 45.3 ± 9.1 % of the study area, followed by high-ﬂat (24.8 ± 9.3 %), alluvial (17.3 ± 4.0 %), ridge (6.7 ± 2.7 %) and swamp (5.9 ± 4.6 %).

Fruiting phenology data collection Two phenology data sets were collected during this study, using the same methodology. Phenology studies commenced in January 1998 and continued until July 2002. Data were collected in all but two of 55 months (March ’98 and February ’02). i) Phenology plots Ten 0.1-ha phenology plots (100 x 10 m), were established with two plots located in each of the five topographical types. All trees ≥10 cm dbh were measured, labelled, and identified by botanists from the Wanariset Herbarium. The total number of live trees in the 1-ha of plots at the start of the study was 549 individuals covering some 186 species. Every month around the same date (third week), all trees in these plots were observed with 10 x 40 Leica binoculars. The relative abundance of flowers, young and mature fruits, and young and old leaves, in relation to size of the crown, was estimated, though for analysis here only presence/absence of fruit is considered. ii) Specific sun bear fruit trees The second data set contained 104 trees of 11 species (≥10 cm dbh) known or thought to be important in the diet of the sun bear. The species of trees in this sample were based on primary fruits found in sun bear scats and trees frequently covered with sun bear claw marks (signs of bears climbing the trees), as well as information from local people. Individual trees of these species were encountered by chance in the forest, measured, labelled, and reobserved on a monthly basis. The following taxa were included: Monocarpia kalimantanensis (n=10) (Annonaceae), Durio dulcis (n=10), D. oxleyanus (n=10), and D. zibethinus (n=2) (Bombacaceae), Dacryodes rugosa (n=11), D. rostrata (n=10), and Santiria spp. (n=10) (Burseraceae), Ficus spp. (n=11), and Artocarpus integer (n=10) (Moraceae), Ochanostachys amentaceae (n=10) (Olacaceae), and Tetramerista glabra (n=10) (Tetramerista- ceae). Three of the 11 taxa were added later during the study period, when it was found that bears fed substantially on them (Monocarpia June 1999; Ficus spp. June 1999; Santiria spp. June 2000).

Frugivory in sun bears The study on the extent of frugivory in sun bears was based on faecal samples and direct feeding observations. Faeces were systematically collected between October 1997 and 24 Chapter 2

July 2002 from both wild and reintroduced bears, with sun bear scats being clearly identifiable due to their scent, texture and size. Collection occurred along transects and trails that intersected the home ranges of three wild radiocollared female sun bears, monitored between 1999 and 2001, and at specific sites where these bears were located by triangulation. Randomly located scats from unmarked wild bears were collected as well throughout the forest whenever encountered. Scats from three captive-reared bears that were radiocollared and gradually reintroduced into the forest were also collected. Scats from these bears were included only after they had been acclimated to living in the forest for at least 6 months. All faecal samples were weighed and washed through a 0.2 mm mesh-sized sieve and dried in the sun. Identification of seeds, and seedlings after germination trials, were done by the Wanariset Herbarium. Scat contents were sorted into a five main categories i) fruit (whole or broken seed counts), ii) flowers, iii) insects, iv) pith (meristem), v) debris (termite nest remains; wood; resin). A visual estimate was made of the relative volume of each category. To calculate the frequency of occurrence of fruit in scats for each month, the running mean was calculated for three consecutive months. This was done as fruit scats collected in a given month could be up to several weeks old when collected, representing fruits eaten that month or the month before. Secondly, faecal data collected late in the month, might be more comparative with the fruit situation for the following month, as phenology data was collected towards the end of each month. Faecal samples were collected in all but five out of 57 months of scat collection. Direct feeding observations were obtained on a regular basis from reintroduced bears as well as ad libitum from wild sun bears. Reintroduced bears were habituated for visual observations and twice a week observations of each bear were carried out. Data collection from observations was continuous, and time of initiation and termination of each activity recorded. Bears were observed on average 7.0 ± 1.5 hours during observation days (n=711), approximately 50% of their active hours in a day (G.M. Fredriksson, unpubl. data). During foraging observations we recorded the total time handling a food source (e.g. foraging in or around a fruiting tree, or handling a termite nest), as well as actual time spent feeding on that food source (how much time spent actually swallowing food), and which part of a food item was eaten (e.g. pulp, seeds, exocarp). With larger food items (e.g. large fruits), the exact number of items consumed was recorded. Samples were collected of all food items for subsequent identification. Data from the initial 6-8 months of observations were not used for data analyses here, because during this time released bears still received supplementary foods. The period of direct observations used for analysis spanned 31 months from October 1998–April 2001. Phenological influences on sun bear frugivory 25

Classification of fruiting types and spatial distribution of sun bear fruit species The flowering and fruiting patterns of all trees combined, and separate species/genera, were classified according to timing and frequency of flowering and fruiting events. The ‘mast fruiting’ period was defined as the period where fruit production was 1.96 standard deviations above the mean, following the definition for mast fruiting proposed by Wich and van Schaik (2000). When all fruiting events of a species occurred during this mast fruiting period, the species was categorized as a ‘mast-species’. When fruiting frequency was >1 /yr the species would be categorized as ‘sub-annual’, when it fruited once a year it was classified as ‘annual’, and when fruiting occurred less than once a year, but not during mast fruiting periods, it was classified as ‘supra-annual’ (following Sakai et al., 1999). When a species included individuals of more than one fruiting type, we assigned the fruiting type of the majority of individuals as the fruiting type for this species. We use the term ‘fruiting event’ to describe a tree with fruit during one monthly observation period or several consecutive observation periods. The term ‘sun bear food tree’ refers to trees from species that were found to occur in the diet of the sun bear. The distribution of sun bear fruit species among topographical types was determined by classifying all trees (≥ 10 cm dbh) from the ten 0.1 ha phenology plots into sun bear food and non-food species, and calculating the percentage of sun bear food trees for each topographical type.

Data analysis Two analyses were performed to evaluate the temporal pattern of fruiting events. First an index of aggregation, Morisita’s Index, Id, (Morisita, 1962), independent of sample size, was calculated based on temporal distribution of fruiting events for 17 three-month periods (sensu

Sakai et al., 1999) from January 1998 to July 2002. Id is near 1 in distributions that are essen- tially Poisson, >1 in clumped distributions, and <1 in cases of regular or seasonal reproduction

(Morisita, 1962; Krebs, 1999). Second, a χ2 test for goodness of fit was performed to examine whether the observed distribution of fruiting events significantly deviated from random distribution. This test was performed assuming that fruiting events occurred at random throughout the 17 three-month periods (when sample size of fruiting events was ≥70) or 8 six-month periods (when sample size of fruiting events was between 35-69) as the expected number of fruiting events in a unit period (3 or 6 mo) must be five or more (Sokal and Rohlf, 1981). These two analyses were conducted on separate genera important in the diet of the sun bear, as well as for all trees monitored for their monthly phenology combined. 26 Chapter 2

To test for differences in number of trees fruiting between periods, the study period was divided into four periods of 12 months starting May ’98, after the ﬁrst mast-fruiting event. Kruskal-Wallis one-way analyses of variance and Chi-squared tests were applied to test for differences between periods, and habitat types with signiﬁcance levels set at p=0.05. When multiple comparisons between treatments were conducted (Siegel and Castellan, 1988), Bonferroni techniques were applied to limit overall experiment-wise error (Sokal and Rohlf, 1981). All values reported are mean ± 1 standard deviation (SD) and statistical analyses were performed with Sigma Stat 1.0.

Results

Temporal variation in fruiting phenology Mean percentage of all trees fruiting in the ten 0.1-ha plots over the 53 months (Jan ’98- July ’02) was 2.4 ± 3.4 %; masting was thus defined as >9.1% of trees fruiting (2.4 + [1.96*3.4]) Two distinct mast-fruiting peaks lasting 2 months each were discerned, with the first in progress at the start of data collection in January-February 1998, and a second peak almost four years later, during November-December 2001 (Fig. 2). Both masting events took place during ENSO episodes (NOAA-CIRES, 2003). During the first masting event, 15.7 % of trees (86 of 549) fruited, and during the second 10.3 % of trees (49 of 474) carried fruit. During the inter-mast period, fruit production av- eraged 1.6 ± 0.6 % of trees fruiting ha-1 month-1. Fruit production differed significantly between years (years 1-4 in Fig. 2A), following the 1998 masting event (Kruskal-Wallis, H=18.1 df=3, p<0.001). Fruit production was significantly lower in the first two years (April 1998-March 2000) than in the fourth year (April 2001-March 2002) (Dunn-Šidák, Q>3.6, p<0.05 for both years). In the ten 0.1-ha plots, 26.2% of trees (144 of 549) were species encountered in the diet of sun bears, although only 37% (53 of 144) of these trees were reproductively active during the study, representing 9.7% of all trees (≥10 cm dbh) found in 1 ha. From these 53 reproductively active sun bear fruit trees, 128 fruiting events were recorded over 53 months of monitoring, with 52 events (40.6%) during the two masting events (4 months). During mast fruiting months, 13.0 ± 2.2 sun bear fruit trees ha-1 month-1 fruited, whereas during the inter- mast interval only 1.6 ± 1.8 sun bear fruit trees ha-1 month-1 produced fruit. In the first two years after the 1998 masting event fruiting by sun bear tree species was very low with only Phenological influences on sun bear frugivory 27

Figure 2. A. Temporal changes in fruiting phenology of trees (≥10 cm dbh, n=549) from ten 0.1-ha plots, showing two mast fruiting events. B. Fruiting phenology of selected sun bear fruit trees (n=104, data set II), mast and non-mast species combined. Horizontal line indicates the threshold above which masting is deﬁned to occur.

0.6 ± 0.8 sun bear fruit trees ha-1 month-1 producing fruit. In the third year fruiting increased to 2.5 ± 2.1 sun bear fruit trees ha-1 month-1 producing fruit. Fruiting by selected sun bear fruit trees (second phenology data set) also differed sig- niﬁcantly over years (Kruskal-Wallis, H=14.8, df=3, p=0.002), with three years after the 1998 masting event having lower fruit production than the fourth year (Dunn-Šidák, Q>2.8, p<0.05 28 Chapter 2

for all years) (years 1-4 in Fig. 2B). Fruit production was almost absent in the first year following the 1998 mast fruiting event. Mean percentage of sun bear trees fruiting, over the 53 months was 8.9 ± 10.2 %; hence, mast-fruiting for these was defined as when >29 % of trees under observation were fruiting. Two distinct mast-fruiting peaks for these selected sun bear trees coincided with the two mast fruiting events among all trees (Fig. 2). Three times during the study period did temperature drop below the ‘critical’ value of 21ºC, proposed as the low night-time temperature threshold triggering the onset of general flowering (Ashton et al., 1988). Each of these minima lasted only one night and lowest temperature recorded was 20.7 ºC (July ’02). These minor drops in temperature did not correlate to a subsequent onset of general-flowering. The temperature drop in July 2002 occurred when a smaller general flowering event was already well under way.

Fruiting patterns and types In total, 639 fruiting events were recorded from all trees (n=549) in the ten 0.1 ha plots (Table 1). More than half of these fruiting events (330 of 639; 51.6%) were concentrated during the mast-fruiting periods. Fruiting events were signiﬁcantly clustered over 3-month periods

2 (Table 1) (Id=2.52, χ =946.9, df=16, p<0.001). A high proportion of trees (61.4%) were never observed to fruit during the study period (Table 1). The most common fruiting type in the plots was mast-fruiting (16.4% of trees), followed by supra-annual fruiting pattern (14.4% of trees). Only 2 % of trees displayed an annual fruiting pattern and 5.8% of individuals a sub-annual pattern. Among 104 sun bear trees that we monitored, 378 fruiting events were recorded.

2 These were also found to be clustered over 3-month periods (Id=1.77, χ =363.6, df=16, p<0.001). Several species/genera displayed highly clumped fruiting patterns (Durio spp., Artocar- pus integer, Dacryodes spp., Ochanostachys amentaceae) (Table 1). Intermediate fruiting patterns were observed for Tetramerista glabra and Santiria spp. The former species had signiﬁcantly lower number of trees fruiting though in the ﬁrst 18 months after the 1998 masting and ENSO event than in the subsequent 18 months (Mann-Whitney U Rank Sum test: Mean rank 1=14.3, Mean rank 2=22.7, n=36, Z=-2.614, p=0.017). The lowest index of aggregation (most random fruiting) was found in Ficus spp. Mono- carpia kalimantanensis had the second lowest index of aggregation, but fruiting was not ran-

2 domly distributed (Id=1.07, χ =19.72, df=11, p<0.05) (Table 1). Mean dbh of selected sun bear fruit trees was larger (44.7 ± 21.1 cm) than the mean dbh of all monitored trees in the 0.1-ha plots (21.1 ± 14.3 cm) (Mann-Whitney U Rank Sum test: Mean rank1=501, Mean rank 2=285, n=617, Z=9.4, p<0.001). Still, 31.7% of these selected sun bear fruit trees were never observed Phenological inﬂuences on sun bear frugivory 29 8 1 1 3 2 4 1 1 4 1 1 8 Fruiting pattern 9 5 13.217.534.6 18.8 12.8 0.7 10.6 2.5 4.2 2.9 6.4 20.2 10b 12 Non-fruiting Mast-Fruiting Supra-annual Annual Sub-annual 2 χ ) d Morisita index ( I events) N (fruiting N (trees) N 5353 54953 144 63952 405 128 2.52 102 511 2.35 *** 378 2.55 *** 61.4a *** 1.77 63.1 *** 60.8 16.4 31.7 14.4 2 5.8 5252 10 21 44 44 5.69 6.72 *** *** 11 5238 22 11 32 29 4.69 0.99 - - 10 2 37 10 118 1.07 * 1 53 10 9 5.19 - 4 25 10 8 2.32 - 5 52 10 94 2.06 *** (months) Artocarpus integer rugosa Dacryodes and D. rostrata D. oxleyanus Durio dulcis and D. Ficus spp. Monocarpia kalimantanensis Ochanostachys amentaceae Santiria spp. Tetramerista glabra Tetramerista 1) Phenology plots (1 ha) combined: all trees species Bear food species Non-bear food 2) Selected fruit trees (second data set): Table 1. Morisita index and fruiting patterns main sun bear fruit the 1 ha phenology plots ( ≥ 10 cm DBH) and for for genera/species. Table b=no of trees a=% of trees; *** indicates p<0.001; * indicates p=0.05; ‘-’ = too small sample size; 30 Chapter 2

fruiting. Based on the majority of observations, taxa were divided into those only found fruiting during masting events (mast species: Durio spp., A. integer, Dacryodes spp., and O. amentaceae) and those that also fruit during inter-mast periods (non-mast species: Ficus spp., Santiria spp., T. glabra, and M. kalimantanensis) (Table 1).

Faecal analyses A total of 1209 sun bear scats were collected (77 between Oct 1997-April 1998, 280 in the first year after the 1998 mast (May 1998-April 1999), 305 in second year, 391 in the third year, 137 in the fourth year, and 19 in May-June 2002). Fruit remains were encountered in 58.5% of faecal samples (706 of 1209), and in 46 of 52 months (88.5%) that scats were collected. Insect remains (primarily termites) were encountered in 75.4% of scats (911 out of 1209). Among all 706 fruit scats, most (74.2%) contained remains of one fruit species, 20.8% contained two species, 4.2% contained three, and 0.7% contained four species of fruit. Of 706 scats collected with fruit remains, 40.4% (238) contained 100% fruit. Remains of 83 distinct fruit species belonging to 39 genera in 23 different families were identified in sun bear scats. The most common genera encountered in scats were Ficus spp. (23.2% of scats), Dacryodes spp. (12.5% of scats; D. rugosa and D. rostrata combined), Syzigium spp. (12.3% of scats) and Santiria spp. (11.2% of scats) (Appendix I). Bears primarily fed on fruits from trees (60% of identified species) with remaining fruit species being primarily from monocots (Appendix II). Only 4 species of liana were positively identified to be consumed by sun bears and liana fruit resources made up <1 % of fruit scats. Moraceae was the most species rich family occurring in the diet of sun bears, with 14 species, followed by Palmae which featured with 10 species (Appendix II). The family Euphor- biaceae, which dominated in the phenology plots (21.4% of stems ≥ 10 cm dbh), featured with only 4 species in the diet, three of which were Baccaurea spp. (Appendix II). The 11 selected sun bear fruit tree taxa monitored for their phenology were encountered in 66.1% of fruit scats, indicating their importance in the fruit diet of sun bears. Frequency of occurrence of fruit in scats differed between years (Fig. 3) (Kruskal-Wallis H=23.0, df=3, P<0.001). The first year after the 1998 mast had significantly lower occurrence of fruit than the three subsequent years (Dunn-Šidák, Q>3.1, p<0.05 for all years). Sun bears appeared to be almost entirely frugivorous during the two mast-fruiting events (fruits found in 73 of 74 scats and 83 of 85 scats, respectively). Sun bears primarily fed on mast fruit species during the short period when they were available (Fig 4), ignoring the ‘non-mast’ species, although these also had higher fruiting activity during the masting events, possibly indicating a Phenological influences on sun bear frugivory 31

Figure 3. Frequency of occurrence of fruit in scats (n=706 out of 1209) based on running mean of 3 consecutive months.

Figure 4. Faecal samples containing fruit remains divided into mast and non-mast species (706 of 1206 scats collected). Bars indicate scats containing mast fruit species, lines indicate scats containing non-mast fruit species. Most scats collected in the ﬁrst year contained insect matter. 32 Chapter 2

preference for the mast species. By contrast, fruits were nearly absent in the diet of sun bears for a period of 12 months following the 1998 masting event (Fig. 4). All sun bear scats collected during this ﬁrst post ENSO/mast period primarily contained insect remains.

Temporal variation in use of fruit resources Based on faecal samples collected between Oct 1997-July 2002, the family Moraceae dominated the fruit diet of sun bears (26% of fruit scats), followed by Burseraceae (19%), Myrtaceae (9%), Annonaceae (8%), Palmae (8%), Tetrameristaceae (6%), and Bombacaceae (3%). The importance of various fruit families, genera, and species varied substantially over seasons and years (Fig. 5 and Fig. 6). In the first year, incorporating the first mast-fruiting event, the main plant genera in the diet were Artocarpus (Moraceae), Dacryodes (Burseraceae), and Durio (Bombacaceae). Sun bears consumed copious amounts of these, mostly large seeded (up to 45 mm), fruits. For example, a single scat could contain >400 Dacryodes rugosa seeds (seed length 15 mm; max scat weight 400 gr). In the second year, with the lowest overall fruit availability, Ficus (Moraceae) was the dominant genus. Figs were the most frequently occurring fruit genus encountered in scats throughout the study period (164 of 706 fruit scats), and also occurred in 58.7% of months that scats were collected. Still, occurrence of scats containing fig remains were clumped temporally

2 (Id=1.96, χ =170.7, df=14, p<0.001). After the 1997-98 ENSO figs hardly featured in faecal samples for a period of 12 months. No correlation was found between number of fruiting fig trees and number of scats encountered each month with fig remains (Spearman Rank Correlation: R2=0.22, p=0.193, n=37). A number of taxa from the Palmae family also featured in the diet during these periods with extreme fruit lows (viz., Oncosperma horridum, Borassodendron borneensis, Polydocarpus sp., the rattans Calamus spp., Daemonorops spp., Korthalsia spp.). Fruits of O. horridum, clusters of small hard drupes, were encountered in 49 scats. Certain species in the family Annonaceae were also found to be important for sun bears during the inter-mast period. Two unidentified species of Polyalthia with large (up to 55 mm) fatty fruits were frequently encountered in scats (51 of 706) in the third year of the study, whereas Polyalthia sumatrana, which reaches high densities in the forest, was not encountered in scats at all. Fruits of Monocarpia kalimantanensis, though available throughout most of the study period, were only consumed during the fruit low period. In the third year Ficus spp., Santiria spp. and Tetramerista glabra dominated the fruit diet. In the fourth year Syzigium spp. dominated the fruit diet and to a lesser extent figs. The ginger Phenological influences on sun bear frugivory 33

Hornstedtia cf. reticulata (Zingiberaceae) also appeared in scats. In the ﬁfth year Dacryodes spp. and Durio spp. were again consumed copiously during the masting event, followed by Santiria spp. (post masting event) (Fig. 5).

Figure 5. Annual ﬂuctuations in contribution of main plant genera to the sun bear diet. The category ‘rattan’ covers several genera.

Figure 6. Annual changes in contribution of plant families to the sun bear diet. 34 Chapter 2

Direct feeding observations From direct observations (n=4977 hours) sun bears were found to feed on fruits for 10.3 ± 4.3% of their feeding time (n=3 bears), although the observation period did not include the mast fruiting events. Remaining feeding time was primarily spent on insect matter (termites, ants, beetle larvae, cockroaches, stingless bees), as well as small quantities of flowers and vegetative matter. Although relatively little time was spent feeding on fruits, sun bears were still observed to feed on fruits on 43.3 ± 6.0% of days that observation were carried out (n=711 days). We directly observed 568 fruit feeding events during which bears fed on at least 58 different fruit species of 30 genera and 20 families (Appendix II). Plants of 62 fruit-feeding events remained unidentified. Sun bears mainly fed on fruits that had dropped to the forest floor, and trees were only climbed in 25.4% of fruit-feeding events (144 of 568). Average diameter of trees climbed by sun bears for fruit feeding was 33.6 cm dbh (range 5-150 cm). The three main genera observed to be fed upon during direct observations almost make up 50% of fruit feeding time (Appendix I). Several fruit species which rank high in occurrence from scats, especially mast fruiting species, were not observed to be fed upon during direct observations (Appendix I), as the observation period did not cover the masting events. The genus Litsea (Lauraceae) ranks fourth in frequency during direct observations but was rarely found in scats. This is probably due to the fact that seeds from these fruits are soft and were probably destroyed during feeding, with few remains of Litsea spp. identified in scats. Fruits of the palm Borassodendron borneensis were relatively frequently observed to be fed upon by the bears (ranking ninth), although this primarily constituted of chewing on the fibres which surround the extremely hard endocarp. In scats one would occasionally encounter fibres, which in most cases remained unidentified. The total number of different fruit species fed upon by sun bears, from both scats and observations combined was at least 115, belonging to 54 genera in 30 families (Appendix II).

Fruit resources in relation to topographical types Fruit feeding observations were found unevenly distributed over topographical types when looking at their relative availability (χ2=476.6, df=4, p<0.001). More fruit-feeding was observed in swamp (χ2=429.5, df=1, p<0.001) compared to the other topographical types combined, and signiﬁcantly less fruit feeding observations in slope habitat (χ2=106.1, df=1, p<0.001) than expected. Actual fruiting events by sun bear fruit trees, derived from the ten 0.1 ha phenology plots, were also found unequally distributed among topographical types (χ2=35.1, df=4, p<0.001), with more fruiting events in swamp (χ2=13.1, df=1, p<0.001) and ridge topographical Phenological inﬂuences on sun bear frugivory 35

Table 2. Chi-square test results of the distribution of sun bear fruit feeding observations and fruiting events of sun bear fruit trees in relation to availability of topographical types.

Topograph- Availability aSun bear bFruiting events cFruit feeding ical types topogr. types fruit trees sun bear fruit observations trees % ± SD no of fruit no of fruiting χ2 no of obs χ2 χ2 trees events FFO vs ATT FFO vs FE Swamp 5.9 ± 4.6 40 42 ** > 150 * > * < Alluvial 17.3 ± 4.0 39 21 125 * > High-ﬂat 24.8 ± 9.3 14 11 * < 136 ** > Slope 45.3 ± 9.1 28 13 135 * < ** > Ridge 6.7 ± 2.7 23 41 ** > 22 ** < Total 144 128 568 aNumber of sun bear fruit trees in ten 0.1 ha phenology plots (0.2 ha/topographical type); bNumber of fruiting events from sun bear trees in phenology plots between Jan 1998-July 2002; cNumber of fruit feeding observations during 711 days of observations; FFO= Fruit Feeding Observations; ATT= Availability of topographical types; FE = Fruiting Events; * p<0.01; ** p<0.001; > indicates more than expected; < indicates less than expected

type (χ2=11.6, df=1, p<0.001) compared to all other topographical types combined, and less fruiting events in high-flat topographical type (χ2=10.4, df=1, p<0.01) than expected (Table 2). The distribution of fruit feeding observations in relation to actual fruiting events (fruit production) in the different topographical types was also unequally distributed (χ2=417.9, df=4, p<0.001). Despite a high number of fruiting events, ridge topographical type was significantly less used (χ2=206.9, df=1, p<0.001) compared to all other topographical types combined. More fruit feeding observations where observed in high-flat (χ2=170.4, df=1, p<0.001) and slope habitat (χ2=115.3, df=1, p<0.001) than expected from the number of actual fruiting events (Table 2).

Discussion

Fruiting phenology in Sungai Wain was highly aseasonal and ﬂuctuated largely between years. The majority of reproductive fruit trees used by sun bears displayed a supra-annual or mast fruiting pattern, indicating that fruit production by species occurring in the diet of the sun bear is irregular and consists of a short boom and prolonged bust situation. Two mast fruiting events that occurred during the 53 month of observations were spaced four years apart, 36 Chapter 2

falling well in the ~2-10 year interval that has been reported for other sites in Southeast Asia (Medway, 1972; Appanah, 1985; Ashton et al., 1988, Curran and Leighton, 2000; Wich and van Schaik, 2000; Sakai, 2002; Marshall, 2004). During the two masting events 15.7% and 10.3% of trees respectively produced fruits. This coincides with the figure reported by Sakai et al. (1999) for ‘general flowering’ when >10% of individuals under observation were flowering in Sarawak (Malaysian Borneo). A combination of exhausted energy reserves from mast fruiting species, and possibly disrupted fruiting patterns of inter-mast fruit resources resulted in a prolonged period of fruit scarcity. Fruit production slowly built up over 4 years to the second masting event. Both masting events took place during ENSO episodes (NOAA-CIRES, 2003). The strong influence of ENSO events in East Kalimantan can also be witnessed from prolonged droughts and associated forest fires that repeatedly affect the region (Leighton and Wirawan, 1986; Hoffman et al., 1999; Siegert et al., 2001). No temperature drop, proposed as a potential trigger for general flowering (Ashton et al., 1988; Yasuda et al., 1999), was observed in the ~2 months preceding general flowering in 2001, indicating that possibly other factors play a role in the onset of mass- flowering (see Corlett and LaFrankie, 1998). Over the ~5 years study period sun bears were found to feed on a large variety of fruits (115 species). Although 60% of scats collected contained fruit remains sun bears were found to spend only 10% of their feeding time on fruits during direct observations. The remainder of feeding time was spent on invertebrates. This discrepancy could be due to two factors. Firstly, decay rates of scats containing primarily insect remains are much quicker than those containing fruit remains. Insect scats disappear within 24 hours due to dung beetle activity, whereas fruit scats remain visible up to 3 weeks (G.M. Fredriksson unpubl. data). This indicates that studying the diet of sun bears based on faecal analysis alone could strongly bias the results towards frugivory. On the other hand, most fruits are large compared to insects and during a short fruit foraging session large quantities of fruits can be consumed, the remains of which can be expelled in several scats. During mast fruiting events sun bears were found to be almost purely frugivorous. Mast fruiting species fed upon by sun bears produce succulent fleshy fruits, and are probably important in the ecology of sun bears for building up energy reserves or restoring lost energy reserves for the prolonged inter-mast periods when few fruit resources are available. The main sun bear mast fruit genera (Durio spp., Dacryodes spp., and Artocarpus spp.) have high energy, protein, fat and potassium values (Voon and Kueh, 1999). The ability to store large amounts of energy as fat to survive periods of food shortages is well developed in bear species living Phenological influences on sun bear frugivory 37

in temperate regions, where bears hibernate during winter. Orangutans, which occur sympatrically with sun bears throughout much of the lowland forests in Sumatra and Borneo have substantial dietary overlap with sun bears (Galdikas, 1988; Leighton, 1993; Knott, 1998; G. M. Fredriksson, unpubl. data), at least on Borneo. Orangtans in West Kalimantan were found to gorge themselves during a masting event on non-dipterocarp fruits and consumed 2-3 times the amount of calories needed for daily energy requirements, putting on fat reserves (Knott, 1998). Subsequently, during the inter-mast period orangutans were living off these stored reserves and loosing weight (Knott, 1998). Sun bears in East Borneo appear seriously affected by prolonged fruit scarcity with potentially elevated mortality due to predation (Fredriksson, 2005) or starvation (G.M. Fredriksson, unpubl. data; Wong et al., in press). Most of the mast species consumed by sun bears and orangutans are higher in caloric content than non-mast fruits (Leighton, 1993; Knott, 1998; Voon and Kueh, 1999; G.M. Fredriksson, unpubl. data). This might explain why intermast species were not consumed by bears during the mast period, even though many of these also produced larger crops during general masts. Densities of large strangling figs in Sungai Wain are low, but small hemi-epiphytic figs, producing minor fruit crops, are relatively abundant (G.M. Fredriksson unpubl. data). Despite the scarcity of large stranglers, figs were the most commonly consumed fruit genus by sun bears during the prolonged intermast period, witnessed both from scats and direct observations. Although figs are low in overall energy value compared to many of the mast fruit species, they contain high protein and calcium values (O’Brien et al., 1998; Voon and Kueh, 1999). Figs also displayed the most random fruiting pattern and were the most frequently available fruit genus. Probably due to the combination of these factors, figs appear one of the main fallback fruit resources for sun bears, as for a variety of other wildlife in the Southeast Asia (e.g. Leighton and Leighton, 1983; Sugardjito et al., 1987; van Schaik, 1996; Kinnaird et al., 1999). Figs are also encountered in the diet of several other bear species inhabiting (sub) tropical regions (e.g. Sloth bear Melursus ursinus, Joshi et al., 1997; Asiatic Black bear Ursus thibetanus, Hwang et al., 2003; Spectacled bear Tremarctos ornatus, Peyton, 1980). In the first year after the strong ENSO, few scats with fig remains were encountered even though fruiting by all other intermast species was negligible. An increase in fig consumption was only observed after May 1999 (13 months after the ENSO event). Monitoring of fig phenology commenced in June 1999, after which fig fruits appeared available on a continuous basis. Harrison (2001) reported that flower and fruit production of dioecious figs decreased or even ceased after the 1997-98 ENSO drought in Sarawak (Malaysian Borneo). Under normal conditions figs produce fruits all year round in order to maintain species-specific pollinator 38 Chapter 2

relationships with fig wasps which have a short live span (Harrison, 2001; Shanahan et al., 2001). Fig wasps became locally extinct after the ENSO event and for some fig species took more than two years to recover (Harrison, 2001). Although monoecious figs in Sarawak did not suf- fer extinction of pollinators, Harrison (2001) found that pollination rates were lower after the ENSO drought for some of these (e.g. F. benjamina). It is therefore possible that sun bears hardly consumed figs in the first year after the ENSO event, as simply fewer fig fruits were available. Fruit production of Tetramerista glabra, another important intermast fruit resource, also appeared disrupted after the 1998 ENSO event. This species normally produces fruit year round (Soerianegara and Lemmens, 1993), but after the 1998 mast T. glabra only became available again in early 2000. Although T. glabra featured in only 8% of fruit scats, most large trees of this species in Sungai Wain are covered with multiple-age claw marks, indicating frequent and repeated usage by sun bears. T. glabra has been identified as an important fruit species during lean periods for orangutans (Singleton and van Schaik, 2001) and bearded pigs (Curran and Leighton, 2000). During the initial period post-ENSO, when fruits like Ficus and T. glabra were virtually absent, various fruits from the Palmae family featured in the bear diet. Although few palms were included in the monthly phenology, our general observations suggested that many species in the palm family displayed a continuous fruiting pattern. Many fruits in the Palmae family are rich in oils and carbohydrates and have been found to be important food for primates in Southeast Asia during food lean periods (Lucas and Corlett, 1991). Fruits of the ginger Hornstedtia cf. reticulata started appearing in the diet after 2001. Densities of this ginger species increased substantially in later stages of the study (G.M. Fredriksson, pers. obs.), probably due to increased light conditions at the forest floor level, following elevated mortality of large trees after the 1998 ENSO drought (van Nieuwstadt, 2002). Acorns (Lithocarpus spp. and Quercus spp. Fam. Fagaceae), popular in the diet of several other bear species (Asiatic black bear: Huygens and Hayashi, 2001; Hwang et al., 2003; American black bear Ursus americanus, Inman and Pelton 2002) were rarely encountered in the diet of sun bears at this lowland site, despite a relatively high density of these genera. On the other hand, half of scats collected in the mountainous interior of Borneo during other surveys, contained acorns remains (G.M. Fredriksson, unpubl. data), in accord with Davies and Payne’s (1982: 91) statement that sun bears ‘feed on large quantities of the hard seeds of the Fagaceae family’. It is possible that sun bears living at higher elevations feed more on oaks because densities of species with succulent fruits decrease rapidly at higher altitudes (Djojosudharmo and van Schaik, 1992), whereas Fagaceae increase in density (Pendry and Proctor, 1996). Interestingly, sun bears were found feeding more on fruits in high-flat forest type, com- Phenological influences on sun bear frugivory 39

pared to other topographical types, despite lower fruit production. This could be due to the fact that sun bears used this forest type more intensively because insect food resources reach higher densities in this drier forest type (G.M. Fredriksson unpubl. data). Sun bears focus on the more stable invertebrate food resource during the intermast period, opportunistically feeding on fruits encountered whilst foraging for insects, making the overall foraging strategy more efficient. Conversely, sun bears spent less time feeding on fruit resources in ridge topographical types despite the fact that fruit productivity was higher than expected. Ridge tops are disproportion- ally used by humans for travel through the forest. In Sumatra, Griffiths and van Schaik (1993) found that sun bears switched to a more nocturnal activity pattern in areas of high human usage, and although at this study site sun bears remained diurnal (G.M. Fredriksson unpubl. data), it is possible they spatially avoided this topographical type due to higher levels of human passage. Overall at this lowland study site, three plant families (Moraceae [26%], Burseraceae [19%], and Myrtaceae [9%]) made up more than 50% of the sun bear fruit diet throughout the study period. The density and distribution of these genera could be a contributing factor to regional differences in sun bear densities, and conservation of fruit trees from these genera should be incorporated in forest management practices as they contribute disproportionately to the diet of sun bears. In order to obtain a comprehensive picture concerning which fruit resources are important in the diet of sun bears it is essential to carry out long term observations, taking into account the multi-year phenological cycles observed in Sundaic forests. The proportional rep- resentation of various food items depended heavily on the overall phenological status of the forest during the sampling period. During the initial stages of this study it appeared that sun bears were pure frugivores, though if the study would have commenced several months later, sun bears would have appeared primarily insectivorous. From our analysis a picture emerges of the general feeding ecology of sun bears. They are adaptable and easily switch to more nutritious resources whenever these become available, while they appear able to effectively store fat and thus survive during periods of low food availability. Sun bears share these traits with other large frugivores and omnivores in island Southeast Asia, such as the orangutan (Leighton and Leighton 1983; Knott, 1998) and bearded pig (Caldecott, 1988). We speculate that these species have evolved these traits in response to the specific ecological challenges posed by an environment dominated by supra-annual periods of low food availability interspersed by brief periods of glut. To test this it is interesting to compare the evolution of sun bears with its nearest relatives in mainland Southeast Asia, where such intra-annual cycles are less dominant. 40 Chapter 2

The taxomic classifcations and phylogenetic relationships within the Ursidae remain subject to controversies. The main problem being that the family Ursidae represents a typical example of rapid evolutionary radiation and recent speciation events, dating back to mid- Miocene about 6-8 Mya (Kurtén, 1968; Goldman et al., 1989; Talbot and Shields, 1996; Waits et al., 1999; Yu et al., 2004). Previous phylogenetic studies placed the sun bear as a sister taxon to the youngest diverging clade, containing brown bears and polar bears (e.g. Talbot and Shields, 1996; Waits et al., 1999). A more recent analysis, based on nuclear DNA (Yu et al., 2004) and morphology (Sacco and van Valkenburgh, 2004) provides strong support for the sun and sloth bear being sister taxa. This seems more reasonable from the morphological standpoint, the sloth bear and sun bear being Asian subtropical or tropical species, markedly distinguished from other bears, including the sympatric Asiatic black bear, by their morphological and behavioral specialization which are likely due to recent adaptive change (Hall, 1981; Nowak and Paradiso, 1983; Goldman et al., 1989). The sun bear, which is the only bear species living in island Southeast Asia, probably di- verged between 4 and 2.5 mya from the other species of Southeast Asia (see Talbot and Shields, 1996; Waits et al., 1999). It remains unclear what caused the rapid split between these species, although Meijaard (2003; 2004b) hypothesized a vicariance model in which the ancestral sun bear became isolated in what is now Sundaland. By that time, the rising Himalayas had resulted in increasing dry Asian climates and stronger monsoonal patterns, causing a change in the vegetation of this region (Morley, 2000). It is possible that by the time sun bears evolved in Sunda- land they did so in an environment more and more affected by supra-annual fruiting patterns. Did these environmental conditions shape the present-day ecology of Bornean sun bears? The closest relative of the sun bear, the sloth bear (Yu et al., 2004), currently ranges from Nepal south through India and Sri Lanka and is the ursid species most specifically adapted to feeding on insects (Laurie and Seidensticker, 1977; Joshi et al., 1997). Laurie and Seidensticker (1977) speculated that the sloth bear was able to enter the myrmecophagous niche as com- petitors that feed on insects are allopatric in sloth bear habitat. For the sun bear it might not have been an option to evolve into the insectivorous niche due to a number of sympatrically occurring myrmecophages (pangolins: Manis spp.), nor to evolve into a more carnivorous niche due to the presence of several more efficient feline com- petitors, and a relatively low prey biomass in Southeast Asia (Meijaard, 2004a). Hence, the sun bear appears to have evolved more in the direction of primates; being agile in trees and able to access a variety of fruit resources when available, but also feeding substantially on insects, which are the most stable food resource in tropical forests (see Kikkawa and Dwyer, 1992). The sun Phenological influences on sun bear frugivory 41

bear has evolved a combination of morphological adaptations to exploit these two main food resources in the tropical rainforest. Some of these useful for tree climbing like small body size (weight between 27-65 kg [Lekagul and McNeely, 1977]), extremely long claws, naked soles, ﬂat- tened chest, and strongly inward curving feet (Pocock, 1941), but also several physical features which are indicative of insect feeding such as the longest tongue among the ursids, mobile lips, a nearly naked snout, huge canines (practical for opening hardwood in search for insects), and exceptionally large paws comparative to body size useful for digging up termites and ants, as well as breaking into logs. They are thus well adapted to use a wide variety of resources in different feeding strata. Additionally, the Bornean sun bear is signiﬁcantly smaller than the Sumatran or Malaysian/mainland sun bear (Meijaard, 2004b), possibly another adaptation linked to evolution in the most nutrient poor part of Sundaland (see MacKinnon et al., 1996). All of these factors combined make evolution of sun bears in the typical Sundaic environment a likely scenario.

Acknowledgements We thank the Indonesian Institute of Sciences (LIPI) for granting GMF permission to carry out research on sun bears in East Kalimantan. GMF would like to express gratitude to the Forest Research Institute in Samarinda for their assistance. The Tropenbos Foundation, the Borneo Orangutan Foundation, and the Conservation Endowment Fund of the American Zoo and Aquarium Association provided financial assistance. Dave Garshelis is thanked for logistical, financial and technical assistance. The Wanariset Herbarium, especially Ambrianshyah and Arifin are thanked for identification of plant samples. Field assistants Lukas, Damy and Adi are thanked for various assistance in the forest throughout the years, as well as many people from the Sungai Wain village. SBJ. Menken, D. Garshelis, V. Nijman and E. Meijaard are thanked for useful comments of previous drafts. MJ Lammertink and G. Usher are thanked for the map. SAW was funded by the Dutch Foundation for the Advancement of Science (NWO) while working on this paper.

Literature cited Appanah S. 1985. General flowering in the climax rain forests of South-east Asia. Journal of Tropical Ecology 1: 225-240. Ashton PS, Givnish TJ, Appanah S. 1988. Staggered flowering in the Dipterocarpaceae - New insights into floral induction and the evolution of mast fruiting in the aseasonal tropics. American Naturalist 132: 44-66. 42 Chapter 2

Berlage HP. 1949. Regenval in Indonesie. Koninklijk Magnetisch en Meteorologisch Observatorium te Ba- tavia. Verhandelingen no. 37. Caldecott J. 1988. Hunting and wildlife management in Sarawak. IUCN Tropical forest programme. Conklin-Brittain, NL, Wrangham, RW, Hunt, KD. 1998. Dietary response of chimpanzees and Cercopith- ecines to seasonal variation in fruit abundance: II. Macronutrients. International Journal of Primatol- ogy 19: 949-970. Corlett RT, LaFrankie JV. 1998. Potential impacts of climate change on tropical Asian forests through an influence on phenology. Climatic Change 39: 439-453. Curran LM, Caniago I, Paoli GD, Astianti D, Kusneti M, Leighton M, Nirarita CE, Haeruman H. 1999. Impact of El Niño and logging on canopy tree recruitment in Borneo. Science 286: 2184-2188. Curran LM, Leighton M. 2000. Vertebrate responses to spatiotemporal variation in seed production of mast-fruiting dipterocarpaceae. Ecological Monographs 70: 101-128. Davies G, Payne J. 1982. A faunal survey of Sabah. Kuala Lumpur, Malaysia: IUCN/WWF. 1-297. Djojosudharmo S, van Schaik CP. 1992. Why are orang utans so rare in the highlands? Tropical Biodiversity 1: 11-22. Fogden MPL. 1972. The seasonality and population dynamics of equatorial forest birds in Sarawak. Ibis 114: 307-341. Foster MS. 1977. Ecological and nutritional effects of food scarcity on a tropical frugivorous bird and its fruit source. Ecology 58: 75-85. Fredriksson GM. 2002. Extinguishing the 1998 forest fires and subsequent coal fires in the Sungai Wain Protection Forest, East Kalimantan, Indonesia. In Moore P, Ganz D, Tan LC, Enters T and Durst PB,

eds. Communities in flames: proceedings of an international conference on community involvement in fire management. Bangkok, Thailand: FAO and FireFight SE Asia, 74-81. Fredriksson GM. 2005. Predation on sun bears by reticulated python in East Kalimantan, Indonesian Bor- neo. The Raffles Bulletin of Zoology 53:165-168. Fredriksson GM, Nijman V. 2004. Habitat-use and conservation of two elusive ground birds (Carpococcyx radiatus and Polyplectron schleiermacheri) in Sungai Wain protection forest, East Kalimantan, Indo- nesian Borneo. Oryx 38: 297-303. Galdikas BMF. 1988. Orangutan diet, range, and activity at Tanjung Puting, Central Borneo. International Journal of Primatology 9: 1-35. Goldizen AW, Terborgh J, Cornejo F, Porras DT, Evans R. 1988. Seasonal food shortage, weight-loss, and the timing of births in saddle-back tamarins (Saguinus fuscicollis). Journal of Animal Ecology 57: 893-901. Goldman D, Giri PR, O’Brien S. 1989. Molecular genetic-distance estimates among the ursidae as indicated by one- and two-dimensional protein electrophoresis. Evolution 43: 282-295. Phenological influences on sun bear frugivory 43

Griffiths M, van Schaik CP. 1993. The impact of human traffic on the abundance and activity periods of Sumatran rain forest wildlife. Conservation Biology 7: 623-626. Hall ER. 1981. The Mammals of North America. Wiley, New York. Harrison RD. 2001. Drought and the consequences of El Niño in Borneo: a case study of figs. Population Ecology 43: 63-75. Hoffmann AA, Hinrichs A, Siegert F. 1999. Fire damage in East Kalimantan in 1997/98 related to land use and vegetation classes: satellite radar inventory results and proposals for further actions: IFFM- SFMP. 1-44. Huygens OC, Hayashi H. 2001. Use of stone pine seeds and oak acorns by Asiatic black bears in central Japan. Ursus 12: 47-50. Hwang M-H, Garshelis DL, Wang Y. 2002. Diets of Asiatic black bears in Taiwan, with methodological and geographical comparisons. Ursus 13: 111-125. Inman RM, Pelton MR. 2002. Energetic production by soft and hard mast foods of American black bears in the Smoky Mountains. Ursus 13: 57-68. Janzen DH. 1974. Tropical blackwater rivers, animals, and mast fruiting by the dipterocarpaceae. Biotropica 6: 69-103. Joshi AR, Garshelis DL, Smith JLD. 1995. Home ranges of sloth bears in Nepal: Implications for conservation. Journal of Wildlife Management 59: 204-214. Joshi AR, Garshelis DL, Smith JLD. 1997. Seasonal and habitat-related diets of sloth bears in Nepal. Journal of Mammalogy 78: 584-597. Kikkawa J, Dwyer PD. 1992. Use of scattered resources in rain forest of humid tropical lowlands. Biotropica

24: 293-308. Kinnaird MF, O’Brien TG, Suryadi S. 1999. The Importance of ﬁgs to Sulawesi’s imperiled wildlife. Tropical Biodiversity 6: 5-18. Knott CD. 1998. Changes in orangutan caloric intake, energy balance, and ketones in response to ﬂuctuat- ing fruit availability. International Journal of Primatology 19: 1061-1079. Krebs CJ. 1999. Ecological Methodology. 2 nd eds. Benjamin/Cummings. Kunkun JG. 1985. Short information on Malayan sun bear in Indonesia. Asiatic Bear Conference, 67-69. Kurtén B. 1968. The Pleistocene Mammals of Europe, Aldine, Chicago, MA. Laurie A, Seidensticker J. 1977. Behavioural ecology of the sloth bear (Melursus ursinus). Journal of Zoology, Lond 182: 187-204. Leighton M. 1993. Modeling dietary selectivity by Bornean orangutans-Evidence for integration of multiple criteria in fruit selection. International Journal of Primatology 14: 257-313. 44 Chapter 2

Leighton M, Leighton DR. 1983. Vertebrate responses to fruiting seasonality within a Bornean rain forest. In: Sutton SL, Whitmore TC and Chadwick AC, eds. Tropical Rain Forest Ecology and Management. Cambridge UK: Blackwell Scientiﬁc Publications. 181-196. Leighton M, Wirawan N. 1986. Catastrophic drought and ﬁre in Borneo tropical rain forest associated with the 1982-1983 El Niño southern oscillation event. In: Prance GT, ed. Tropical rain forest and world atmosphere. Boulder, Colorado, USA: Westview Press. 75-101. Lekagul MDB, McNeely BA. 1977. Mammals of Thailand. Kurusapha Ladprao Press, Thailand. Lucas PW, Corlett RT. 1991. Relationship between the diet of Macaca fascicularis and forest phenology. Folia Primatologica 57: 201-215. MacKinnon K, Hatta G, Halim H, Mangalik A. 1996. The ecology of Kalimantan. Periplus Editions Ltd., Hong Kong. Marshall AJ. 2004. Population ecology of gibbons and leaf monkeys across a gradient of Bornean forest types Harvard University, USA. McConkey K, Galetti M. 1999. Seed dispersal by the sun bear (Helarctos malayanus) in Central Borneo. Journal of Tropical Ecology 15: 237-241. McPhaden MJ. 1999. El Niño-The child prodigy of 1997-98. Nature 398: 559-562. Medway L. 1972. Phenology of a tropical rain forest in Malaya. Biological Journal of the Linnean Society 4: 117-146. Meijaard E. 2003. Mammals of south-east Asian islands and their Late Pleistocene environments. Journal of Biogeography 30: 1245-1257. Meijaard E. 2004a. The biogeographic history of the Javan leopard Panthera pardus based on a craniometric

analysis. Journal of Mammalogy 85: 302-310. Meijaard E. 2004b. Craniometric differences among Malayan sun bears (Ursus malayanus); evolutionary and taxonomic implications. Rafﬂes Bulletin of Zoology 52: 665-672. Momose K, Yumoto T, Nagamitsu T, Kato M, Nagamasu H, Sakai S, Harrison RD, Itioka T, Hamid AA, Inoue T. 1998. Pollination biology in a lowland dipterocarp forest in Sarawak, Malaysia. I. Characteristics of the plant-pollinator community in a lowland dipterocarp forest. American Journal of Botany 85: 1477-1501. Morisita M. 1962. Id-Index, a measure of dispersion of individuals. Researches in Population Ecology 4: 1-7. Morley RJ. 2000. Origin and evolution of tropical rain forests. John Wiley & Sons, Ltd, Chichester, United Kingdom. NOAA-CIRES. 2003. Climate Diagnostics Centre. http://www.cdc.noaa.gov/people/klaus.wolter/MEI/. Nowak RM, Paradiso JL. 1983. Walker’s Mammals of the World. Johns Hopkins University Press, Baltimore, MD. Phenological inﬂuences on sun bear frugivory 45

O’Brien TG, Kinnaird MF, Dierenfeld ES, Conklin-Brittain NL, Wrangham RW, Silver SC. 1998. What’s so special about figs? Nature 392: 668. Pendry CA, Proctor J. 1997. Altitudinal zonation of rain forest on Bukit Belalong, Brunei: soils, forest structure and floristics. Journal of Tropical Ecology 13: 221-241. Peres CA. 1994. Primate responses to phenological changes in an Amazonian terra-firme forest. Biotropica 26: 98-112. Peyton B. 1980. Ecology, distribution, and food habits of spectacled bears, Tremarctos Ornatus, in Peru. Journal of Mammalogy 61: 639-652. Pocock RI. 1941. The fauna of British India, including Ceylon and Burma. Mammalia-Vol II. Carnivora (suborders Aeluroidea (part) and Arctoidea). Taylor and Francis ltd, London, UK. Sacco T, van Valkenburgh B. 2004. Ecomorphological indicators of feeding behaviour in the bears (Carnivora: Ursidae). Journal of Zoology, Lond 263: 41-54. Sakai S. 2002. General flowering in lowland mixed dipterocarp forests of South-east Asia. Biological Journal of the Linnean Society 75: 233-247. Sakai S, Momose K, Yumoto T, Nagamitsu T, Nagamasu H, Hamid AA, Nakashizuka T. 1999. Plant reproductive phenology over four years including an episode of general flowering in a lowland dipterocarp forest, Sarawak, Malaysia. American Journal of Botany 86: 1414-1436. Servheen C. 1999. Sun bear conservation action plan. In: Servheen C, Herrero S and Peyton B, eds. Bears. Status Survey and Conservation Action Plan. Gland, Switserland: IUCN. 219-223. Shanahan M, Compton SG, So S, Corlett R. 2001. Fig-eating by vertebrate frugivores: a global review. Biologi- cal Reviews 76: 529-572.

Siegel S, Castellan NJJ. 1988. Nonparametric statistics for the behavioural sciences. McGraw-Hill, New York. Siegert F, Ruecker G, Hinrichs A, Hoffmann AA. 2001. Increased damage from ﬁres in logged forests during droughts caused by El Niño. Nature 414: 437-440. SigmaStat 1.0, 1992-1994, Jandel Corporation. Singleton I, van Schaik CP. 2001. Orangutan home range size and its determinants in a Sumatran swamp forest. International Journal of Primatology 22: 877-911. Soerianegara I, Lemmens RH. 1993. Plant resources of South-East Asia. Timber trees: major commercial timbers. Pudoc, Wageningen, Netherlands. 454-457. Sokal RR, Rohlf J. 1981. Biometry. W.H. Freeman, New York, USA. Sork VL. 1993. Evolutionary ecology of mast-seeding in temperate and tropical oaks (Quercus spp.). Vegeta- tio 107/108: 133-147. Strushaker TT, Leyland L. 1977. Palm-nut smashing by Cebus a. apella in Colombia. Biotropica 9: 124-126. 46 Chapter 2

Sugardjito J, te Boekhorst IJA, van Hooff JARAM. 1987. Ecological constraints on the grouping of wild orangutans (Pongo pygmaeus) in Gunung Leuser National Park. International Journal of Primatology 8: 17-41. Talbot S, Shields G. 1996. A phylogeny of the bears (Ursidae) inferred from complete sequences of three mitochondrial genes. Molecular Phylogenetics and Evolution 5: 567-575. Terborgh J. 1983. Five new world primates: A study in comparative ecology. Princeton University Press, Prince- ton, USA. Terborgh J. 1986. Keystone plant resources in the tropical forest. In: Soulé ME, ed. Conservation Biology: the science of scarcity and diversity. Sunderland, Mass. USA: Sinauer Associates. 330-344. van Nieuwstadt MGL. 2002. Trial by fire-Postfire development of a dipterocarp forest. Unpublished PhD Thesis, University of Utrecht. 1-143. van Schaik CP. 1986. Phenological changes in a Sumatran rain forest. Journal of Tropical Ecology 2: 327-347. van Schaik CP. 1996. Phenology: Forest seasonality in an everwet environment. In: van Schaik CP and Supri- atna J, eds. Mt Leuser, a Sumatran Sanctuary. 120-131. van Schaik CP, Terborgh JW, Wright SJ. 1993. The phenology of tropical forests - Adaptive significance and consequences for primary consumers. Annual Review of Ecology and Systematics 24: 353-377. van Schaik CP, van Noordwijk MA. 1985. Interannual variability in fruit abundance and reproductive seasonality in Sumatran Long-tailed macaques (Macaca fascicularis). Journal of Zoology, Lond 206: 533-549. Voon BH, Kueh HS. 1999. The nutritional value of indigenous fruits and vegetables in Sarawak. Asia Pacific Journal of Clinical Nutrition 8: 24-31. Waits LP, Sullivan J, O’Brien SJ, Ward RH. 1999. Rapid radiation events in the family Ursidae indicated by

likelihood phylogenetic estimation from multiple fragments of mtDNA. Molecular Phylogenetics and Evolution 13: 82-92. Walsh RPD, Newbery DM. 1999. The ecoclimatology of Danum, Sabah, in the context of the world’s rainforest regions, with particular reference to dry periods and their impact. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences 354: 1869-1883. Whitmore TC. 1985. Tropical rainforests of the Far East. Clarendon Press, Oxford. Wich SA, van Schaik CP. 2000. The impact of El Niño on mast fruiting in Sumatra and elsewhere in Malesia. Journal of Tropical Ecology 16: 563-577. Wong ST, Servheen C, Ambu L. 2002. Food habits of malayan sun bears in lowland tropical forest of Borneo. Ursus 13: 127-136. Wong ST, Servheen C, Ambu L. Norhayati A. in press. Impacts of fruit production cycles on Malayan sun bears and bearded pigs in lowland tropical forest of Sabah, Malaysian Borneo. Journal of Tropical Ecology. Phenological inﬂuences on sun bear frugivory 47

Wood GHS. 1956. The dipterocarp ﬂowering season in North Borneo, 1955. Malaysian Forester 19: 193-201. Yasuda M, Matsumoto J, Osada N, Ichikawa S, Kachi N, Tani M, Okuda T, Furukawa A, Nik AR, Manokaran N. 1999. The mechanism of general ﬂowering in Dipterocarpaceae in the Malay Peninsula. Journal of Tropical Ecology 15: 437-449. Yu L, Li QW, Ryder OA, Zhang YP. 2004. Phylogeny of the bears (Ursidae) based on nuclear and mitochondrial genes. Molecular Phylogenetics and Evolution 32: 480-494. 48 Chapter 2

Appendix 1:

Main sun bear fruit taxa from faecal analyses (N = 1209, October 1997 to June 2002) and direct observations (N = 4977 h, October 1998 to April 2001). Faecal analyses Family % Feeding Family % observations Ficus spp. Moraceae 23.2 Ficus spp. Moraceae 17.6 Syzigium spp. Myrtaceae 11.8 Syzigium spp. Myrtaceae 17.3 Santiria spp. Burseraceae 11.2 Santiria spp. Burseraceae 14.2 Dacryodes rugosa* Burseraceae 10.3 Litsea sp. Lauraceae 6.4 Tetramerista glabra Tetrameristaceae 8.1 Rattan spp. Palmae 6.4 Polyalthia spp. Annonaceae 7.4 Monocarpia Annonaceae 5.8 kalimantanensis Oncosperma horridum Palmae 6.9 Oncosperma horridum Palmae 5.5 Artocarpus integer* Moraceae 5.8 Tetramerista glabra Tetrameristaceae 4.0 Cryptocarya sp. Lauraceae 2.3 Borassodendron Palmae 2.8 borneensis Hornstedtia cf reticulata Zingiberaceae 2.3 Polydocarpus sp. Palmae 2.6 Monocarpia Annonaceae 2.3 Diospyros spp. Ebenaceae 1.2 kalimantanensis Dacryodes rostrata* Burseraceae 2.1 Cryptocarya sp. Lauraceae 1.0 Artocarpus anisophyllus* Moraceae 2.0 Garcinia spp. Guttiferae 0.9 Durio dulcis* Bombacaceae 1.8 Polyalthia spp. Annonaceae 0.8 Durio oxleyanus* Bombacaceae 1.8 Knema sp. Myristicaeae 0.8 Rattan spp. Palmae 1.7 Mangifera spp. Anacardiaceae 0.8 Baccaurea spp. Euphorbiaceae 1.6 Baccaurea spp. Euphorbiaceae 0.8 Artocarpus nitidus Moraceae 1.4 Lithocarpus gracilis Fagaceae 0.7 Garcinia spp. Guttiferae 1.4 Aglaia sp. Meliaceae 0.6 Mangifera spp. Anacardiaceae 1.3 Barringtonia sp. Euphorbiaceae 0.4 * Mast-fruiting species; percentages indicate % of faecal samples containing a particular taxon, or percentage of fruit feeding time spent on a taxon; ‘Rattan’ incorporates several genera. Phenological inﬂuences on sun bear frugivory 49

Appendix 2:

Overall list of plant species occurring in the diet of the sun bear (faecal analyses and observations) in Sungai Wain between October 1997 to July 2002. Species Family Scat/ Part eaten Growth form Observation* Aglaia sp. Meliaceae S fr tree Alangium ridley Alangiaceae S fr tree Annonaceae sp. 1 Annonaceae S/O fr liana Annonaceae sp. 2 Annonaceae S/O fr liana Artocarpus anisophyllus Moraceae S/O fr tree Artocarpus dadah Moraceae S/O fr tree Artocarpus integer Moraceae S/O fr tree Artocarpus nitidus Moraceae S/O fr tree Artocarpus sp. 1 Moraceae O fr tree Artocarpus sp. 2 Moraceae O fr tree Baccaurea bracteata Euphorbiaceae S/O fr tree Baccaurea macrocapra Euphorbiaceae S/O fr small tree Baccaurea sp. 1 Euphorbiaceae S fr small tree Barringtonia sp. 1 Lecythidacae O ﬂo small tree Borassodendron borneensis Palmae S/O fr, ﬂo big palm Calamus sp. 1 Palmae S/O fr rattan palm Calamus sp. 2 Palmae S/O fr rattan palm Crypteronia sp. Crypteroniaceae S/O fr tree Cryptocarya sp. Lauraceae S/O fr tree Dacryodes rostrata Burseraceae S/O fr tree Dacryodes rugosa Burseraceae S/O fr tree Daemonorops sp. 1 Palmae S/O fr rattan palm Dehaasia sp. Lauraceae S/O fr tree Dialium indum Caesalpiniaceae S/O fr tree Diospyros sp. 1 Ebenaceae S/O fr tree Diospyros sp. 2 Ebenaceae S/O fr tree Diospyros sp. 3 Ebenaceae S/O fr tree Durio dulcis Bombacaceae S fr tree Durio graveolens Bombacaceae S fr tree Durio lanceolata Bombacaceae S/O fr tree Durio oxleyanus Bombacaceae S/O fr tree Dysoxylum sp. Meliaceae S fr tree Eugenia polyanthe Myrtaceae S/O fr small tree Ficus benjamina Moraceae S/O fr strangler Ficus lowii Moraceae O fr small tree Ficus sp. 1 Moraceae O fr strangler Ficus sp. 2 Moraceae S/O fr strangler Ficus sp. 3 Moraceae O fr strangler 50 Chapter 2

Appendix 2 Continued

Species Family Scat/ Part eaten Growth form Observation* Ficus sp. 4 Moraceae O fr strangler Ficus sp. 5 Moraceae O fr strangler Ficus sp. 6 Moraceae O fr tree Garcinia mangostana Guttiferae S/O fr tree Garcinia parvifolia Guttiferae S/O fr tree Garcinia sp. 1 Guttiferae S/O fr tree Glochidion sp. Euphorbiaceae S fr tree Hornstedtia cf reticulata Zingiberaceae S fr undergrowth Horsfieldia sp. Myristicaceae S/O fr tree Ilex sp. Aquifoliaceae S fr tree Knema laterica Myristicaceae S/O fr tree Knema sp. 1 Myristicaceae S/O fr tree Korthalsia sp. 1 Palmae S/O fr rattan palm Korthalsia sp. 2 Palmae S/O fr rattan palm Lansium domesticum Meliaceae S fr tree Licuala spinosa Palmae O fr, lb, flo undergrowth palm Lithocarpus gracilis Fagaceae S/O fr tree Lithocarpus sp. 1 Fagaceae S/O fr tree Litsea angulata Lauraceae S/O fr tree Litsea sp. 1 Lauraceae S/O fr tree Litsea sp. 2 Lauraceae S/O fr tree Madhuca kingiana Sapotaceae S/O flo tree Magnoliaceae sp. 1 Magnoliaceae O fr tree Magnoliaceae sp. 2 Magnoliaceae O fr tree Mangifera caesia Anacardiaceae O fr tree Mangifera foetida Anacardiaceae O fr tree Mangifera sp. 1 Anacardiaceae O fr tree Mangifera torquenda Anacardiaceae O fr tree Marantaceae sp. 1 Marantaceae O lb, flo undergrowth Microcos sp. Tiliaceae S/O fr tree Monocarpia kalimantanensis Annonaceae S/O fr tree Nephelium sp. Sapindaceae S/O fr tree Oncosperma horridum Palmae S/O fr, flo palm Palaquium sp. Sapotaceae S/O fr tree Pandanus sp. 1 Pandanaceae O lb, fr undergrowth Pinanga sp. Palmae O lb, flo small palm Polyalthia sp. 1 Annonaceae S/O fr tree Polyalthia sp. 2 Annonaceae S/O fr tree Polydocarpus sp. Palmae S/O fr palm Pternandra sp. Melastomataceae S/O fr small tree Quercus argentata Fagaceae S/O fr tree Quercus sp. 1 Fagaceae S/O fr tree Phenological influences on sun bear frugivory 51

Appendix 2 Continued

Species Family Scat/ Part eaten Growth form Observation* Santiria oblongifolia Burseraceae S/O fr tree Santiria tomentosa Burseraceae S/O fr tree Sida sp. Malvaceae S fr shrub Syzigium tawahense Myrtaceae O fr tree Syzigium sp. 1 Myrtaceae S/O fr tree Syzigium sp. 2 Myrtaceae S/O fr tree Syzigium sp. 3 Myrtaceae S/O fr tree Tetramerista glabra Tetrameristaceae S/O fr tree Uvariastrum sp. Annonaceae S/O fr liana Walsura sp. 1 Meliaceae S/O fr tree Willughbeia angustifolia Apocynaceae S/O fr liana Xerospermum norhonianum Sapindaceae S/O fr tree Xerospermum sp. 1 Sapindaceae S/O fr tree Unidentiﬁed (22 diff species) S/O fr

Total species = 115; Total genera = 54; Total families= 30 *S=encountered in scats; O=direct feeding observation; S/O= both encountered in the diet from scats and observations; fr=fruit; ﬂo=ﬂower; lb=leafbase/pith Chapter 3

Movements and activity patterns of female sun bears in East Kalimantan, Indonesian Borneo: implications for conservation

With David L. Garshelis 54 Chapter 3

Abstract Sun bears (Helarctos malayanus) in Indonesia are primarily threatened by destruction of their forest habitat, with poaching for body parts a secondary threat. Protected areas are increasingly becoming important for the conservation of this species, as non-protected forest areas are rapidly being converted to other landuses, but relevant data on spatial and habitat requirements of sun bears are limited. Radio-telemetry was used to study the movements and activity patterns of Bornean sun bears (H. m. euryspilus) in a small lowland rain forest reserve in East Kalimantan, Indonesian Borneo. Three wild and two rehabilitated-released female bears were ﬁtted with motion-sensitive VHF radio-collars; three of these bears were tracked continuously for 1-3 years during 1999-2003. A total of 5,470 locations were recorded. Home range sizes (12-month period) were 4-5 km2 (95% MCP and 95% ﬁxed kernel). Home range overlap was considerable. Small home ranges may have been related to a diet composed predominantly of insects. No mast fruiting event occurred during the period of study. Activity monitoring (169 days, 4,049 hours, 24,294 readings) revealed that sun bears were predominantly diurnal and crepuscular throughout the year with activity commencing about 0.5 h before sunrise and lasting until 2.5 h after sunset, and low levels (20-30%) of activity at night. Bears were active 61.6% of the day (14.8 ± 2.3 hours), on average, with higher levels of activity associated with insect feeding than fruit feeding. A combination of relatively small home range size, bounded entirely within the reserve, and avoidance of areas with human disturbance (trails, burned forest) suggest that small reserves are still valuable for the conservation of sun bears. Long-term persistence will depend on prevention of poaching and maintaining connectivity to other forested areas. Movements and activity patterns of female sun bears 55

Introduction Habitat reduction, combined with unsustainable levels of hunting, has put many bear populations at risk (Garshelis, 2009). Ursids (bears) generally have large home ranges (Garshe- lis 2004), often with distinct movements to seasonal ranges in response to temporal variations in the distribution and availability of food (Garshelis and Pelton 1981, Blanchard and Knight 1991, Joshi et al. 1995, Ferguson et al. 1999, Liu et al. 2002, Hwang et al. 2010, Castellanos 2011). Radio-telemetry has provided important insights into the lives of bears, and has been instrumental for obtaining data on movement patterns and habitat use (Garshelis 2004) that has been of use for management and conservation planning. Although this technology has been used for several decades to study bears in northern regions (see Garshelis 2004) only a handful of radio-telemetry studies have been carried out on bears in the tropics (Nomura et al. 2004, Wong et al. 2004, Paisley and Garshelis 2006, Hwang et al. 2010, Castellanos 2011). The sun bear (Helarctos malayanus), ranging throughout much of Southeast Asia (Fig. 1), has been listed as ‘Vulnerable’ based on signiﬁcant habitat loss over the last 30 years (Fredriks- son et al. 2008). Southeast Asia experiences some of the highest levels of forest conversion in the world including deforestation, forest degradation and fragmentation (FAO 2010, Sodhi et al. 2010, Broich et al. 2011, Miettinen et al. 2011), and clearing for rapid expansion of the plantation sector (Koh 2007, Butler and Laurance 2008, Fitzherbert et al. 2008). Levels of threat to this bear species vary across the region with poaching for body parts being the greatest threat on mainland Southeast Asia (Dang 2006, Nea and Nong 2006, Vinitpornsawan et al. 2006), whereas habitat loss and fragmentation are currently considered the greatest threat in Indonesia and Malaysia (Tumbelaka and Fredriksson 2006, Wong 2006). Sun bears are a forest-dependent species, found in a variety of forest types (Fredriks- son et al. 2006a, Steinmetz et al. 2011) up to 2500 m above sea level (a.s.l.) and have been found to persist in logged-over (Wong et al. 2004, Meijaard et al. 2005, Linkie et al. 2007), and other disturbed forest areas (Fredriksson et al. 2006b). They have even been found to make foraging excursions into oil palm plantations (Nomura et al. 2004). Sun bears on the island of Borneo have been recognized as a subspecies, H. m. euryspilus, substantially (~30%) smaller than their mainland and Sumatran counterpart H. m. malayanus (Horsﬁeld 1825, Meijaard 2004). Some 45% of the extant distribution range of sun bears lies in Indonesia (IUCN Bear Specialist Group, Japan Mapping Workshop 2006), making it the single most important country for the conservation of this species. Some 24 million ha of forest are under some form of protected status on Sumatra and Kalimantan (Indonesian Borneo) (FWI/GFW 2002), although the area of forest under protection has not increased in the last decade (FAO 2011). Protected 56 Chapter 3

Figure 1. Sun bear distribution range in Southeast Asia and location of the Sungai Wain Protection Forest, East Kalimantan, Indonesian Borneo.

areas are seen as one of the most effective tools for conservation in the tropics (Kramer et al. 1997, Bruner et al. 2001, Terborgh et al. 2002), despite the fact that the effectiveness of many of these protected areas is limited due to lack of, or weak, in-situ management implementation (e.g. FWI/GFW 2002, Kinnaird et al. 2003, Curran et al. 2004, Gaveau et al. 2009, Linkie et al. 2010, Harrison 2011). The conservation value of large expanses of unprotected logged- over areas in Borneo has also been highlighted (Meijaard et al. 2005, Meijaard and Sheil 2008), although many of these areas are under further threat by corporate conversion (Butler and Laurance 2008). The size of protected areas, or available safe habitat, is important for many species, especially those with large ranges (Powell et al. 1996, Beringer et al. 1998, Woodroffe and Ginsberg 1998, Woodroffe 2001, Wielgus 2002). Movements and activity patterns of female sun bears 57

Lowland dipterocarp forest, richest in terms of biodiversity and previously covering large expanses of Borneo, is now becoming increasingly rare, fragmented, and affected by various forms of human disturbance (Curran et al. 2004). In order assess the spatial, ecological, and habitat requirements of sun bears in this threatened forest type in Kalimantan, we studied their ranging and activity patterns in relation to fruiting patterns and human disturbance in a small lowland forest reserve in East Kalimantan, Indonesian Borneo.

Methods

Study site This study was carried out in a lowland dipterocarp forest, the Sungai Wain Protection Forest (SWPF), near Balikpapan, East Kalimantan, Indonesian Borneo (1º 05’ S and 116º 49’ E), at the easternmost edge of the sun bears’ distribution range (Fig. 1). The reserve covers a watercatchment area of circa 10,000 ha. Approximately 50% of the reserve was burned by forest ﬁres in March-April 1998 (Fredriksson 2002), during one of the most severe El Niño- Southern Oscillation (ENSO) related droughts ever recorded (McPhaden 1999), leaving an unburned central core of 4,000 ha of primary forest as well as unburned forest along rivers. The geomorphology of the reserve is of eroded sedimentary formations with a relatively low elevational range (30 - 150m), intersected by many small rivers. No roads traverse the reserve and access is only by foot. One main trail leads northwards from the nearest village and forest edge following a ridge separating 2 main sub-watersheds in the reserve. Less than 2 km before this ends, the track splits with the west branch descending to a research station located in the middle of the reserve. This trail was used as the main access route to the research station, for ferrying logistical supplies to the research station twice a week by porters, as well as for varied research purposes. There was little other foot trafﬁc on this trail.

Fruiting phenology, and rainfall The most common tree families above 10 cm dbh (diameter at breast height) were Euphorbiaceae, Dipterocarpaceae, Sapotaceae and Myrtaceae. The relative dominance of Dip- terocarpaceae increased substantially in the larger size classes. Together, the 25 most common species formed 40% of the total stem density (van Nieuwstadt 2002). Two phenology data sets were collected during this study, using the same methodology. Phenology studies commenced in January 1998 and continued until July 2002. Ten 0.1-ha phenology plots (100 x 10 m) were 58 Chapter 3

established where all trees ≥10 cm dbh were measured, labeled, and identified by botanists from the Wanariset Herbarium. The total number of live trees in the 1-ha plots at the start of the study was 549 individuals of 186 species. Every month around the same date (3rd week), all trees in these plots were observed with 10 x 40 binoculars. The relative abundance of flowers, young and mature fruits, and young and old leaves, in relation to size of the crown, was estimated, though for analysis here we only considered presence or absence of fruit. The second data set contained only trees known, or thought to be, important in the diet of the sun bear; we chose 104 trees of 11 species (≥10 cm dbh) at scattered sites throughout the forest (Fredriksson et al. 2006a). Trees in this forest undergo synchronized supra-annual flowering and fruiting (masting), providing a burst of food for sun bears (Fredriksson et al. 2006a), but no such events occurred during this portion of our study. Daily rainfall, minimum-maximum temperature and humidity data were collected between January 1998 and October 2002, at ground level in the primary forest. Although rainfall patterns in the region are not consistent on an annual basis, a distinction could be made between a ‘wet’ season (Nov-April) and a ‘dry’ season (May-Oct) (t=4.96, df=6, p=0.002). Rainfall type falls within the category of ‘slightly seasonal’ according to Mohr’s index, with average annual rainfall 2968 ± 510 mm y-1. Temperature was stable throughout the study period with average maximum of 29.7 ± 0.7ºC and minimum 23.2 ± 0.4ºC. The study area lies almost on the equator so there is little variation in the timing of sunrise (~06:00 h) and sunset (~18:00 h) year-round.

Bear captures and radio-tracking We captured sun bears in locally-constructed barrel traps spaced 1-2 km apart. We also set Aldrich spring-activated foot snares in cubby sets (Johnson and Pelton 1980). Traps were checked daily by foot along an 8-hour trapline. Traps were baited with a variety of food items, including honey, fruits, mixtures of cow blood and fish juices, with chicken intestines finally proving the most productive lure. We trapped from October 1998 until July 2000. We immobilized captured bears with a mixture of Zolatil (tiletamineHCl/zolazepam- HCl; Virbac, France) at 5 mg/kg, and Domitor (medetomidine hydrochloride, Orion Pharma, Espoo, Finland) at 0.027 mg/kg via a blowpipe. Bears were weighed, ear tagged, measured, and fitted with conventional VHF (very high frequency; 164-166 MHz) radio-collars (Advanced Telemetry Systems, Isanti, MN, USA). Each collar contained an activity sensor that altered the base signal of approximately 65 pulses per minute (ppm) when stationary with additional pulses (170 ppm, range 138-210 ppm) when tilted or moved. We attached radio-collars with a leather Movements and activity patterns of female sun bears 59

breakaway link that was designed to drop off the bear’s neck after about 1–2 years (Garshelis and McLaughlin 1998). We administered Antisedan (atipamezole hydrochloride, Orion Pharma, Espoo, Finland) at the capture site to reverse the anesthesia, after completion of handling. We also radio-collared two female sun bears that had been orphaned in the wild as cubs and were undergoing a rehabilitation and release programme (Fredriksson 2001); they were incorporated into this study when they started becoming independent at 1.5-2 years of age, and able to live in the forest. These bears were habituated for behavioural observations by the primary investigator and one assistant, which aided in verifying activity associated with pulses emitted by the radio-collars, telemetry accuracy, movement patterns, and provided unique insight into various other aspects of sun bear ecology. The rehabilitated bears received supplemental food (primarily fruits) in the beginning of the study, which we gradually reduced and fully discontinued by the time these bears were 2-2.5 years of age. We located collared bears by ground-based triangulation with a hand held 2-element directional antenna (Rubber ducky, Telonics Inc., Mesa, AZ, USA). Radio-locations were taken along an extensive foot-trail system (~60 km), including the main ridge trail that was used for quick movement between telemetry readings and more efﬁcient movement between distant locations in the forest. We attempted to obtain locations of all collared bears three times a day: morning (06:00-10:00 h), midday (10:00-14:00) and late afternoon (14:00-18:00). Bearings were plotted on a 1:35,000 topographic map. We assigned a precision rating to each location based on the size of the error polygon, and discounted those where the location could not be ﬁxed within 150 m.

Home range and movements Home range sizes were calculated using 3 methods: 100% and 95% minimum convex polygons (MCP) as well as a ﬁxed kernel probability density estimator with a band width using least squares cross validation to calculate the smoothing parameter, and using the 95% isopleth to determine area (Powell 2000). The 95% MCP were generated by discarding the 5% of points furthest from the mean location of all points used for generation of the 100% MCP (Worton 1987). MCP ranges are indicative of the total area needed to sustain the bears, including some heavily-used patches separated by gaps with little use, whereas probabilistic home range models (kernel) estimate the area actually used most of the time; the former is more useful for conservation (our purpose), whereas the latter may be more biologically relevant. In addition, MCP is less sensitive than other methods to variations in sampling intensity and lack of independence among successive radiolocations (Garshelis 1983). We calculated home range 60 Chapter 3

sizes for monthly and annual (12-month) periods. For monthly range calculations we only used months with more than 15 days of location data. In order to test whether fruiting patterns and rainfall had an effect on ranging patterns we used monthly phenology data (percentage of monitored trees in fruit) and monthly ranging data calculated by 100% MCP and kernel (95%), using a stepwise linear regression. To test whether human trafﬁc along the ridge-trail affected sun bear movement patterns, we generated a 70% minimum convex polygon using all combined locations, and compared all bear locations within that polygon to the same number of random points within the polygon. Speciﬁcally, we compared the distribution of bear locations to random points with respect to the main trail, using a raster grid with a 20-m cell size. To test the accuracy of our triangulated movement data, we measured real distances traveled by one of our rehabilitated bears by closely following it for several hours on different days using a Walktax Distance Measurer (Forestry Suppliers Inc, Jackson, MS, USA). This created a step-by-step track of its travel route. We then estimated the total average distance that this bear traveled in a day by expanding the sampled distance during active hours to the average number of hours of total activity per day (see below). These distance measurements were compared with distances obtained from 4 subsequent 24-hour tracking sessions where locations of this same bear were obtained by triangulation. We used ArcGis 10/ArcView spatial analyst tool (ESRI Inc, Redlands, CA, USA) to plot and measure movements. MCP and kernels were generated using Geospatial Modeling Environment (Spatial Ecology LLC).

Activity patterns We quantiﬁed diel activity by listening to radio signals at 10-minute intervals for 24- hour periods, once a week for each bear during Feb 1999 - Oct 2002. We categorized signals as either active or inactive. To test whether monthly activity patterns (percentage time active for all bears pooled) were affected by monthly changes in fruiting phenology or rainfall we carried out a linear regression, using reduced major axis regression.

Results

Study subjects We captured 3 female sun bears in barrel traps during 1,538 barrel-trapnights (513 barrel-trapnights/bear capture) over a period of 22 months. One of the rehabilitated sun bears Movements and activity patterns of female sun bears 61

was also inadvertently captured in an Aldrich foot snare after 1,048 snare-trapnights. All foot snares were removed after this event and not redeployed, as the bear developed severe ne- crosis on the wrist after having been caught for less than 3 hours. Based on sign, we discerned that bears visited trapsites but were not captured on 66 occasions. Two of the captured sun bears were caught on consecutive days (5-6 July, 1999), after almost one year of trapping. Both were emaciated, weighing 23-25 kg, ~35% less than two female sun bears caught at a similar time in Sabah (Malaysian Borneo) (35 and 39 kg; Nomura et al. 2004). One (Nenek) appeared to be at least 20 years old based on the extreme wear of her canine teeth and molars, and scattered white hairs covering her feet and ﬂanks. The other (Bibi) also had worn canines but appeared to be somewhat younger (Table 1). These two females were each later found to have a cub, which we estimated to be 1-2 months and 3-4 months old, for Bibi and Nenek, respectively. Bibi was only monitored for 1 month, after which she was swallowed by a large reticulated python (Python reticulatus) (Fredriksson 2005a). Nenek was monitored for 2 months, after which she was found dead, partially scavenged by bearded pigs (Sus barbatus). Neither of the cubs was seen again but both were too small to have survived independently. The third bear (Liar) was captured in March 2000. This female appeared in a better physical condition, both in terms of body mass (30 kg) and tooth condition. We estimated her age at 10-15 years. At the time of capture she did not appear to be lactating and no subsequent observations were made of this bear in the company of cub(s). This bear was monitored for almost 1 year (Table 1), at which time the leather breakaway link degraded and the collar dropped off.

Table 1. Identity, condition, period of monitoring and total number of locations (coordinates) obtained for 5 radio-collared female sun bears.

Total Weight Physical Months number Bear Sex (in kg) Condition Estimated Age Monitoring period monitored locations Nenek F 25 Very poor Very old-20 yr+ 5 July 1999-9 Sep 1999 2 162 Bibi F 23 Very poor Adult- 15-20 yr 6 July 1999-31 July 1999 1 66 Liar F 30 Fair Adult-10-15 yr 13 March 2000-20 Feb 2001 11 877 Ganja F 34 Fair (sub) adult (2-6 yr) 1 Nov 1998-24 April 2003 54 (35 A)* 2,579 Ucil F 36 Fair (sub) adult (2-6 yr) 8 Jan 1999-8 June 2002 42 (27 A)* 1,786 5,470

* Data collected over the last 35 months (out of 54) and 27 (out of 42) months, respectively, were used for calculating activity patterns. 62 Chapter 3

Table 2. Annual (12-month period) home range sizes (in km2) calculated by 100% and 95% Minimum Convex Polygon (MCP), and 95% ﬁxed Kernel.

Home ranges (in hectares) for 3 female sun bears with simultaneous monitoring over 12-month periods

Bear Tracking period # mo 100% MCP n locations 95% MCP n locations 95% Kernel

Liar March 2000-Feb 2001 12 762 877 449 833 485 Ganja March 2000-Feb 2001 12 812 610 438 579 539 Ucil March 2000-Feb 2001 12 708 602 452 572 502 average 760 445 521 SD 74 10 26

Bear Tracking period # mo 100% MCP n locations 95% MCP n locations 95% Kernel

Ucil May 2000-April 2001 12 864 628 466 597 537 Ucil May 2001-April 2002 12 688 851 493 808 518

Ganja May 2000-April 2001 12 825 648 478 616 570 Ganja May 2001-April 2002 12 655 907 441 862 499 Ganja May 2002-April 2003 12 584 477 451 453 542 average 723 466 533 SD 118 21 27

Table 3. Monthly range sizes (in km2) calculated for months with >15 days data, between July 1999- October 2002.

Bear 100% MCP 95% MCP 95% Kernel Days # mo Locations

Nenek 1.6 (0.6) 1.2 (0.3) 2.2 (0.8) 56 2 151 Bibi 3.0 2.3 4.1 25 1 66 Liar 2.7 (1.0) 2.1 (0.9) 4.1 (1.5) 321 12 877 Ganja 2.9 (0.9) 2.2 (0.7) 4.5 (1.3) 876 38 2,281 Ucil 2.9 (0.9) 2.3 (0.8) 4.4 (1.4) 580 25 1,548

Average 2.8 2.2 4.3 SD 0.9 0.8 1.4 range (1.1-5.5) (0.9-4.4) (1.7-8.7) Movements and activity patterns of female sun bears 63

During the period that these bears were radio-tracked another 6-8 sun bears, judged to be different independent individuals, were encountered in the study area, including at least 4 other females with different-aged cubs (one with twins). The two rehabilitated females (Ganja and Ucil) were monitored over a period of 4.5 and 3.5 years respectively.

Ranging patterns We obtained 5,470 locations for the three wild and two rehabilitated bears during July 1999-June 2002 (Table 1). We were highly successful in obtaining radio-locations of the collared bears because of access to a trail system: location data were missing for only 9 days of 437 days (2.1%) of continuous attempted daily monitoring of the three wild sun bears. Yearly ranges were calculated for the 12-month period Mar 2000-Feb 2001, during which we collected simultaneous data for 3 female sun bears monitored over the longest period of time (Table 2). Their 12-month 100% MCP home range sizes were 7-8 km2. Home ranges were only 4-5 km2 when calculated with the outer 5% of locations excluded (95% MCP) or using the 95% kernel method, indicating that a sizeable portion of the range was used infrequently. Home range sizes remained consistent through time (especially 95% MCP and kernel) for the 2 rehabilitated sun bears, which were tracked in multiple years (Table 2). Ranges of these two bears, Bibi and Nenek, overlapped each other by 54 and 43%, respectively, for the short period of time (1 month) they were radio-tracked simultaneously. Each of the ﬁve bears had monthly home range sizes (95% MCP) of 2 km2 (average 2.2 ± 0.8[SD] km2; monthly variation 0.85-4.4 km2; Table 3). Monthly range sizes were slightly larger when more trees were fruiting (linear regression with 100% MCP: R2=0.079, p=0.015; with 95% kernel: slope=55, R2=0.073, p=0.020), with range size increasing with increasing number of trees in fruit. Rainfall had a minor effect on monthly ranging patterns (linear regression with 100% MCP: p=0.302; with 95% kernel: p=0.041).

Distribution of locations Fewer bear locations (317 of 3818 = 8.3%) in the 70% MCP formed by pooling all bears were within 100 m of the main trail than would have been expected from a random distribution (910 locations, 23.8%, χ2 =507, df=1, p=0.001). Very few bear locations (n=58, 1%) were obtained in the forest area that was affected by forest ﬁres in 1998. Some of the locations that were mapped in the burned parts of the forest could have been in unburned river corridors or in burned forest areas close to rivers, as location error and coarseness of the GIS river layer precluded differentiation on this scale. 64 Chapter 3

Figure 2. Annual home ranges of three radio-collared female sun bears in the SWPF, East Kalimantan, Indonesian Borneo, during March 2000-Feb 2001. Ranges were calculated using 100% and 95% MCP, and 95% ﬁxed kernel. Within the 95% ﬁxed kernel border, progressively darker gradients represent 80%, 60%, 40%, 20%, an 10% contours. Ganja and Ucil were rehabilitated and released, whereas Liar was wild-caught.

From the shape of the 95% kernel (Fig. 2) generated for bear Liar it appears that she avoided entering the burned forest bordering the southwestern part of her range. All of the radio-collared bears remained within the boundaries of the reserve during the time they were monitored. However, based on interview data and surveys along forest edge areas (Fredriksson 2005b), a number of untagged bears ventured near the forest edge.

Daily movement We obtained measurements of actual distance traveled by one bear (Ganja) on 16 days (during Nov 2000-Mar 2001) that she was visually observed. On these days we observed the bear for an average of 6.5 hours (range 5-8 hours) during which she traveled an average of 2,025 ± 444 m (SD) (range 1,053-2,671 m). Based on an average of 14.8 hours activity per day, which excluded all periods of non-activity (see below), this extrapolated to 4,604 ± 888 m (SD) travelled daily. In fact, this is an under-estimate as during each day of direct observations, the bear would usually be “inactive” for some of the time, thus the “active” period in which she travelled the average 2,025 m would be obtained in less than 6.5 hour. During this active time the bear fed at a rapid rate, zigzagging between primarily invertebrate food resources. Distances measured using telemetry locations obtained 3x per day (n=604 days with locations in morning, midday, and afternoon and another location the next day) yielded a much lower estimate of 24-hour movements (1,481 ± 606 m (SD)). Movements and activity patterns of female sun bears 65

Table 4. 24-hour activity monitoring session and percentage of time that sun bears were found active.

Bear ID 24-h sessions # hours # readings % time active (± SD)

Nenek 7 168 1,008 62.0 ± 11.0 Bibi 4 96 576 69.8 ± 5.6 Liar 42 1,004 6,022 57.7 ± 7.1 Ganja 89 2,135 12,807 64.2 ± 9.0 Ucil 27 647 3,881 58.0 ± 9.9 Total 169 days 4,050 24,294 61.6 ± 9.4

Figure 3. 24-hour activity pattern of 5 radio-collared female sun bears monitored between 1999-2003. Standard error (SE) bars are shown.

Activity patterns We obtained 24-hour activity data during 169 days of monitoring (producing >24,000 readings, Table 4) over a period of 3 years (Feb 1999-Jan 2002). We were unable to categorize activity on only 42 attempts (0.2%). 66 Chapter 3

Overall, female sun bears were active 14.8 ± 2.3 h (SD) per day (61.6 ± 9.4% SD, all 24-hour sessions combined, n=169). Average activity per day among four of the five bears was consistent (58–64%), with no indication that the rehabilitated bears were different than the wild bears. One wild bear (Bibi) had a higher level of activity (69%) for the 1-month that she was monitored, but this could have represented normal short-term variation. Sun bears were distinctly more active during daylight hours (85.5 ± 4.8 % SD) than at night (37.9 ± 20.0% (SD), n=169, t=19.6, df=142, p=0.001). They tended to become active about 0.5 h before sunrise and ceased activity about 2.5 h after sunset (Fig. 3). Peaks of activity generally occurred during 07:00-09:00 and 16:00-18:00 h (Fig. 3), with diminished activity during 10:00-14:00 h. The human-habituated individuals that we visually observed frequently slept for brief periods during the hottest time of day on a fallen log or on the forest floor. When these bears were cubs (< 1 year), they spent large parts of the day resting alone in trees. Activity levels were higher during the first 12-month period of data collection (Feb 1999-Jan 2000) than during the subsequent 2 years (average % activity = 69, 58, 57 for years 1-3, respectively; year 1 vs 2, t=6.7, df=21, p<0.001, 2-tailed; year 1 vs 3, t=5.1, df=21, p<0.001). This period of higher activity coincided with the period of lowest fruit availability, when bears were primarily foraging for invertebrates. Monthly activity decreased slightly with increased fruiting phenology, although not significantly so (R2=0.067, p=0.139). Rainfall did not seem to affect monthly activity levels (R2=0.063, p=0.151), although rain did seem to affect daily activities. Anecdotally we observed that bears occasionally sought refuge in a tree nest for several hours during persistent rainfall, whereas light rainfall after a hot dry day would trigger intense play activity among the rehabilitated bears.

Discussion

Home ranges Our study presents the ﬁrst year-round ranging and activity data of female sun bears in their range. Wong et al. (2004) presented similar data for male sun bears in northeastern Borneo. In this aseasonal lowland forest at the easternmost edge of their distribution range, where fruit resource availability is highly inﬂuenced by supra-annual community mast fruiting events, and where there is minimal seasonal variation in day length, ranges of female sun bears were relatively small, and entirely encompassed within the reserve boundaries. Home range sizes of bears are related to body size (McNab 1963, Gittleman and Harvey 1982, Gomp- Movements and activity patterns of female sun bears 67

per and Gittleman 1991), metabolic needs (McNab 1983) and distribution and abundance of food (McLoughlin et al. 1999, 2000, Gompper and Gittleman 1991), resulting in large inter- and intraspecific variation. Insectivorous carnivores usually have relatively small home ranges (Gittleman and Harvey 1982) seemingly because protein-rich insect prey, especially ants and termites, are abundant and omnipresent over relatively small areas (e.g. Wood and Sands 1978, Redford 1987). As such, the sloth bear (Melursus ursinus), the most insectivorous (mymecopha- gus — ant and termite eating) ursid, appears to have the smallest home ranges of any species of bear. During our study, sun bears had a comparable diet to sloth bears, spending up to 90% of their foraging time consuming insects. Ants and termites are among the few food resources consistently common in lowland rainforests (Collins 1989, Kikkawa and Dwyer 1992) and sun bears rely heavily on this resource during prolonged inter-mast intervals that occur on Borneo (Wong et al. 2005, Fredriksson et al. 2006a). The female sun bears in our study had somewhat larger home ranges than female sloth bears during periods (5-12 months) when sloth bears relied predominantly on insects (Nepal x = 4.7 km2 [100% MCP], Sri Lanka = 2.5 km2 [fixed kernel]; Joshi et al. 1995, Ratnayeke et al. 2007). However, sloth bears significantly expanded their ranges during seasons, or in places, where fruits were common in their diet (Joshi et al. 1995, K. Yoganand, unpubl. report). Sun bears on Borneo become much more frugivorous during rare mast fruiting events (Wong et al. 2002, 2005, Fredriksson et al. 2006a). Mast fruiting events occur 2-10 years apart (Medway 1972, Ashton et al. 1988, Curran and Leigthon 2000). No mast fruiting event occurred within the period that we were radio-tracking, so we cannot assess ranging patterns during times when these bear are predominantly frugivorous. Nevertheless, we did observe a slight increase in their area of use when more trees were in fruit, suggesting that it is energetically profitable for sun bears to range farther to find fruit. Fruit trees favored by sun bears are widely scattered and found at low densities throughout this forest (Fredriksson et al. 2006a). Yearly home range sizes (daily locations for 12-month period) for female sun bears in this study ( x = 5.1 km2, 95% fixed kernel) were smaller than those reported for male sun bears in Malaysian Borneo ( x = 14.8km2, 95% adaptive kernel) during a period when they also fed largely on insects (Wong et al. 2004). This was true even though three of four male bears in Wong et al.’s (2004) study were tracked for <1 year and were periodically out of range of the tracking equipment. Nomura et al. (2004) also recorded smaller ranges for female than male sun bears, although home range size calculated during their study was affected by a limited number of locations and bears being frequently out of telemetry range. Ursids, in general, display large variation in home range sizes between individuals, sexes, species, and even among 68 Chapter 3

populations within the same species (Garshelis 2004, 2009). Larger range size for male bears (2-5x) is a common trait seen in all bear species where range size has been studied. Typically, the ranges of adult males overlap those of several females (Koehler and Pierce 2003, Dahle and Swenson 2003, Garshelis 2004, 2009, Castellanos 2011). No data are available yet on the effects of cub rearing on ranging patterns in sun bears. Although two emaciated female sun bears with small cubs were caught during this study, both died shortly after radio-tracking began, so our data on them are very limited. Moreover, the fact that we caught these two bears on consecutive days, after a year of unsuccessful trapping, suggested that they were in dire condition and finally succumbed to our bait. At the time when these bears were caught there had been a lack of fruits for almost 1.5 years (Fredriksson et al. 2006a). Hence, their movements may not have been indicative of healthy females with cubs. Sun bears during this study displayed high levels of range overlap, which might have been exacerbated by potentially elevated densities due to an influx of bears from the uninhab- itable burned areas. During the fires we heard frequent vocalizations (seemingly indicative of aggression), which were virtually never heard in subsequent years. Sun bear use of burned-over areas remained low for close to a decade post fires, based on repeated sign transects there (Fredriksson et al. in prep). Both this study and the study by Wong et al. (2004) monitored only a small number of individuals in a limited geographical area. We expected home range sizes for sun bears to be relatively small due the small size of this subspecies and their predominantly insect dominated diet. It is possible that ranges of sun bears in western parts of their range (Sumatra and mainland Southeast Asia) are somewhat larger, due to more available fruit (Wich et al. 2011) and hence a higher proportion of fruit in their diet (Steinmetz et al. 2012). Sun bears have long been considered the most arboreal species of bear (Ewer 1973), so this three-dimensional use of the forest might suggest that they would have correspondingly small home ranges. However, most of the foraging that we observed (75%) was on the forest floor, where bears found both insects and fallen fruits. This might be different during a mast- fruiting event, or in places where sun bears are more frugivorous.

Activity patterns Sun bears at our study site had a clear diurnal activity pattern similar to that reported for sun bears in Sabah, approximately 700 km north of our study site (Wong et al. 2004). Daily activity rhythms and time budgets in bears have been related to seasonal changes in sunrise and sunset, weather conditions, food type and abundance, human disturbance, presence of other Movements and activity patterns of female sun bears 69

bears (of the same or another species) and the bear’s sex, age and family associations (e.g., Amstrup and Beecham 1976, Garshelis and Pelton 1980, Ayers et al. 1986, Roth and Huber 1986, Clevenger et al. 1990, Reid et al. 1991, Lariviere et al. 1994, MacHutchon et al. 1998, Holm et al. 1999, MacHutchon 2001, Wagner et al. 2001, Paisley and Garshelis 2006, Hwang and Garshelis 2007). Our study area had minimal seasonal variation in day length and the sun bears showed only short-term changes in activity patterns related to rainfall. Contrary to the strictly diurnal pattern of activity among our radio-collared bears, sun bears that raided farmers’ crops adjacent to our study site were reported to be nocturnal (Fredriksson 2005b). This is consistent with other reports indicating a switch to nocturnal behavior when sun bears were near humans. Griffiths and van Schaik (1993) reported nocturnal activity among sun bears in a forest area with human activity in Aceh, Northern Sumatra, and Nomura et al. (2004) reported that sun bears forayed into oil palm plantations in Sabah only at night. Likewise, Wong et al. (2004) found that that camera trap photos of bears showed a more crepuscular-nocturnal pattern than the telemetry data, and suggested that this was because camera traps were set in places with more human odors. Other bear species are also known to become more nocturnal in areas with high human activity (Beckmann and Berger 2003, Kaczensky et al. 2006, Schwartz et al. 2010). Activity levels were higher during the first year of monitoring, coinciding with the latter part of a prolonged period (1.5 years) when virtually no fruit was available in the forest (Fredriksson 2006a). Sun bears at this time were feeding predominantly on invertebrate food resources. To obtain sufficient quantities of this widely available, though frequently difficult to obtain food source (i.e. termites living in hard mounds, large logs) sun bears were active for longer periods of time. During this time they would travel relatively long distances (> 4.5 km over a 24-hour period) but in a small area, zigzagging between predominantly invertebrate food resources and feeding at rapid rates, a foraging behavior described for other insectivores (Gittleman and Harvey 1982).

Distribution of locations Consistent with their temporal response to humans, sun bears also showed spatial avoidance of human activity. Collared sun bears in this study tended to avoid a ridge trail used by humans during day-time. Accordingly, we conjecture that the low capture rates during this study might also have been related to an aversion to human smells and objects, as sun bears in the interior of the Sungai Wain Protection Forest have little experience with humans. Sun bears that were apparently more desensitized to human inﬂuences (close to logging operations 70 Chapter 3

or near oil palm plantations) were caught more easily in similar traps (Nomura et. al. 2004). Sun bears during this study also rarely ventured into forest areas that had previously been burned. Avoidance of these areas was probably caused by a variety of factors, starting with the inaccessible nature of burned forest areas for several years post-fire event due to a thick impenetrable understory of ferns. Secondly, food availability was significantly reduced, with close to 80% mortality in sun bear fruit trees (Fredriksson et al. 2006b), as well as a significant decrease in invertebrate food resources important in the diet of sun bears (Fredriksson et al. in prep). Reduced canopy cover due to high tree mortality post-fires, lack of cover and resultant changes in the micro-climate (i.e., higher temperature) in burned forest areas might also have affected sun bear use of these areas.

Conservation implications Insular Southeast Asia still suffers the one of the highest deforestation rates in the tropics (Butler and Laurance 2008, Sodhi et al. 2010), with lowland and peat swamp areas disap- pearing fastest (e.g. Miettinen et al. 2011). Small patches of lowland forest, considered the richest in terms of biodiversity, are now becoming the last remaining examples of this habitat, one example of which is the SWPF. This small reserve is currently one of the last protected areas with (partially) unburned coastal lowland rainforest in East Kalimantan, a forest type previously covering several million ha. Can these small reserves continue to play a role for the conservation of sun bears? Several of the results from our study indicate that they can. The sun bears monitored in SWPF were found to have relatively small home ranges, with signiﬁcant home range overlap. Sun bears here did not display seasonal movements, nor did ranges of female sun bears monitored in the center of the reserve go near the forest edge. In addition, a considerable number of (breeding) sun bears were observed in the area. Sun bears in SWPF also tended to avoid humans, both spatially as well as temporally, a useful trait as humans pose the prime threat to this species. Sun bear populations can thus seemingly persist, even in such small reserves; the same result was found for mainly insectivorous sloth bears, which also have small ranges (Ratnayeke et al. 2007). However, persistence of large wildlife in these small reserves strongly depends on poaching pressure (Harrison 2011). Bears may be killed for their parts, or as a result of conﬂicts with humans. Bears living near forest edges are attracted to crops, especially when natural food supplies are low, making border areas potential demographic sinks (Woodroffe and Ginsberg 1998). Sun bears are also a frequent by-catch in snares placed for primarily for ungulates at forest edges in Kalimantan (G. Fredriksson pers. obs.). Poaching of bears for their Movements and activity patterns of female sun bears 71

body parts has been on the rise in Southeast Asia (Mills et al. 1995, Dang 2006, Shepherd and Nijman 2007, Shepherd and Shepherd 2010, Foley et al. 2011) and in several parts of mainland Southeast Asia sun bears have been hunted close to extinction, even from large protected areas (Scotson 2011). The long-term future of sun bears living in such small reserves will also depend on the habitat matrix surrounding these areas, and their connectivity to other forest areas. The fact that sun bears are a forest dependent species, combined with their tendency to avoid human disturbance, might hamper their movement though human-dominated areas between forest patches.

Acknowledgements We thank the Indonesian Institute of Sciences (LIPI) for granting GMF permission to carry out research on sun bears in East Kalimantan. GMF would like to express gratitude to the Forest Research Institute in Samarinda for their assistance. The Tropenbos Foundation, the Borneo Orangutan Foundation, and the Conservation Endowment Fund of the American Zoo and Aquarium Association provided ﬁnancial assistance. Field assistants Trisno, Lukas, Damy and Adi are thanked for various help in the forest throughout the years, as well as many people from the Sungai Wain village who helped with ferrying logistical supplies and traps. Various students from the Mulawarman University in Samarinda are thanked for their collaboration in the forest during various student projects, especially Wiwin Effendi, Joned, Eko and Ali Mustofa. Ente Rood and Matthew Nowak are thanked for assistance with statistics. Graham Usher is thanked for all help with GIS analysis and maps for this paper.

References Amstrup, S.C., and Beecham, J. (1976). Activity patterns of radio-collared black bears in Idaho. Journal of Wildlife Management 40, 340-348. Ashton, P.S., Givnish, T.J., and Appanah, S. (1988). Staggered flowering in the Dipterocarpaceae - New insights into floral induction and the evolution of mast Fruiting in the aseasonal tropics. Am Nat 132, 44-66. Ayres, L.A., Chow, L.S., and Graber, D.M. (1986). Black Bear Activity Patterns and Human Induced Modifica- tions in Sequoia National Park. Bears: Their Biology and Management 6, 151-154. Beckmann, J. P. and J. Berger. (2003). Rapid ecological and behavioural changes in carnivores: the responses of black bears (Ursus americanus) to altered food. Journal of Zoology 261:207-212. 72 Chapter 3

Beringer, J., S.G. Seibert, A.J. Brody, M.R. Pelton, and Gilder, L.D.V. (1988). The inﬂuence of a small sanctuary on survival rates of black bears in North Carolina. Journal of Wildlife Management 62, 727–734. Blanchard, B.M., and Knight, R.M. (1991). Movements of Yellowstone Grizzly bears. Biological Conservation 58, 41-67. Broich, M., Hansen, M., Stolle, F., Potapov, P., Margono, B.A., and Adusei, B. (2011). Remotely sensed forest cover loss shows high spatial and temporal variation across Sumatera and Kalimantan, Indonesia 2000-2008. Environmental Research Letters 6, 014010. Bruner, A.G., Gullison, R.E., Rice, R.E., and da Fonseca, G.A.B. (2001). Effectiveness of parks in protecting tropical biodiversity. Science 291, 125-128. Butler, R.A., and Laurance, W.F. (2008). New strategies for conserving tropical forests. Trends in Ecology & Evolution 23, 469-472. Castellanos, A. (2011). Andean bear home ranges in the Intag region, Ecuador. Ursus 22, 65-73. Clevenger, A.P., Purroy, F.J., and Pelton, M.R. (1990). Movement and Activity Patterns of a European Brown Bear in the Cantabrian Mountains, Spain. Bears: Their Biology and Management 8, 205-211. Collins, N.M. (1989). Termites. In Tropical Rain Forest Ecosystems Biogeographical and Ecological Studies, H. Lieth, and M.J.A. Werger, eds. (Amsterdam, Elsevier), pp. 455-471. Curran, L.M., and Leighton, M. (2000). Vertebrate responses to spatiotemporal variation in seed production of mast-fruiting dipterocarpaceae. EcolMonogr 70, 101-128. Curran, L.M., Trigg, S.N., McDonald, A.K., Astianti, D., Hardiono, Y.M., Siregar, P., Caniago, I., and Kasischke, E. (2004). Lowland forest loss in protected areas of Indonesian Borneo. Science 303, 1000-1003. Dahle, B., and Swenson, J.E. (2003). Seasonal range sizes in relation to reproductive strategies in brown

bears Ursus arctos. Journal of Animal Ecology 72, 660-667. Dang, N.X. (2006). The status and conservation of bears in Vietnam. In Understanding Asian bears to secure their future (Ibaraki, Japan, Japan Bear Network), pp.61-65. Ewer, R.F.(1973). The carnivores. Cornell University Press, Ithaca, NY, USA. FAO (2010). Global Forest Resources Assessment: Indonesia. (Rome, Food and Agricultural Organization of the United Nations, Forestry Department). FRA2010/095. FAO (2011).State of the World’s Forests 2011. (Rome, Food and Agricultural Organization of the United Nations). Ferguson, S. H., M. K. Taylor, E. W. Born, A. Rosing-Asvid, and F. Messier (1999). Determinants of Home Range Size for Polar Bears (Ursus maritimus). Ecology Letters 2:311-318. Fitzherbert, E.B., Struebig, M.J., Morel, A., Danielsen, F., Brühl, C.A., Donald, P.F., and Phalan, B. (2008). How will oil palm expansion affect biodiversity? Trends in Ecology & Evolution 23, 538-545. Movements and activity patterns of female sun bears 73

Foley, K.E., Stengel, C.J., and Shepherd, C.R. (2011). Pills, Powders, Vials and Flakes: the bear bile trade in Asia (Petaling Jaya, Selangor, Malaysia, TRAFFIC Southeast Asia). FWI/GFW (2002). The State of the Forest: Indonesia (Bogor, Indonesia, and Washington D.C., Forest Watch Indonesia, Global Forest Watch), pp. 104. Fredriksson, G.M. (2001). Conservation threats facing sun bears, Helarctos malayanus, and experiences with sun bear reintroductions in East Kalimantan, Indonesia. In: The evaluation of bear rehabilitation projects from a conservationist viewpoint (Ouwehands Zoo, Rhenen, Netherlands). Fredriksson, G.M. (2002). Extinguishing the 1998 forest fires and subsequent coal fires in the Sungai Wain Protection Forest, East Kalimantan, Indonesia. In: Communities in flames: proceedings of an international conference on community involvement in fire management (Bangkok, Thailand, FAO and FireFight SE Asia). Fredriksson, G.M. (2005a). Predation on sun bears by reticulated python in East Kalimantan, Indonesian Borneo. The Raffles Bulletin of Zoology 53, 197-200. Fredriksson, G.M. (2005b). Human sun bear conflicts in East Kalimantan, Indonesian Borneo.Ursus 16, 130-137. Fredriksson, G.M., Wich, S.A., and Trisno (2006a). Frugivory in sun bears (Helarctos malayanus) is linked to El Niño-related fluctuations in fruiting phenology, East Kalimantan, Indonesia. Biological Journal of the Linnean Society 89, 489-508. Fredriksson, G.M., Danielsen, L.S., and Swenson, J.E. (2006b). Impacts of El Niño related drought and forest fires on sun bear fruit resources in lowland dipterocarp forest of East Borneo. Biodiversity and Conservation 15, 1271-1301.

Fredriksson, G., Steinmetz, R., Wong, S. &Garshelis, D.L. (IUCN SSC Bear Specialist Group) 2008.Helarc- tosmalayanus. In: IUCN 2012. IUCN Red List of Threatened Species. Version 2012.1. . Downloaded on 20 July 2012. Garshelis, D.L., and Pelton, M.R. (1980). Activity of black bears in the Great Smoky Mountains National Park.Journal of Mammalogy 61, 8-19. Garshelis, D.L., and Pelton, M.R. (1981). Movements of black bears in the Great Smoky Mountains National Park. Journal of Wildlife Management 45, 912-925. Garshelis, D. L. (1983). The role of sampling intensity in the selection of a home range model. Pages 321-334. Int. Conf. Wildl. Biotelem. Garshelis, D.L., and McLaughlin, C.R. (1998). Review and evaluation of breakaway devices for bear radiocollars. Ursus 10, 459-465. Garshelis, D.L. (2004). Variation in Ursid life histories – is there an outlier? In Giant Pandas: biology and conservation: 53–73. Lindburg, D. & Baragona,K.(Ed.). Berkeley: University of California Press. 74 Chapter 3

Garshelis, D.L. (2009). Family Ursidae (Bears). Pages 448–497 in D.E. Wilson and R.A. Mittermeier, editors. Handbook of the mammals of the world. Volume 1. Carnivores. Lynx Edicions, New York, USA. Gaveau, D.L.A., Epting, J., Lyne, O., Linkie, M., Kumara, I., Kanninen, M., and Leader-Williams, N. (2009). Evalu- ating whether protected areas reduce tropical deforestation in Sumatra. Journal of Biogeography 36, 2165-2175. Gittleman, J.L., and Harvey, P.H. (1982). Carnivore home-range size, metabolic needs and ecology. Behavio- ral Ecology and Sociobiology 10, 57-63. Gompper, M.E., and Gittleman, J.L. (1991). Home range scaling: intraspecific and comparative trends. Oe- cologia 87, 343-348. Griffiths, M., and Van Schaik, C.P. (1993).The impact of human traffic on the abundance and activity periods of Sumatran rain forest wildlife. Conservation Biology 7, 623-626. Harrison, R.D. (2011). Emptying the Forest: Hunting and the Extirpation of Wildlife from Tropical Nature Reserves. BioScience 61, 919-924. Holm, G.W., Lindzey, F.G., and Moody, D.S. (1999).Interactions of sympatric black and grizzly bears in northwest Wyoming. Ursus 11, 99-108. Horsfield, T. (1825). Description of the Helarctos euryspilus; exhibiting in the bear from the island of Borneo, the type of a subgenus of Ursus. Zoological Journal 2, 221-234. Hwang, M.-H., and Garshelis, D.L. (2007). Activity patterns of Asiatic black bears (Ursus thibetanus) in the Central Mountains of Taiwan. Journal of Zoology 271, 203-209. Hwang, M.-H., Garshelis, D.L., Wu, Y.-H., and Wang, Y. (2010). Home ranges of Asiatic black bears in the Cen- tral Mountains of Taiwan: Gauging whether a reserve is big enough. Ursus 21, 81-96.

Johnson, K.G., and Pelton, M.R. (1980). Prebaiting and snaring techniques for black bears.WildlSoc Bull 8, 46-54. Joshi, A.R., Garshelis, D.L., and Smith, J.L.D. (1995). Home ranges of sloth bears in Nepal: Implications for conservation. Journal of Wildlife Management 59, 204-214. Kaczensky, P., D. Huber, F. Knauer, H. Roth, A. Wagner, and J. Kusak (2006). Activity patterns of brown bears (Ursus arctos) in Slovenia and Croatia. Journal of Zoology 269:474-485. Kikkawa, J., and Dwyer, P.D. (1992). Use of scattered resources in rain forest of humid tropical lowlands. Biotropica 24, 293-308. Kinnaird, M.F., Sanderson, E.W., O’Brien, T.G., Wibisono, H.T., and Woolmer, G. (2003). Deforestation trends in a tropical landscape and implications for endangered large mammals. Conservation Biology 17, 245-257. Koehler, G.M., and Pierce, D.J. (2003). Black bear home-range sizes in Washington: climatic, vegetative, and social inﬂuences. Journal of Mammalogy 84, 81-91. Movements and activity patterns of female sun bears 75

Koh, L.P. (2007). Potential Habitat and Biodiversity Losses from Intensified Biodiesel Feedstock Production. Conservation Biology 21, 1373-1375. Kramer, R.A., van Schaik, C.P., and Johnson, J., eds. (1997). Last Stand: Protected areas and the defense of tropical biodiversity (New York, Oxford University Press). Lariviere, S., Huot, J., and Samson, C. (1994).Daily activity patterns of female black bears in a northern mixed-forest environment. Journal of Mammalogy 75, 613-620. Linkie, M., Dinata, Y., Nugroho, A., and Haidir, I.A. (2007). Estimating occupancy of a data deficient mammalian species living in tropical rainforests: Sun bears in the Kerinci Seblat region, Sumatra. Biological Conservation 137, 20-27. Linkie, M., Rood, E., and Smith, R. (2010).Modelling the effectiveness of enforcement strategies for avoiding tropical deforestation in Kerinci Seblat National Park, Sumatra. Biodiversity and Conservation 19, 973-984. Liu, X., A. K. Skidmore, T. Wang, Y. Yong, and H. H. T. Prins (2002). Giant Panda Movements in Foping Nature Reserve, China. The Journal of Wildlife Management 66:1179-1188. MacHutchon, A.G. (2001). Grizzly bear activity budget and pattern in the Firth River Valley, Yukon. Ursus 12, 189-198. MacHutchon, A.G., Himmer, S., Davis, H., and Gallagher, M. (1998). Temporal and spatial activity patterns among coastal bear populations. Ursus 10, 539-546. McLoughlin, P.D., Case, R.L., Gau, R.J., Ferguson, S.H., and Messier, F. (1999). Annual and seasonal movement patterns of barren-ground grizzly bears in the central Northwest Territories. Ursus 11, 79-86. McLoughlin, P.D., Ferguson, S.H., and Messier, F. (2000). Intraspecific Variation in Home Range Overlap

with Habitat Quality: A Comparison among Brown Bear Populations. Evolutionary Ecology 14, 39-60. McNab, B. K. 1963. Bioenergetics and the determination of home range size. American Naturalist 97:133-140. McNab, B. K. 1983.Ecological and behavioral consequences of adaptation to various food resources. Pages 664-697 in J. F. Eisenberg and D. G. Kleiman, editors. Advances in the study of mammalian behavior. American Society of Mammalogists Special Publication 7. McPhaden, M.J. (1999). El Niño -The child prodigy of 1997-98.Nature 398, 559-562. Medway, L. (1972). Phenology of a tropical rain forest in Malaya.Biological Journal of the Linnean Society 4, 117-146. Meijaard, E. (2004).Craniometric differences among Malayan sun bears (Ursus malayanus); evolutionary and taxonomic implications.Rafﬂes Bulletin of Zoology 52, 665-672. 76 Chapter 3

Meijaard, E., Sheil, D., Rosenbaum, B., Iskandar, D., Augeri, D., Setyawati, T., Duckworth, W., Lammertink, M.J., Rachmatika, I., Nasi, R., et al. (2005). Life after logging: reconciling wildlife conservation and production forestry in Indonesian Borneo (Bogor, Indonesia, CIFOR, WCS and UNESCO). Meijaard, E. and D. Sheil. 2008. The persistence and conservation of Borneo’s mammals in lowland rain forests managed for timber: observations, overviews and opportunities. Ecological Research 23:21-34. Miettinen, J., Shi, C., and Liew, S.C. (2011). Deforestation rates in insular Southeast Asia between 2000 and 2010. Global Change Biology 17, 2261-2270. Mills, J.A., Chan, S., and Ishihara, A.P. (1995).The bear facts. The East Asian market for bear gall bladder (Cambridge, TRAFFIC International), pp. 41. Nea, C., and Nong, D. (2006). The conservation of bears in Cambodia. In Understanding Asian bears to secure their future (Ibaraki, Japan, Japan Bear Network), pp. 57-60. Nomura, F., Higashi, S., Ambu, L., and Mohamed, M. (2004).Notes on oil palm plantation use and seasonal spatial relationships of sun bears in Sabah, Malaysia. Ursus 15, 227–231. Paisley, S., and Garshelis, D.L. (2006).Activity patterns and time budgets of Andean bears (Tremarctos ornatus) in the Apolobamba Range of Bolivia. Journal of Zoology 268, 25–34. Powell, R.A., Zimmerman, J.W., Seaman, D.E., and Gilliam, J.F. (1996). Demographic analyses of a hunted black bear population with access to a refuge. Conservation Biology 10, 224-234. Powell, R. A. ()2000. Animal home ranges and territories and home range estimators. in L. Boitani and T. K. Fuller, editors. Research techniques in animal ecology. Columbia University Press, New York. Ratnayeke, S., van Manen, F.T., and Padmalal, U.K.G.K. (2007). Home Ranges and Habitat Use of Sloth Bears

Melursusursinusinornatus in Wasgomuwa National Park, Sri Lanka. Wildlife Biology 13, 272-284. Redford, K. H. (1987). Ants and termites as food.Patterns of Mammalian Myrmecophagy. Pages 349-397 in H. H. Genoways, editor. Current Mammalogy. Plenum press, Nebraska. Reid, D., Jiang, M., Teng, Q., Qin, Z., and Hu, J. (1991). Ecology of the asiatic black bear (Ursus thibetanus) in Sichuan, China. Mammalia 55, 221-237. Roth, H.U., and Huber, D. (1986). Diel Activity of Brown Bears in Plitvice Lakes National Park, Yugoslavia. Bears: Their Biology and Management 6, 177-181. Schwartz, C. C., S. L. Cain, S. Podruzny, S. Cherry, and L. Frattaroli (2010). Contrasting Activity Patterns of Sympatric and Allopatric Black and Grizzly Bears. Journal of Wildlife Management 74:1628-1638. Scotson, L. (2011). The distribution and status of Asiatic black bears Ursus thibetanus and sun bears Helarc- tos malayanus in Xe Pian National Protected Area, Lao PDR (Free the Bears Fund, WCS, WWF). Movements and activity patterns of female sun bears 77

Shepherd, C., and Nijman, V. (2008). The trade in bear parts from Myanmar: an illustration of the ineffective- ness of enforcement of international wildlife trade regulations. Biodiversity and Conservation 17, 35-42. Shepherd, C.R., and Shepherd, L.A. (2010). The Poaching and Trade of Malayan Sun Bears in Peninsular Ma- laysia: New legislation to provide stronger deterrents. TRAFFIC Bulletin 23, 49-52. Sodhi, N., Posa, M., Lee, T., Bickford, D., Koh, L., and Brook, B. (2010).The state and conservation of South- east Asian biodiversity. Biodiversity and Conservation 19, 317-328. Steinmetz, R., Garshelis, D.L., Chutipong, W., and Seuaturien, N. (2011).The shared preference niche of sympatric Asiatic Black Bears and Sun Bears in a tropical forest mosaic. PLoS One 6, e14509. Steinmetz, R., Garshelis, D.L., Chutipong, W., and Seuaturien, N. (2012).Foraging ecology and coexistence of Asiatic black bears and sun bears in a seasonal tropical forest in Southeast Asia. Journal of Mammalogy Terborgh, J., Schaik, C.P.v., L.Davenport, and Rao, M., eds. (2002). Making parks work (Island Press,USA). Tumbelaka, L., and Fredriksson, G.M. (2006). The status of sun bears in Indonesia. In Understanding Asian bears to secure their future (Ibaraki, Japan, Japan Bear Network), pp. 73-78. van Nieuwstadt, M.G.L. (2002). Trial by ﬁre-Postﬁre development of a dipterocarp forest (PhD thesis Uni- versity of Utrecht, Netherlands), pp. 1-143. Vinitpornsawan, S., Steinmetz, R., and Kanchanasakha, B. (2006).The status of bears in Thailand.In Under- standing Asian bears to secure their future (Ibaraki, Japan, Japan Bear Network), pp. 50-56. Wagner, R.O., Hightower, D.A., and Pace, R.M., III (2001). Measuring levels and patterns of activity in black bears. Ursus 12, 181-188.

Wich, S. A., E. R. Vogel, M. D. Larsen, G. M. Fredriksson, M. Leighton, C. P. Yeager, F. Q. Brearley, C. P. van Schaik, and A. J. Marshall. 2011. Forest Fruit Production Is Higher on Sumatra Than on Borneo. PLoS ONE 6:Published online 2011 June 2028.doi: 2010.1371/ journal.pone.0021278. Wielgus, R.B. (2002). Minimum viable population and reserve sizes for naturally regulated grizzly bears in British Columbia. Biological Conservation 106, 381–388. Wong, S.T., Servheen, C., and Ambu, L. (2002). Food habits of malayan sun bears in lowland tropical forest of Borneo. Ursus 13, 127-136. Wong, S.T., Servheen, C., and Ambu, L. (2004). Home range, movement and activity patterns, and bedding sites of Malayan sun bears, Helarctos malayanus in the rainforest of Borneo. Biological Conserva- tion 119, 168-181. Wong, S.T., Servheen, C., Ambu, L., and Norhayati, A. (2005). Impacts of fruit production cycles on Malayan sun bears and bearded pigs in lowland tropical forest of Sabah, Malaysian Borneo. Journal of Tropi- cal Ecology 21, 627–639. 78 Chapter 3

Wong, S.T. (2006). The status and conservation of bears in Malaysia. In Understanding Asian bears to secure their future (Ibaraki, Japan, Japan Bear Network), pp.66-72. Wood, T. G. and W. A. Sands (1978). The role of termites in ecosystems. Pages 245-292 in M. V. Brian, editor. Production ecology of ants and termites. Cambridge University Press, Cambridge, UK. Woodroffe, R., and Ginsberg, J.R. (1998). Edge effects and the extinction of populations inside protected areas. Science 280, 2126-2128. Woodroffe, R. (2001). Strategies for carnivore conservation: Lessons from contemporary extinctions. In Carnivore conservation, J.L. Gittleman, S.M. Funk, D.W. Macdonald, and R.K. Wayne, eds. (Cam- bridge, UK, Cambridge University Press), pp. 61–92. Worton, B. J. 1987. A review of models of home range for animal movement. Ecological Modeling 38:277-298. Chapter 4

Impacts of El Niño related drought and forest fires on sun bear fruit resources in lowland dipterocarp forest of East Borneo

With L.S. Danielsen and J.E. Swenson

Biodiversity and Conservation (2006), 15(4): 1271-1301 82 Chapter 4

Abstract

Droughts and forest fires, induced by the El Niño/Southern Oscillation (ENSO) event, have increased considerably over the last decades affecting millions of hectares of rainforest. We investigated the effects of the 1997-98 forest fires and drought, associated with an exceptionally severe ENSO event, on fruit species important in the diet of Malayan sun bears (Helarc- tos malayanus) in lowland dipterocarp forest, East Kalimantan, Indonesian Borneo. Densities of sun bear fruit trees (≥10 cm DBH) were reduced by ~80%, from 167 ± 41 (SD) fruit trees ha-1 in unburned forest to 37 ± 18 fruit trees ha-1 in burned forest. Densities of hemi-epiphytic figs, one of the main fallback resources for sun bears during periods of food scarcity, declined by 95% in burned forest. Species diversity of sun bear food trees decreased by 44% in burned forest. Drought also affected sun bear fruit trees in unburned primary forest, with elevated mortality rates for the duration of 2 years, returning to levels reported as normal in region in the third year after the ENSO event. Mortality in unburned forest near the burn-edge was higher (25 ± 5% of trees ≥10 cm DBH dead) than in the forest interior (14 ± 5% of trees), indicating possible edge effects. Combined effects of fire and drought in burned primary forest resulted in an overall tree mortality of 78 ± 11% (≥10 cm DBH) 33 months after the fire event. Disturbance due to fires has resulted in a serious decline of fruit resources for sun bears and, due to the scale of fire damage, in a serious decline of prime sun bear habitat. Recovery of sun bear populations in these burned-over forests will depend on regeneration of the forest, its future species composition, and efforts to prevent subsequent fire events. Effects of fires on sun bear fruit resources 83

Introduction

Severe droughts, associated with increased occurrences of El Niño/Southern Oscil- lation (ENSO) events, have become more frequent over the past decades (e.g. Timmermann et al. 1999). These droughts have created conditions conducive for uncontrolled fires, which have damaged extensive areas of forest throughout the tropics, with fires in Southeast Asia being particularly severe on the islands of Borneo and Sumatra (Goldammer and Mutch 2000; Tacconi 2003). In recent years, forest fires have caused more deforestation than intentional clearing in some tropical regions (e.g. Cochrane et al. 1999). On Borneo between 3-5 million ha of primary forest were affected by fires in 1982-83 during a severe ENSO related drought (Lennertz and Panzer 1984; Malingreau et al. 1985). Smaller fire events occurred in 1990, 1992 and 1994, all coinciding with ENSO episodes (Salafsky 1998). During the severe 1997-98 ENSO event primary forest areas burned easily (Siegert et al. 2001) as prior drought stress led to the shedding of leaves by evergreen species and accumulation of dry litter on the forest floor (GMF pers. obs.). In 1997-98 in the Indonesian province of East Kalimantan alone, 5.2 million ha of land were affected by the fires, 2.6 million ha of which were forest, including several protected lowland reserves (Hoffman et al. 1999; Siegert et al. 2001; Fuller et al. 2004). Areas previously affected by fires have also become susceptible to more intense fires due to higher fuel loads and rapid desiccation, now even during ‘normal’ dry seasons (Cochrane and Schulze 1999; Cochrane et al. 1999; Laurance 2003). Several studies have investigated the effects of the 1997-98 drought and fires on tree mortality and forest structure (e.g. Nakagawa et al. 2000; Williamson et al. 2000; van Nieu- wstadt 2002; Slik et al. 2002; Potts 2003; Slik and Eichhorn 2003; Slik 2004). Their findings showed that drought significantly increased mortality rates in unburned forest (Nakagawa et al. 2000; van Nieuwstadt 2002; Potts 2003; Slik 2004; van Nieuwstadt and Sheil 2005), that overall tree mortality was extremely high in burned areas (van Nieuwstadt and Sheil 2005), that fire resulted in a strong reduction of climax tree density (Slik and Eichhorn 2003), and that species composition changed after fire damage due to disproportionate mortality of certain tree species groups and tree size classes (Slik et al. 2002; Slik 2004). Shifts in species composition in natural forest occur slowly under ‘normal’ conditions (Swaine et al. 1987), but catastrophic disturbances like repeated fires can reduce structural and biological complexity in forests (Schindele et al. 1989). Fire-return intervals of less than 90 years can eliminate rainforest tree species, whereas intervals of less than 20 years may eradicate tree growth entirely resulting in savanna-like landscapes (Cochrane et al. 1999). 84 Chapter 4

Few studies however have investigated the effects of drought and forest fires on wildlife or their food resources in Indonesia (Doi 1988; Anggraini et al. 2000; O’Brien et al. 2003). Forest-dependent species tend to become less abundant or even locally extinct and other, less-forest-dependent species invade the area or increase in abundance (Doi 1988; Anggraini et al. 2000; see also Barlow et al. 2002; Peres et al. 2003). Some of the larger, long-lived species persist, reluctant or unable to relocate themselves (Suzuki 1992; Anggraini et al. 2000), although effects on their life-history remain largely unknown (but see O’Brien et al. 2003). Furthermore, the proximate causes of why individual species perish or flourish – e.g. decreased or increased food-availability, reduced nest-sites – remain unstudied. We studied the effects of fires on a variety of food resources important in the diet of one of the largest extant mammals on Borneo, the Malayan sun bear. Sun bears are partly frugivores (McConkey and Galetti 1999; Wong et al. 2002; Fredriksson et al. in press), although during periodic mast-fruiting events fruit makes up almost 100% of the diet (Fredriksson et al. in press). These mast-fruiting events provide the opportunity for sun bears to gorge themselves on large amounts of succulent fruits, probably enabling them to build up, or recover, fat and energy reserves for the prolonged period of fruit lows preceding and following these supra- annual mast fruiting events (Fredriksson et al. in press). Sun bears are considered important actors in forest dynamics, in part due to their seed dispersal abilities (McConkey and Galetti 1999), especially of large-seeded fruits (Fredriksson unpubl. data). Preliminary data suggests that sun bears make little use of burned forest areas for several years after a fire event (Doi 1988; Fredriksson unpubl. data). The aims of this study were i) to investigate the effect of forest fires on the density of sun bear fruit resources; ii) to quantify the effect of forest fires on species diversity of sun bear fruit resources; and iii) to study the effect of drought on mortality of sun bear food trees in unburned forest and based on the above iv) discuss the effects of these changes on sun bear populations in fire affected areas.

Methods

Study area The study was carried out in the lowland dipterocarp forest of the Sungai Wain Protec- tion Forest (SWPF), a reserve near Balikpapan, East Kalimantan, Indonesian Borneo (1º 16’ S and 116º 54’ E) (Fig. 1). The reserve covers a watercatchment area of ca 100 km2. Average Effects of ﬁres on sun bear fruit resources 85

Figure 1. Map showing the location of Sungai Wain forest on the island of Borneo. The enlargement shows the Sungai Wain forest, indicating the unburned area (dark grey), and burned areas (light grey). The location of pairs of burned and unburned edge plots is shown. The circular inset shows an example of the lay out of one pair burned-unburned edge plots over the ﬁre edge.

annual rainfall was 2740 ± 530 mm (1998-2003). The topography of the reserve consists of gentle to sometimes steep hills, and is intersected by many small rivers. The area varies in altitude from 30 to 150 metres a.s.l. Trees with stems greater than 10 cm DBH (diameter at breast height) are dominated by the families Euphorbiaceae, Dipterocarpaceae, Sapotaceae, and Myrtaceae. The relative dominance of Dipterocarpaceae increases substantially in the larger size classes. Due to an altitudinal gradient, with rivers running in a north-south direction, the southern part of the reserve is moister. Dipterocarpaceae, dominant in the northern part of the reserve (above 10 cm DBH) decrease in abundance towards the moister south where Sa- potaceae and Euphorbiaceae become more dominant. The 25 most common tree species form 40% of the total stem density (van Nieuwstadt 2002). Several palm genera (Borassodendron, Oncosperma, Polydocarpus, Licuala) are common in the subcanopy and understory (especially rattans), and ginger species (Costaceae and Zingiberaceae) as well as Marantaceae, Araceae, and Pandanaceae are common in the understory. This paper only deals with trees >10 cm DBH. All growth forms of ﬁgs (Ficus spp.) were included. Only hemi-epiphytic ﬁgs with a diameter of ≥3 cm were included as potential food resources, as these were observed to bear fruits and to be fed upon by sun bears (Fredriksson pers. obs). 86 Chapter 4

History of forest fires at the study site Most of the reserve was unaffected by fires at the start of the study in 1997, except for a small area near the eastern border that had burned in 1982-83. The prolonged drought of the 1982-83 ENSO, and several subsequent shorter ENSO droughts, probably caused elevated levels of mortality among large trees, resulting in an irregular canopy cover. Consequently the primary forest became increasingly vulnerable to desiccation during droughts and more susceptible to fires. Drought in the region started in May 1997 and continued intermittently till late April 1998, with 6 out of 12 months having no rainfall at all, or well below 100 mm. Fires entered the SWPF in March 1998, initially from a neighbouring state-owned logging concession, but subsequently also from surrounding agricultural fields. Fires moved slowly through the leaf litter and remained mainly in the undergrowth. Occasionally the fire reached into the crowns of older trees with hollow trunks, or dead standing trees with resin residues. In burned areas, all leaf litter and surface soil humus was reduced to ash, and mortality of seedlings and saplings was close to 100% (Fredriksson pers. obs.). Fire breaks were created over a period of 2 months but nevertheless approximately 50% of the reserve was affected by the fires (Fredriks- son 2002), leaving an unburned central core of some 4,000 ha of primary forest.

Permanent sampling plots In the SWPF, 18 permanent sample plots (PSP), of 20 x 200 m (0.4 ha) each, were established in once-burned forest and adjacent unburned forest after the fire event in 1998 (Fig. 1) by the Wanariset Research Station (Tropenbos-Kalimantan Project). The set-up of the PSPs was designed around man-made firebreaks of ~1.5 m wide (see van Nieuwstadt 2002). Because the fire-breaks did not correspond to any topographical feature in the places where the PSPs were positioned, this allowed for a random sampling scheme with paired plots of unburned and burned forest at a short distance from each other. The PSPs were laid out in nine pairs, each pair of PSPs adjacent to each other over the firebreak between burned and unburned forest, forming one contiguous transect of 20 x 400 m, half in burned and half in unburned forest. The unburned PSPs lie adjacent to the burn edge and these plots are labelled ‘unburned edge plots’. The PSPs were nested in three groups and spread over a total area of circa 20 km2 (Fig. 1). The distance between two pairs of PSPs was more than 500 m. We counted and measured all trees (≥10 cm DBH) in the PSPs 33 months after the fires, and calculated the percentage of live trees in both burned and unburned edge PSPs. Liana’s were not sampled, as they only make up a small proportion (<1%) of fruits encountered in the diet of sun bears at the study site (Fredriksson et al. in press). Effects of fires on sun bear fruit resources 87

Mortality rates Annual mortality rates were calculated from ten 0.1-ha phenology plots (total 1 ha, all trees ≥10 cm DBH, n=549 trees at the start of the study) which were monitored on a monthly basis between January 1998–July 2004. These phenology plots were positioned in unburned forest at least 1 km from the burn edge, and are subsequently called ‘unburned interior’ plots. Annual mortality rates for the interior plots are calculated based on exact 12 month periods (Jan-Dec).

1/t Mortality rate was determined as: m={1-(Nt /N0) }* 100

where ‘m’ is mortality per year, N0 is the initial number of live individuals and Nt is the number of live individuals at re-census interval t (e.g. Sheil and May 1996). Percentage of live trees was calculated for the interior plots for the same time interval as for the burned and unburned edge plots.

Densities of tree species important in the diet of the sun bear A list of 115 fruit species eaten by sun bears in the study area was available from Fre- driksson et al. (in press). All trees (≥ 10 cm DBH) that provided these fruits were subsequently labelled ‘sun bear fruit trees’. Densities of sun bear fruit trees and all Ficus spp. were subsequently calculated for the three subsets of plots. Two common species-rich genera which occur in the diet of the sun bear (Syzigium spp. [Fam. Myrtaceae]; Diospyros spp. [Fam. Ebenaceae]) posed a problem for density calculations as identification to species level is difficult. Only certain species from these genera were fed upon by the bears, whereas others were consistently ignored. In order to avoid overestimation of the density of these genera, we first calculated the percentage of trees from these genera in the 1 ha interior plots that belonged to species actually fed upon by sun bears, based on leaf and fruit samples collected during direct feeding observations of sun bears. For both genera this was found to be approximately 50% of the individual trees in the 1 ha interior plots. Therefore, the density of these taxa as potential sun bear food resources was reduced by 50% for further analyses in all plots. Two other taxa, Madhuca kingiana [Fam Sapotaceae] and Pternandra spec. [Fam Melastomataceae], were infrequently encountered in the diet of the bears, but are common in the forest. In order to avoid overestimating the densities of sun bear fruit trees, we reduced the density of these species also by 50% in the analyses. 88 Chapter 4

Statistical analysis Differences between burnt and unburned forest plots were compared with paired t- tests. A one-way ANOVA, with a Bonferroni multiple comparison test, was used to check for differences in mortality rates between sampling regimes and years. Standard deviation is always given when the average is presented (average ± SD). An α-level <0.05 was chosen to indicate statistical signiﬁcance.

Results

Density of fruit resources Density of live fruit trees important in the sun bear diet, 33 months after the fire was 167 ± 41 trees ha-1 in unburned edge plots and 37 ± 18 trees ha-1 in burned plots, indicating a drought and fire-related reduction of nearly 78% (paired t-test: t=8.7, df=8, p<0.001) (Table 1). Densities of sun bear fruit trees in the unburned interior plots was 144 trees ha-1 (total area sampled 1 ha) at the start of the study in 1998 and declined to 133 trees ha-1 33 months after the drought. Densities of all important sun bear fruit genera were lower in burned forest compared to unburned edge plots, although this was only significant for 8 out of 19 genera (Table 1). Several genera, generally occurring at low-medium densities, were not represented by any live trees in burned plots (e.g. Monocarpia [Fam. Annonaceae], Quercus [Fam. Fagaceae], Litsea, Cryp- tocarya [Fam. Lauraceae]). The difference in densities for these genera was not significant due to the large variation in number of trees encountered in the 9 unburned edge plots (Table 1). Two of the main sun bear fruit genera, Artocarpus and Dacryodes, which contribute the bulk of fruit eaten by bears during masting events, declined significantly in densities, respectively from 11.9 ± 6.6 trees ha-1 to 1.9 ± 2.4 trees ha-1 and 11.4 ± 7.8 trees ha-1 to 2.5 ± 2.5 trees ha-1 (Table 1). The palm Oncosperma horridum, whose fruits are favoured by sun bears, declined from 8.1 ± 4.2 trees ha-1 in unburned edge plots to zero in burned areas (Table 1). The most important plant genus for sun bears which provides fruits during intermast periods, Ficus spp., declined significantly in burned forest. Most epiphytic, hemi-epiphytic and climber figs were encountered in the unburned edge plots, with only 4 of 74 figs (all sizes combined) observed in burned plots (paired t-test: t=6.5, df=8, p<0.001). Densities of figs important in the sun bear diet declined significantly from 5.6 ± 1.6 figs ha-1 in unburned edge plots and 0.3 ± 0.3 figs ha-1 in the burned plots (paired t-test: t=3.7, df=8, p=0.006), corresponding to a reduction of ~95% following fire. Effects of fires on sun bear fruit resources 89

Table 1. Densities (stems ≥10 cm DBH, average ± SD ha-1) calculated from 18 vegetation plots (total 7.2 ha) of the main fruit-bearing genera important in the sun bear diet in unburned edge forest (UBF-edge) and burned forest (BF) 33 months after the 1997-98 ﬁre event. Genera only include species that have been found to occur in the bear diet. Densities of Ficus mainly comprise of hemi-epiphytic stranglers ≥ 3 cm diam. P indicates signiﬁcance level for paired t-test (df=8).

Density Genera Family UBF-Edge SD Burned SD P Madhuca Sapotaceae 16.5 21.3 2.9 4.4 ns Artocarpus Moraceae 11.9 6.6 1.9 2.4 0.003 Dacryodes Burseraceae 11.4 7.7 2.5 2.5 0.017 Syzigium Myrtaceae 11.4 7 3.1 3.4 0.011 Baccaurea Euphorbiaceae 8.1 3.7 0.6 1.1 0.001 Oncosperma Palmae 8.1 10.5 0 - ns Diospyros Ebeneceae 7.8 6.4 3.3 2.8 ns Santiria Burseraceae 6.1 4.9 0.6 1.1 0.013 Ficus * Moraceae 5.6 3.9 0.3 0.8 0.006 Litsea Lauraceae 5.6 5.3 0 - 0.013 Lithocarpus Fagaceae 4.7 3.2 0.8 1.8 0.005 Garcinia Guttiferae 3.6 3.3 0.8 1.8 ns Polyalthia Annonaceae 3.3 2.5 1.1 1.8 ns Durio Bombacaceae 2.5 3.8 0.6 1.1 ns Quercus Fagaceae 1.4 2.8 0 - ns Cryptocarya Lauraceae 1.1 1.8 0 - ns Mangifera Anacardiaceae 0.6 1.7 0.3 0.8 ns Monocarpia Annonaceae 0.6 1.1 0 - ns Tetramerista Tetrameristaceae 0 - 0.6 1.1 ns All sun bear fruit species combined 166.9 40.7 36.8 17.8 <0.001

No signiﬁcant differences were found in densities for the variably common genera Madhuca and Diospyros. Madhuca kingiana is a dominant tree species in the permanent sample plots in the south of the reserve, which is the moister part of the forest due to topography. One unburned edge plot in the southern part of the forest (0.4 ha) contained 22 trees (≥10 cm DBH) of this species compared to several unburned edge plots in the western or northern part of the forest which contained zero trees of this species. 90 Chapter 4

Species diversity of fruit trees in burned and unburned forest Species diversity of sun bear fruit resources declined signiﬁcantly following ﬁre (paired t-test: t=11.9, df=8, p<0.001). At least 66 species which feature in the diet of sun bear occurred in unburned edge plots, whereas only 37 of these were observed in burned plots, representing a decrease of ~44% in richness of sun bear fruit species in burned forest areas (Appendix 1). The family Lauraceae, representing the sun bear fruit genera Cryptocarya and Litsea, as well as the family Caesalpiniaceae with the sun bear fruit genus Dialium were not encountered with any live trees in burned sampling plots, whereas in the unburned edge plots the genus Litsea (all species combined) occurred in 8 out of 9 plots and were represented with 20 trees. The genus Dialium was found in 6 out of 9 unburned edge plots. A total of 46 sun bear food tree species were encountered in the unburned interior plots (total area sampled 1 ha versus 3.6 ha sampled of burned and unburned edge plots each).

Tree mortality Overall tree mortality (both sun bear fruit trees and tree species not fed upon by bears) in burned forest was extremely high, with 77.5 ± 10.8% trees dead 33 months after the ﬁre event (range 58.2-90.0%, n=9 plots). In unburned edge plots mortality was also high with 24.9 ± 5.4 % of trees dead (range 17.7-35.6%, n=9 plots). Cumulative percentage of dead trees in the unburned interior plots 33 months after the drought was 14.1 ± 5.2 %, some 40% lower than encountered in the unburned edge plots.

Table 2. Annual mortality rates (%) of trees (≥10 cm DBH) in unburned interior plots calculated from ten 0.1 ha phenology plots. Mortality rates are presented separately for sun bear food trees and non-food trees, as well as for all trees combined.

Non-bear trees Bear trees All trees combined (n=405) (n=144) (n=549)

1998 4.69 3.47 4.37 1999 7.25 4.32 6.48 2000 2.79 0.00 2.04 2001 2.01 0.00 1.46 2002 2.35 2.26 2.32 2003 2.40 3.08 2.59 Average 3.58 2.19 3.21 Effects of ﬁres on sun bear fruit resources 91

Mortality rates between years differed significantly (one-way ANOVA F=6.6 df=5, p<0.001) although only mortality in 1999 was significantly different (higher) than all other years (Bonferroni, t>3.8, p<0.006 for all comparisons). Mortality remained elevated for two years in the interior plots and approached ‘normal’ rates reported for the region in the third year after the drought (Table 2). Mortality rates did not differ significantly between sun bear fruit species and species that do not occur in the diet of bears in the interior plots (t-test, t=2.3, df=5, p=0.067), although the p-value suggested a tendency. No significant difference was found in mortality of sun bear food trees and non-food trees between burned and edge plots 33 months after the fire (paired t-test, t>0.04, df=8, p>0.3).

Discussion The 1997-98 fires reduced the density of sun bear fruit tree species by nearly 80%, three years after the fire. Fruit resources are important in the diet of sun bears, partly to regain energy after periods of fruit scarcity and partly to build up fat reserves to cope with prolonged intermast periods (Fredriksson et al. in press). When few fruit resources are available sun bears subsist primarily on insects although densities of these were also highly reduced in burned forest areas (Fredriksson unpubl. data). The reduction in fruit trees measured during this study is almost double the 44% decline reported by Leigthon and Wirawan (1986) for fruit species important in the diet of frugivorous primates like Bornean gibbons (Hylobates muelleri) and orangutans (Pongo pygmaeus) after the 1982-83 fires in Kutai National Park, East Kalimantan. Possibly the figures presented Leigthon and Wirawan (1986) are underestimates of the true extent of damage as their sampling was carried out shortly after the fire event. We documented delayed mortality due to the fires and drought, which continued for at least two years after the ENSO event. Mortality rates approached ‘normal’ rates reported for the region in the third year after the drought (average 1.7 % for 9 study sites in the region see Phillips et al. 1994; Wich et al. 1999; Potts 2003). Fire reduced the important sun bear fruit genus Ficus spp. by 95% three years after the fires. This serious decline might well have negative consequences in terms re-establishment or persistence of frugivore populations, as figs have been found to be one of the main fruit resources during periods of food lows for a variety of wildlife (e.g. Leighton and Leighton 1983). Besides the large reduction of fig densities in burned areas, Harrison (2000) reported on local extinctions of fig wasps after the ENSO drought, which affected fig fruit production, even in forest areas unaffected by fires. Putz and Susilo (1994) however reported that ten years after 92 Chapter 4

the 1982-83 forest fires regeneration of hemi-epiphytic figs by establishment of new terrestrial connections was close to 60%. The reduction of almost 44% in species richness of sun bear fruit taxa in once-burned forest could lead to permanent changes in the composition of sun bear fruit resources in these areas, especially as certain sun bear fruit tree genera disappeared altogether from the burned plots. Fires affect dominant tree species more than rare species (Slik et al. 2002; Potts 2003), but the disappearance of rare species is more worrying as they might become locally extinct (Cochrane and Schulze 1999). Regeneration of the forest will largely depend on the crop of trees that sprouts after the fires. The proportion of seedlings that grew up five years after the 1982-83 fires in Sabah showed a close resemblance to the distribution of families in primary forest, although few Dipterocarp saplings were encountered (Woods 1989). Slik et al. (2002) found that lowland dipterocarp forest, 15 years after being affected by fire, did not show an increase in species richness, although stem density increased, but these primarily belonged to a few pioneer tree species (Macaranga). The recruitment of Macaranga, both in the under- and over-storey, indicated that recovery of species composition in burnt forests takes longer than in selectively logged forest, where after 15 years pioneer species were being replaced by primary forest species (Slik et al. 2002). Whitmore (1985) reported that a lowland dipterocarp forest, extensively damaged by storm and fire in 1880, was still unusually poor in diversity of upper-canopy species when surveyed some 70 years later. Close to 18 % of the burned forest in our study area has remained as unburned forest patches in swampy areas or near streams (Fredriksson and Nijman 2004), which might facilitate a more uniform regeneration due to seed dispersal of forest-interior species into the burned areas. But overall, it will take decades, if not centuries, for many slow-growing climax species to begin fruiting and provide food for wildlife (Whitmore 1985). Fruiting phenologies of remaining fruit trees in burned areas might also deviate from fruiting patterns in unburned forest due to changes in environmental conditions and exposure to higher levels of drought. Kinnaird and O’Brien (1998) reported that occurrence of flowering and fruiting by trees in burned areas was lower than in adjacent unburned areas just after the 1998 fires. The storage of water reserves in trees, which depends on availability of subsoil water, has been found to influence flowering, although water storing capabilities differed greatly between species (Borchert 1994). The possible influences of changed environmental conditions in burned forest on phenological patterns calls for further studies. Drought by itself had significant effects on sun bear food trees in unburned forest, with elevated mortality rates up to two years after the drought. Effects of the drought on Effects of fires on sun bear fruit resources 93

trees have been found to be differential, relative to abundance and habitat factors (Potts 2003), with trees in moister areas less affected (Potts 2003; Slik and Eichhorn 2003; Fredriksson and Nijman 2004). The combined effects of fire and drought were far more severe than the effects of drought alone. Heavily burned forest areas have become dominated by pioneer species (Woods 1989; Nykvist 1996; Toma 1999; Slik et al. 2002), have lower species diversity (Matius et al. 2000; Slik et al. 2002; this study), and decreased soil fertility due to a loss of inorganic nutrients (Nykvist et al. 1994; Malmer 1996). Several studies investigated the regeneration potential of burned forest areas after the 1982-83 fires. Above-ground biomass in a lowland forest in Sabah eight years after the fire was still only a quarter of adjacent unburned forest (Sim and Nykvist 1991, Nykvist 1996). The significant decrease in both density as well as species diversity of sun bear fruit resources due to fire could partially explain the reduced usage of burned forest by sun bears (Fredriksson unpubl. data). Few fresh signs of sun bears were encountered in burned areas up to five years after the fire event, and radio-collared bears rarely entered burned forest. Al- though the extent of the home ranges of our radio-collared bears before the fire was unknown (as they were caught in unburned forest following the fire event), sun bears ranged throughout the reserve prior to fires (Fredriksson 2005, Fredriksson unpubl. data,). Environmental conditions like an increase in temperature due to lack of canopy cover, exposure to rain, and also the inaccessible nature of burned areas due to a thicket of ferns which blocked the understory for approximately four years after the fires, hampering movement for large ground-dwelling mammals, probably discouraged usage by bears. Additionally, densities of various invertebrate food resources, especially termites, declined significantly after the fires (Fredriksson et al. unpubl. data). Doi (1988) reported that sun bears did not recover quickly after the 1982-1983 fires in the largest lowland conservation area in East Kalimantan, the 200,000 ha Kutai National Park, with few bear signs found even in the core of the park three years after the fires. Leighton and Wirawan (1986) reported a decrease in vertebrate densities after the 1982-83 fires, although large-bodied primates reportedly seemed to be the least affected, possibly due to their gener- alized omnivorous diets and behavioural flexibility to switch food types (van Schaik et al. 1993). On the other hand O’Brien et al. (2003) found a significant decrease in siamang (Symphalangus syndactylus) group sizes, as well as infant and juvenile survival, after forest fires in Sumatra to a point where it seems unlikely that groups will survive for more than two generations in burned areas. Tree mortality figures differed substantially between edge and interior plots in unburned forest. Almost 40% more dead standing trees were encountered in the unburned edge 94 Chapter 4

plots compared to interior plots. Although no study as yet has specifically investigated edge effects in Borneo, Laurance et al. (2000, 2001) reported that mortality along forest edges in Amazonia is higher than in the forest interior, with increased tree mortality levels penetrating up to 300 metres from the forest edge. The extremely high mortality recorded in burned areas might also be a cumulative effect of several severe ENSO events over the last decades, each of which probably has caused elevated mortality, resulting in a more open canopy and higher water stress during subsequent droughts. The fact that annual mortality rates in unburned forest were elevated for two years post-ENSO, could indicate that, with an increased rate of ENSO events, such prolonged elevated mortality rates will also be occurring on a more frequent basis. If forests in these drought prone areas experience mortality rates of 6-7 % every 15 years for extended periods of time, recruitment might not balance mortality and significant changes in forest structure could occur even without any direct human influence.

Conclusions Large changes in vegetation structure and environmental conditions in burned forests, coupled with a significant decrease in fruit resources important in the diet of sun bears, have caused a significant reduction of suitable habitat for this bear species in fire-affected areas. The slow regeneration of these burned areas will probably influence re-colonization by bears, even if they have been able to maintain large enough population numbers within the burned forest matrix and adjacent unburned forest areas. Little is known about the ability of sun bears to exploit new pioneer fruit resources which have sprouted since the fires. The most dominant pioneer genus (Macaranga) has dehiscent fruits with small arillate seeds primarily attractive to birds (Davies and Ashton 1999; Slik et al. 2000), and has not (yet) been encountered in the sun bear diet. The damage due to fires in primary forest has far surpassed that encountered in logged-over areas (Woods 1989; Slik et al. 2002). Unburned logged-over areas might have higher potential for biodiversity conservation than primary forest that has burned once, although long-term monitoring of floral regeneration patterns and wildlife diversity and abundance in burned areas needs to be carried out in order to determine the future value of such forest for conservation. The massive spatial scale of these fire disturbances and the relatively short times- pan during which they have affected vast areas (usually 2-3 months of fires) is a new and worrying phenomenon. With the increase in the frequency and severity of ENSO (Timmermann et al. 1999), the future of these burned over forests looks grim as regeneration will only take place Effects of fires on sun bear fruit resources 95

if no further fires affect these areas, whereas repeated fire damage is common (Cochrane et al. 1999, Siegert et al. 2001). These factors, aggravated by low fire prevention activities and a lack of law enforcement in the region, call for highly increased conservation efforts in these drought prone rainforest areas if productive sun bear habitat is to be retained.

Acknowledgements We thank the Indonesian Institute of Sciences (LIPI) for granting GMF permission to carry out research on sun bears in East Kalimantan. GMF would like to express gratitude to the Forest Research Institute in Samarinda for their assistance. The Tropenbos Foundation and the Conservation Endowment Fund of the American Zoo and Aquarium Association provided financial assistance. We especially thank Dave Garshelis for logistical, financial and technical assistance. We thank the Wanariset Herbarium, especially Ambrianshyah and Arifin for identification of plant samples. The Tropenbos Foundation is thanked for usage of the permanent sample plots. Field assistants Trisno, Lukas Nyagang, and Damy da Costa are thanked for various assistance in the field throughout the years, as well as many people from the Sungai Wain village. LSD received financial and technical assistance from the Norwegian University of Life Sciences. We thank SBJ. Menken, D. Garshelis, and V. Nijman for useful comments of previous drafts and G. Usher and M. van Nieuwstadt for the map.

References Anggraini, K., Kinnaird, M. and O’Brien, T. 2000. The effects of fruit availability and habitat disturbance on an assemblage of Sumatran hornbills. Bird Conservation International 10: 189-202. Barlow, J., Haugaasen, T. and Peres, C.A. 2002. Effects of ground ﬁres on understorey bird assemblages in Amazonian forests. Biological Conservation 105: 157-169. Borchert, R. 1994. Soil and stem water storage determine phenology and distribution of tropical dry forest trees. Journal of Ecology 75: 1437-1449. Cochrane, M.A., Alencar, A., Schulze, M.D., Souza, C.M.J., Nepstad, D.C., Lefebvre, P. and Davidson, E.A. 1999. Positive feedbacks in the ﬁre dynamic of closed canopy tropical forests. Science 284: 1832-1835. Cochrane, M.A. and Schulze, M.D. 1999. Fire as a recurrent event in tropical forests of the Eastern Amazon: effects on forest structure, biomass, and species composition. Biotropica 31: 2-16. Davies, S.J. and Ashton, P.S. 1999. Phenology and fecundity in 11 sympatric pioneer species of Macaranga (Euphorbiaceae) in Borneo. American Journal of Botany 86: 1786-1795. 96 Chapter 4

Doi, T. 1988. Present status of the large mammals in the Kutai National Park, after a large scale fire in East Kalimantan, Indonesia. In: Tagawa, H. and Wirawan, N. (eds.), A research on the process of earlier recovery of tropical rain forest, pp. 82-93. Fredriksson, G.M. 2002. Extinguishing the 1998 forest fires and subsequent coal fires in the Sungai Wain Protection Forest, East Kalimantan, Indonesia. In: Moore, P., Ganz, D., Tan, L.C., Enters, T. and Durst, P.B. (eds.), Communities in flames: proceedings of an international conference on community involvement in fire management. FAO and FireFight SE Asia, pp. 74-81. Fredriksson G.M. 2005. Human-sun bear conflicts in East Kalimantan, Indonesian Borneo. Ursus 16:130-137. Fredriksson, G.M. and Nijman, V. 2004. Habitat-use and conservation of two elusive ground birds (Carpo- coccyx radiatus and Polyplectron schleiermacheri) in Sungai Wain protection forest, East Kalimantan, Indonesian Borneo. Oryx 38: 297-303. Fredriksson, G.M., Wich, S.A. and Trisno. in press. Frugivory in sun bears (Helarctos malayanus) is linked to El Niño-related fluctuations in fruiting phenology, East Kalimantan, Indonesia. Biological Journal of the Linnean Society. Fuller D.O., Jessup T.C., and Salim A. 2004. Loss of forest cover in Kalimantan, Indonesia, since the 1997- 1998 El Nino. Conservation Biology 18(1):249-254. Goldammer, J.G. and Mutch, R.W. 2001. Global forest fire assessment 1990-2000. FAO, pp. 1-495. Harrison, R.D. 2000. Repercussions of El Niño: drought causes extinction and the breakdown of mutualism in Borneo. Proceedings of the Royal Society of London Series B-Biological Sciences 267: 911-915. Hoffmann, A.A., Hinrichs, A. and Siegert, F. 1999. Fire damage in East Kalimantan in 1997/98 related to land use and vegetation classes: satellite radar inventory results and proposals for further actions.

IFFM-SFMP, pp. 1-44. Kinnaird, M.F. and O’Brien, T. 1998. Ecological effects of wildfire on lowland rainforest in Sumatra. Conser- vation Biology 12: 954-956. Laurance, W.F. 2003. Slow burn: the insidious effects of surface fires on tropical forests. Trends in Ecology & Evolution 18: 209-212. Laurance, W.F., Delamonica, P., Laurance, S.G., Vasconcelos, H.L. and Lovejoy, T.E. 2000. Rainforest fragmentation kills big trees. Nature 404: 836-836. Laurance, W.F., Williamson, B.G., Delamoˆ Nica, P., Oliveira, A., Lovejoy, T.E., Gascon, C. and Pohl, L. 2001. Effects of a strong drought on Amazonian forest fragments and edges. Journal of Tropical Ecology 17: 771–785. Leighton, M. and Leighton, D.R. 1983. Vertebrate responses to fruiting seasonality within a Bornean rain forest. In: Sutton, S.L., Whitemore, T.C. and Chadwick, A.C. (eds.), Tropical Rain Forest Ecology and Management. Blackwell Scientific Publications, pp. 181-196. Effects of fires on sun bear fruit resources 97

Leighton, M. and Wirawan, N. 1986. Catastrophic drought and fire in Borneo Tropical Rain Forest Associ- ated with the 1982-1983 El Niño southern oscillation event. In: Prance, G.T. (ed.) Tropical rain forest and world atmosphere. Westview Press, pp. 75-101. Lennertz, R. and Panzer, K.F. 1984. Preliminary assessment of the drought and forest fire damage in Kalim- antan Timur. DFS German Forest Inventory Service for Gesellschaft fur Technische Zusammenar- beit (GTZ), pp. 1-45. Malingreau, J. P., Stephens, G. and Fellows, L. 1985. Remote sensing of forest fires: Kalimantan and North Borneo in 1982-83. Ambio 14: 314-321. Malmer, A. 1996. Hydrological effects and nutrient losses of forest plantation establishment on tropical rainforest land in Sabah, Malaysia. Journal of Hydrology 174: 129-148. Matius, P., Toma, T. and Sutisna, M. 2000. Tree species composition of a burned lowland dipterocarp forest in Bukit Suharto, East Kalimantan. In: Guhardja, E., Fatawi, M., Sutisna, M., Mori, T. and Ohta, S. (eds.), Rainforest Ecosystems of East Kalimantan. El Niño, Drought, Fire and Human Impacts. Springer- Verlag, pp. 99-106. McConkey, K. and Galetti, M. 1999. Seed dispersal by the sun bear (Helarctos malayanus) in Central Borneo. Journal of Tropical Ecology 15: 237-241. Nakagawa, M., Tanaka, K., Nakashizuka, T., Ohkubo, T., Kato, T., Maeda, T., Sato, K., Miguchi, H., Nagamasu, H., Ogino, K., Teo, S., Hamid, A.H. and Seng, L. H. 2000. Impact of severe drought associated with the 1997-1998 El Niño in a tropical forest in Sarawak. Journal of Tropical Ecology 16: 355-367. Nykvist, N. 1996. Regrowth of secondary vegetation after the ‘Borneo fire’ of 1982-1983. Journal of Tropi- cal Ecology 12: 307-312.

Nykvist, N., Grip, H., Sim, B.L., Malmer, A. and Wong, F.K. 1994. Nutrient losses in forest plantations in Sabah, Malaysia. Ambio 23: 210-215. O’Brien, T.G., Kinnaird, M.F., Nurcahyo, A., Prasetyaningrum, M. and Iqbal, M. 2003. Fire, demography and the persistance of siamang (Symphalangus syndactylus: Hylobatidae) in a Sumatran rainforest. Animal Conservation 6: 115-121. Peres, C.A., Barlow, J. and Haugaasen, T. 2003. Vertebrate responses to surface wildﬁres in a central Ama- zonian forest. Oryx 37: 97-109. Phillips, O.L., Hall, P., Gentry, A.H., Sawyer, S.A. and Vasquez, R. 1994. Dynamics and species richness of tropical rain forests. Proceedings National Academy of Sciences USA 91: 2805-2809. Potts, M.D. 2003. Drought in a Bornean everwet rain forest. Journal of Ecology 91: 467–474 Putz, F.E. and Susilo, A. 1994. Figs and ﬁre. Biotropica 26: 468-469. Salafsky, N. 1998. Drouht in the rainforest, part II. An update based on the 1994 ENSO event. Climatic Change 39: 601-603. 98 Chapter 4

Schaik, C.P. v., Terborgh, J.W. and Wright, S.J. 1993. The phenology of tropical forests: adaptive significance and consequences for primary consumers. Annual Review of Ecology and Systematics 24: 353-377. Schindele, W., Thoma, W. and Panzer, K.F. 1989. The forest fire 1982/1983 in East Kalimantan. Part 1: The fire, the effects, the damage and technical solutions. German Forest Inventory Service Ltd. Sheil D., and May R.M. 1996. Mortality and recruitment rate evaluations in heterogenous tropical forests. Journal of Ecology 84:91-100. Siegert, F., Ruecker, G., Hinrichs, A. and Hoffmann, A.A. 2001. Increased damage from fires in logged forests during droughts caused by El Niño. Nature 414: 437-440. Sim, B.L. and Nykvist, N. 1991. Impact of forest harvesting and replanting. Journal of Tropical Science 3: 251-284. Slik, F., Verburg, R.W. and Keßler, P. 2002. Effects of fire and selective logging on the tree species composition of lowland dipterocarp forest in East Kalimantan, Indonesia. Biodiversity and Conservation 11: 85-98. Slik, J.W.E., Priyono and Welzen. P.C. 2000. Key to the Macaranga Thou. and Mallotus Lour. species (Euphor- biaceae) of East Kalimantan, Indonesia. Gardens’ Bulletin Singapore 52: 11-87. Slik, J.W.F. 2004. El Niño droughts and their effects on tree species composition and diversity in tropical rain forests. Oecologia 141: 114-120. Slik, J. W.F. and Eichhorn, K.A. O. 2003. Fire survival of lowland tropical rain forest trees in relation to stem diameter and topographic position. Oecologia 137: 446-455. Suzuki, A. 1992. The population of orangutans and other non-human primates and the forest conditions after the 1982-1983’s fires and droughts in Kutai National Park, East Kalimantan. In: Ismail, G.,

Mohamed, M. and Omar, S. (eds.), Forest Biology and Conservation in Borneo. Yayasan Sabah, pp. 190-205. Swaine, M.D., Lieberman, D. and Putz, F.E. 1987. The dynamics of tree populations in tropical forest: a review. Journal of Tropical Ecology 3: 359-366. Tacconi, L. 2003. Fires in Indonesia: Causes, costs and policy implications. CIFOR, pp. 1-34. Timmermann, A., Oberhuber, J., Bacher, A., Esch, M., Latif, M. and Roeckner, E. 1999. Increased El Niño frequency in a climate model forced by future greenhouse warming. Nature 398: 694-696. Toma, T., Matius, P., Hastaniah, Kiyono, Y., Watanabe, R. and Okimori, Y. 2000. Tree species composition of a burned lowland dipterocarp forest in Bukit Suharto, East Kalimantan. In: Guhardja, E., Fatawi, M., Sutisna, M., Mori, T. and Ohta, S. (eds.), Rainforest Ecosystems of East Kalimantan. El Niño, Drought, Fire and Human Impacts. Springer-Verlag, pp. 107-119. van Nieuwstadt, M.G.L. 2002. Trial by fire-Postfire development of a dipterocarp forest. Ph.D. thesis. Uni- versity of Utrecht. Effects of fires on sun bear fruit resources 99

van Nieuwstadt, M.G.L. and Sheil, D. 2005. Drought, ﬁre and tree survival in a Borneo rain forest, East Kalimantan, Indonesia. Journal of Ecology 93: 191–201. Whitmore, T. C. 1985. Tropical rainforests of the Far East. Clarendon Press. Wich, S.A., Steenbeek, R., Sterck, E.H.M., Palombit, R.A. and Usman, S. 1999. Tree mortality and recruitment in an Indonesian rain forest. Tropical Biodiversity 6: 189-195. Williamson, G.B., Laurance, W.F., Oliviera, A.A., Delamonica, P., Gascon, C., Lovejoy, T.E. and Pohl, L. 2000. Amazonian tree mortality during the 1997 El Niño drought. Conservation Biology 14: 1538-1542. Wong, S.T., Servheen, C. and Ambu, L. 2002. Food habits of malayan sun bears in lowland tropical forest of Borneo. Ursus 13: 127-136. Woods, P. 1989. Effects of Logging, Drought, and Fire on Structure and Composition of Tropical Forests in Sabah, Malaysia. Biotropica 21: 290-298. 100 Chapter 4

Appendix 1:

List of sun bear fruit genera/species of trees (≥10 cm DBH) encountered in unburned-edge (UBF) and burned (BF) vegetation plots (total 7.2 ha), 33 months after the 1997-98 ﬁre. A ‘√’ indicates that the genus/species was encountered in the plots, a ‘0’ indicates it was absent. Number of plots indicates in how many of the 0.4 ha plots (9 plots in burned and 9 plots in unburned edge forest) the genus/species was encountered.

Genus Species Family UBF-edge No of plots BF No of plots Aglaia spec. Meliaceae √ 7 0 0 Alangium ridleyi Alangiaceae √ 3 √ 1 Artocarpus anisophyllus Moraceae √ 9 √ 3 Artocarpus dadah Moraceae √ 1 0 0 Artocarpus integer Moraceae √ 3 √ 1 Artocarpus lanceifolius Moraceae √ 2 0 0 Artocarpus nitidus Moraceae √ 3 0 0 Artocarpus spp. Moraceae √ 4 √ 2 Baccaurea bracteata Euphorbiaceae √ 3 0 0 Baccaurea macrocarpa Euphorbiaceae √ 6 0 0 Baccaurea spec. Euphorbiaceae √ 8 √ 2 Baccaurea stipulata Euphorbiaceae √ 1 0 0 Borassodendron borneensis Palmae √ 7 √ 8 Canarium spp. Burseraceae √ 4 0 0 Crypteronia spec. Crypteroniaceae √ 7 √ 1 Cryptocarya spec. Lauraceae √ 3 0 0 Dacryodes costata Burseraceae √ 3 √ 1 Dacryodes rostrata Burseraceae √ 7 √ 1 Dacryodes rugosa Burseraceae √ 2 √ 1 Dacryodes spec. Burseraceae √ 5 √ 4 Dialium indum Ceasalpiniaceae √ 4 0 0 Dialium platysepalum Ceasalpiniaceae √ 1 0 0 Dialium spec. Ceasalpiniaceae √ 4 0 0 Diospyros borneensis Ebenaceae √ 6 √ 4 Diospyros cf buxifolia Ebenaceae 0 0 √ 1 Diospyros spec. Ebenaceae √ 7 √ 4 Durio dulcis Bombacaceae √ 1 √ 1 Durio graveolens Bombacaceae √ 1 0 0 Durio kutejensis Bombacaceae √ 1 0 0 Durio oxleyanus Bombacaceae √ 3 √ 1 Dysoxylum spec. Meliaceae √ 2 √ 1 Eugenia spec. Myrtaceae √ 2 0 0 Eugenia tawahense Myrtaceae √ 5 √ 3 Effects of ﬁres on sun bear fruit resources 101

Appendix 1: Continued

Genus Species Family UBF-edge No of plots BF No of plots Ficus spp Moraceae √ 8 √ 1 Garcinia parvifolia Guttiferae √ 4 0 0 Garcinia spp Guttiferae √ 4 √ 2 Glochidion spec. Euphorbiaceae √ 2 0 0 Ilex cymosa Aquifolicaeae √ 1 √ 1 Irvingia malayana Simaroubaceae √ 1 √ 4 Lansium domesticum Meliaceae √ 1 √ 1 Lansium spec. Meliaceae √ 1 0 0 Lithocarpus gracilis Fagaceae √ 1 0 0 Lithocarpus spp. Fagaceae √ 8 √ 2 Litsea ﬁrma Lauraceae √ 4 0 0 Litsea spp. Lauraceae √ 7 0 0 Madhuca kingiana Sapotaceae √ 6 √ 4 Magnolia lasia Magnoliaceae √ 2 0 0 Mangifera macrocarpa Anacardiaceae √ 1 0 0 Mangifera spp. Anacardiaceae √ 1 √ 1 Monocarpia kalimantanensis Annonaceae √ 2 0 0 Nephelium spec. Sapindaeae √ 5 √ 1 Ochanostachys amentaceae Olacaeae √ 9 √ 2 Oncosperma horridum Palmae √ 6 0 0 Palaquium spp. Sapotaceae √ 9 √ 3 Polyalthia lateriﬂora Annonaceae 0 0 √ 1 Polyalthia laterifolia Annonaceae √ 2 0 0 Polyalthia rumphii Annonaceae √ 3 √ 1 Polyalthia spec. Annonaceae √ 5 √ 2 Polydocarpus spec. Palmae √ 3 √ 2 Prunus volgens Rosaceae √ 6 0 0 Pternandra spec. Melastomataceae √ 8 √ 2 Quercus spp. Fagaceae √ 2 0 0 Sandoricum spp. Meliaceae √ 3 0 0 Santiria cf apiculata Burseraceae √ 1 0 0 Santiria spp. Burseraceae √ 6 √ 2 Santiria tomentosa Burseraceae √ 5 0 0 Syzygium spp. Myrtaceae √ 9 √ 7 Tetramerista glabra Tetrameristaceae 0 0 √ 2 Xerospermum spec. Sapindaeae √ 1 0 0 Unburned edge plots: 66 species Burned plots: 37 species Chapter 5

Human-sun bear conflicts in East Kalimantan, Indonesian Borneo

Ursus (2005), 16: 130-137 104 Chapter 5

Abstract Interviews with farmers in 5 communities situated along the edge of the Sungai Wain Protection Forest, East Kalimantan, Indonesian Borneo, indicated that crop damage caused by sun bears (Helarctos malayanus) was higher than normal following the 1997–98 El Niño South- ern Oscillation Event. Widespread drought and forest fires reduced habitat and fruit availability for sun bears on the islands of Borneo and Sumatra. The main source of antagonism toward bears resulted from the damage they caused to stands of old coconut trees (Cocos nucifera), thereby frequently killing the trees. This prompted farmers to seek removal of the bears. Bear damage to annual crops generally spurred a less hostile reaction. Experiments with metal sheeting affixed to the trunks of coconut trees to deter climbing by bears were successful, at least in the short term (≥3 years). Inexpensive and easily applicable crop-protection devices such as this could help protect sun bears in the future, as increased human-bear conflicts are anticipated due to rapid human population growth, unabated forest destruction and fragmentation, and increased susceptibility of remaining forests to future fires. Human-sun bear conflicts 105

Most species of bears are opportunistic omnivores that may become considered pests when attracted to human-related foods. North American bears (grizzly bears Ursus arctos and American black bears U. americanus) are known to use apiaries, crops, orchard fruits, garbage, and livestock (Ambrose and Sanders 1978, Knight and Judd 1983, Garshelis et al. 1999). They also may afflict considerable damage to timber stands (Stewart et al. 1999). In Japan, Asiatic black bears (U. thibetanus) raid crops, orchards, and fish farms (Huygens and Hayashi 1999). Sloth bears (Melursus ursinus) have been reported to do considerable damage to sugarcane and groundnut plantations (Iswariah 1984). Andean bears (Tremarctos ornatus) in South America have been reported to predate on livestock (Goldstein 2002). Until a few decades ago bounties were commonly used as a means of reducing or eliminating bears to protect crops or livestock (Azuma and Torii 1980, Swenson et al. 1994, Mattson and Merrill 2002). In Southeast Asia, sun bears probably commenced crop-raiding when attractive foods were first planted close to forest habitat. Early reports from colonialists in Indonesia described ways of deterring or killing marauding bears in fruit plantations (O-Viri 1925), even when adjacent forest habitat was still extensive. In recent years, the combined effects of timber harvesting and forest fires have significantly reduced forest coverage in Kalimantan (Curran et al. 2004, Fuller et al. 2004). Increased human encroachment on Indonesian forests have led to increased human-wildlife conflicts (Meijaard 1999, Rijksen and Meijaard 1999), although little information is available on conflicts specifically with sun bears. On the islands of Borneo and Sumatra, sun bear habitat has recently been severely reduced or damaged by forest fires linked to the 1997–98 El Niño Southern Oscillation (ENSO) event. Approximately 5.2 million ha of land, of which 2.6 million ha was forest, were burned over a period of 4 months in the province of East Kalimantan alone (Siegert et al. 2001). Sun bear fruit resources (mainly tree-borne fruits) declined significantly in burned forests, with tree mortality (≥10 cm dbh) reaching >90% in certain areas (van Nieuwstadt 2002, G. Fredriksson, unpublished data). Insects, an alternative bear food (Wong et al. 2002), were also reduced significantly (G. Fredriksson, unpublished data). Suitable sun bear habitat in these fire-affected areas has become progressively fragmented, increasing the chances of edge-related conflict with humans (Woodroffe and Ginsberg 1998). After the large-scale forest fires in 1997–98 a widespread fruiting failure prevailed for more than one year. Although some fruits (e.g. Ficus spp.) are generally available year-round, abundant fruiting in forests in this region occurs at intervals of about 2-10 years (Medway 1972, Ashton et al. 1988, Curran and Leigthon 2000). The combination of rapid loss of habitat as well as inter-annual shortages of food may increasingly compel sun bears to seek nearby human food 106 Chapter 5

sources, especially crops planted along the forest edge. The objectives of this study were: (1) to determine the types and extent of crop damage caused by sun bears; (2) to assess the reactions of farmers to different types of bear-related crop damage; and (3) to ﬁnd ways of alleviating the most disturbing types of sun bear damage.

Study Area The study was carried out in 5 farmer communities along the southern and eastern periphery of the Sungai Wain Protection Forest (SWPF), near Balikpapan, East Kalimantan, Indonesian Borneo (1º 16’ S and 116º 54’ E) (Fig. 1). The reserve covers a lowland dipterocarp forest watercatchment area of approximately 10,000 ha. Forest fires entered the SWPF in early March 1998, initially from a neighboring state- owned logging concession, but subsequently also from surrounding agricultural fields, affecting some 60% of the reserve. After the fires in 1998, connections to forest areas north and west of the reserve were significantly reduced. To the south and east the reserve is bordered by agricultural plots, unproductive grassland and shrublands. At the time this study commenced, the SWPF consisted of approximately 40 km2 primary forest, 40 km2 regenerating burned forest, and 20 km2 affected by human encroachment (Fredriksson 2002).

Methods

Interviews Semi-structured household interviews were conducted in 5 farmer communities around the SWPF (Fig. 1) to compile information on sun bear damage to crops and orchards. Interviews were initiated just after the 1998 ﬁres, and continued in 1999 and 2000. Each year, my assistants and I interviewed 99 farmers representing 40% of the 246 families in these communities. The number of farmers interviewed in each community was relative to the number of inhabitants/farmers in that community. We attempted to carry out the same number of interviews annually in each community, though the identity of interviewees differed slightly over the years. We focused on farmers who frequently visited or worked in their gardens, as opposed to landowners who would only visit their orchards on an irregular basis. In one community all families relied on farming for their income, whereas in the remaining 4 communities approximately 80% of families did. We posed questions regarding ethnic origin of farmers, farming history, farming practices, types and amount of crops grown, and location of the farm Human-sun bear conﬂicts 107

Figure 1. Location of the 5 farmer communities surveyed adjacent to the Sungai Wain Protection Forest, East Kalimantan, Indonesia.

in relation to the forest edge. Crop damage would be recorded through interviews as detailed as possible, with information on species and quantity of crops fed upon, number of trees damaged and type of damage, frequency of bear visits, bear crop raiding behavior, damage to crops by other wildlife species, and methods used to reduce wildlife-related crop damage. Whenever possible we directly observed crop damage in the gardens.

Damage mitigation trials We attempted to reduce sun bear consumption of farmer’s fruits through conditioned taste aversion with thiabendazole (TBZ). TBZ-induced conditioned taste aversion has been used to reduce consumption of human-related foods and livestock by American black bears 108 Chapter 5

(Ternent and Garshelis 1999) and several species of canids (e.g. Gustavson et al. 1983). TBZ powder (Sigma Chemical, Saint Louis, Mo) was mixed into samples of ripe fruit (14-23 mg TBZ/g food) that were targeted by sun bears in gardens or orchards. TBZ was mixed thor- oughly in the soft, ripe pulp of breadfruit (Artocaprus heterophyllus), pineapple (Ananas sp.), and papaya (Carica papaya). Diluted TBZ, as opposed to TBZ powder, was injected into snakefruit (Salacca spp.) because it has a hard mesocarp. Treated fruits were placed where farmers indicated that sun bears had entered their gardens recently. We attempted to inhibit sun bears from climbing and damaging coconut palms by nailing smooth metal sheeting around their trunks, between the heights of 0.5-1 m from the base extending 2-3 m up the trunk. Two types of metal sheeting were used: new zinc sheets as well as recycled metal sheets produced from old food containers. The price of the latter was 50% cheaper/m2 than zinc plates. In one garden all remaining undamaged coconut trees (n=15) were protected with metal sheeting. In ﬁve gardens, which had been repeatedly targeted by sun bears, 2-4 coconut trees were protected with metal sheeting interspersed among unprotected palm trees. As one of the aims was to test a cheap method how to discourage sun bears from damaging trees, farmers were involved in placement of the metal sheeting. During subsequent interviews protected trees were monitored for success rate of the metal sheeting in deterring bears from climbing to them, as well as condition of the sheeting. Unprotected coconut trees growing in the vicinity of protected trees (n=75) were monitored simultaneously.

Results

Farmer profiles and practices Most farmers near the SWPF were immigrants originating from other islands (e.g. Sulawesi, Java), with only 18% originating from Borneo (Dayak or Pasir tribes). The average size of gardens, owned or leased, was 2.6 = 0.4 (SD) ha. Farmers had lived in the area on average for 20 = 7 (SD) years. Mixed orchards (43%), snakefruit plantations (24%), bananas (8%, Musa spp.), vegetables (7 %, e.g. cassava [Manihot esculenta], beans [Fabaceae], spinach [Basellaceae]), rice (Poaceae, 7%) and coconut trees (Palmae, 6%), constituted the main crop coverage by area. Snakefruit is a continuously fruiting low-growing palm species, and several tree species in the mixed orchards also produced fruits throughout the year, even though their wild congeners in the forest were more seasonal. Human-sun bear conﬂicts 109

Crop damage The main wildlife species reported to raid gardens throughout the study period were bearded pigs (Sus barbatus, 98% of gardens), followed by sun bears (43%), barking deer (Mun- tiacus spp., 42%), civets (20%, [Viverridae]), squirrels (12%, [Sciuridae]), and a variety of other species including macaques and pythons (raiding chicken coops). Nearly one-quarter (22%) of farmers reported that sun bears raided their gardens before the 1997–98 forest fires and fruiting failure. This increased to over half (56%) in 1998 and 1999, the first two post-fire years, and declined to 39% in the year 2000. Sun bears most often damaged coconut trees (36%), followed by snakefruit (32%), and a variety of fruits in mixed orchards (breadfruit [Moraceae] 12%; durian [Bombacaceae] 8%; rambutan [Sapindaceae] 8%; mango [Anacardiaceae] 4%). Damage to coconut trees was particularly disliked by farmers because sun bears fed primarily on the growthshoot (palmite), frequently killing over 20 years old fruit-bearing trees (Fig. 2). Only on rare occasions did sun bears feed on the coconut fruits. Farmers developed less antagonism toward bears that fed on ripe

Figure 2. Coconut palm tree damaged (and probably killed) by a sun bear in a garden adjacent to the Sungai Wain Protection Forest, East Kalimantan, Indonesia 1998 (photo courtesy D. & J. Garshelis). 110 Chapter 5

fruits from orchards or snakefruit, because trees or snakefruit palms were rarely damaged and usually only a small proportion of the overall crop was consumed by bears. The only account of attempted livestock depredation involved a sun bear trying to break into a chicken coop. Farmers generally considered financial losses related to sun bear damage to be low and few farmers expressed the feeling that financial compensation would be justified. Moreover, no active wildlife authority existed in the district to whom farmers could forward any complaints or damage claims. In a few cases, where entire coconut stands had been killed by bears or when a large proportion of a small stand of snakefruit had been consumed by a bear, farmers asked our survey team for compensation or removal of the bear.

Mitigation practices and trials Most farmers (57%) used dogs to guard their gardens, followed by small wooden fences (32%), and regular nocturnal check-ups of their gardens (6%). None of these were efﬁcient in keeping sun bears out as bears still entered these gardens. Poisoned fruit baits were put out by a small number of farmers. Although these poisoned baits were primarily targeted against bearded pigs, a variety of wildlife could be killed by this. Some farmers surrounded their garden with wire neck snares, mainly for pigs and barking deer, but a few also put out locally made foot snares designed to capture bears. Two farmers indicated that they had looked into hiring a hunter who could spear a bear climbing down from a coconut tree. Only 2 farmers admitted to killing bears before this study period, but this type of information was difﬁcult to obtain during interviews, as such killing is illegal and farmers may have feared prosecution. Sun bears did not consume any of the 15 TBZ-treated fruit samples that we placed in gardens in an attempt to create conditioned taste aversions toward garden fruits. None of the ~30 coconut trees that we protected with metal sheeting were subsequently climbed by bears. Several unprotected trees (n=27) growing in proximity to these protected trees were still climbed and damaged by bears. An additional small number of coconut trees were covered with metal sheeting by farmers themselves and none of these were subsequently climbed by bears, at least during the study period.

Sun bear crop-raiding behavior All farmers that reported sun bear visits to their gardens stated that sun bears entered their gardens during the night. Frequently sun bears built nests in small orchard trees, usually close to the main trunk, 2–5 m from the ground. These nests were created by bears breaking surrounding branches toward them and constructing a ‘V’ shaped structure (in contrast to the Human-sun bear conﬂicts 111

ﬂat platforms usually out on limbs constructed by orangutans [Pongo pygmaeus], which also live in the surrounding forest). These nests appeared to function as resting platforms rather than feeding platforms, which were occasionally encountered in fruiting trees within the forest. Such nests in orchards possibly provide some security, being off the ground, and by offering a wider olfactory range to detect approaching humans or dogs. Bears probably left these nests before daybreak as no farmer reported seeing bears in the morning.

Discussion

Main causes for human-carnivore conflicts concern the perceived risk of predation on humans and livestock (Sillero-Zubiri and Laurenson 2001, Treves and Karanth 2003). In places where bears are feared and hated, it is generally for either or both of these reasons (polar bear Ursus maritimus: Gjertz and Persen 1987; brown bear and American black bear: Herrero 1985; sloth bear: Rajpurohit and Krausman 2000; Asiatic black bear: Chauhan 2003; Andean bear: Goldstein 2002). No reliable reports of sun bears killing humans or livestock were found in the literature or discovered during this study from local informants. Hence, there has been little negative publicity regarding this species. During this study, however, some local farmers found them to be troublesome pests, especially in the 2 years after the fires and subsequent fruiting failure in the neighboring forest. Although bearded pigs were the main agricultural pest in the gardens around the SWPF, farmers were less antagonistic toward this species than toward sun bears. This was because bearded pigs primarily targeted annual crops (corn, cassava, pineapples), whereas bears frequently killed productive coconut trees by feeding on the palmite. If bears had just eaten the coconuts, the effect would have been much less detrimental, and even could have been benefi- cial to fruit production (Siex and Strushaker 1999). Few recent reports have been written about sun bear crop raiding. Fetherstonhaugh (1940: p. 21) reported that “the Malayan sun bear is an inoffensive jungle dweller and unlike some species, conflicts very little with human activities when it comes in contact with cultivation, the glaring exception being coconuts to which the bears are a positive menace … but there is nothing to fear from their presence near other forms of cultivation”. O-Viri (1925) on the other hand gives a lengthy description of how sun bears were devastating pests in coconut plantations where they fed on the palmite, as well as causing damage to papaya plantations, sugarcane, pineapple and fruit orchards. Several Dutch colonial sources mention damage in planta- 112 Chapter 5

tions due to sun bear depredations, especially to coconut stands (van Balen 1917, Feuilletau-de Bruyn 1933, Nederlandsch-Indische Vereeniging 1939). In other parts of Borneo and Sumatra, sun bears have been reported to enter sugarcane fields (L. Nyagang, local resident, personal communication), and more recently to feed on fruits in oil palm plantations (Nomura 2003; T. Maddox, Zoological Society of London, personal communication). Sun bears were frequently killed by colonial plantation owners at the turn of the century by various means, the most effective being the shooting of bears when they would climb down a fruit tree. Poisons, primarily phosphor, were already used to kill nuisance pigs in those days, and although meant to target sun bears as well, these were rarely effective (O-Viri 1925). TBZ trials attempted during this study were unsuccessful, possibly either because of the availability of a wide array of untreated fruits in gardens or due to human or chemical scent residues on treated fruits which repelled the bears. Reducing availability of crops near forested areas is likely to be the most effective means of mitigating human-wildlife conflicts. Naughton-Treves et al. (1998) recommended that, in general, crops attractive to wildlife should be planted >500 m from the forest edge, and that non-palatable plant barriers should be planted between the forest and human agricultural fields. This recommendation would not be feasible around Sungai Wain, where 250+ farmers would need to be translocated and their fields replaced with unpalatable vegetation. High human population densities, and the complicating factors linked to landownership and compensation issues obstruct these mitigation options in the Indonesian context. Electric fencing has been found to be an effective deterrent against damage to agricultural crops and apiaries in North America and Japan (Jonker et al. 1998, Garshelis et al. 1999, Huygens and Hayashi 1999). We did not experiment with electric fences, as this type of mitigation was not deemed applicable for small-scale private farmers in Indonesia. Farmers in Borneo or Sumatra would not have funds or easy access to procure the necessary materials. This method though, should be tested in the future as it could be effective for mitigating sun bear-human conflicts near certain protected areas or commercial plantations. Compensation for financial losses due to bear depredation has been used on private lands in North America and Europe (Cozza et al. 1996, Wagner et al. 1997). Although popular with the public, compensation does not address the source of the problem (Witmer and Whit- taker 2000). In some cases translocation or removal of the nuisance bears has been applied (e.g. Azuma and Torii 1980, Garshelis 1989). Neither of these mitigation procedures would currently apply to the Indonesian context as forest conservation and wildlife management have not reached that level of attention or public support. Human-sun bear conflicts 113

Figure 3. Balikpapan district logo, East Kalimantan, Indonesia, featuring the sun bear as its new mascot.

Increased human-wildlife conflicts in Indonesia are likely linked to the rapid reduction of forest habitats. We noticed a sharp increase in conflicts with bears shortly after much of the forest burned, coincident with a fruiting failure. Despite the increased problems that this caused, no government officials were available to assist local farmers with their wildlife conflicts. This led local farmers to feel that they had no choice but to attempt killing the nuisance animals themselves, even though this was prohibited by law. Often farmers requested help from our research team, and we worked with them to develop non-lethal means of protecting their crops. Metal sheeting placed around the trunks of large coconut trees functioned well to deter sun bears from climbing up to the growth shoot. These metal sheets can be obtained locally, at a low cost, and are easily installed. The sheeting will likely need to be replaced every 3–4 years, as it deteriorates quickly due to high humidity, but this appears to be manageable to 114 Chapter 5

these farmers. This method is likely to become more widespread as its effectiveness becomes known, and this process could be hastened through conservation education programs. During this study conservation education programs were initiated targeting schools, local communities and local government agencies and focusing on forest functions and the role of wildlife slowly brought about a local change in attitude towards forest conservation and sun bears. In 2001 the sun bear was chosen to become the mascot of the district where this study was carried out (Fig. 3). This seemed to instill a sense of pride and ownership of the species. In addition, a multi-stakeholder management body was established for management of the reserve, fully funded by the local government, where the issue of human-wildlife conﬂicts has ﬁnally received a place on the management agenda. It still remains to be seen, though, whether this slow attitude change towards bears and forest conservation will translate into behavioral changes and tolerance of low-level agricultural losses related to crop raiding, associated with living and farming at a forest edge.

Acknowledgements I would like to express my gratitude to the Indonesian Institute of Sciences (LIPI) for permitting me to do field research on sun bears, as well as to the Forest Research Insti- tute (BPK-Samarinda) for their research invitation and support. I would like to thank my field assistants Lukas Nyagang, Trisno, Rachman, Agusdin, Babam and Iman from Sungai Wain, and students Whiwien, Eko, Joned and Ali from the Mulawarman University, Samarinda, for their assistance with the surveys. Financial support for field work was received from the Tropenbos Foundation. D. Garshelis provided ongoing logistical and technical support for the study, provided useful comments on an earlier draft of the manuscript, as well as the TBZ for the taste- aversion trials. S.B.J. Menken, V. Nijman and M. Lammertink, O. Huygens and E. Meijaard are also thanked for useful comments on earlier drafts of this paper.

Literature cited Ambrose, J.T., and O.T. Sanders. 1978. Magnitude of black bear depredation on apiaries in North Carolina. Proceedings of the Eastern Black Bear Workshop 4:167–177. Ashton, P.S., T.J. Givnish, and S. Appanah. 1988. Staggered Flowering in the Dipterocarpaceae - New Insights into Floral Induction and the Evolution of Mast Fruiting in the Aseasonal Tropics. American Natu- ralist 132:44–66. Human-sun bear conﬂicts 115

Azuma, S., and H. Torii. 1980. Impact of human activities on survival of the Japanese black bear. Int. Conf. Bear Res. and Manage. 4:171–179. Chauhan, N.P.S. 2003. Human casualities and livestock depredation by black and brown bears in the Indian Himalaya, 1989–98. Ursus 14:84–87. Cozza, K., R. Fico, M.L. Battistini, and E. Rogers. 1996. The damage-conservation interface illustrated by predation on domestic livestock in central Italy. Biological Conservation 78:329–336. Curran, L.M., and M. Leighton. 2000. Vertebrate responses to spatiotemporal variation in seed production of mast-fruiting dipterocarpaceae. Ecological Monographs 70:101–128. Curran, L.M., S.N. Trigg, A.K. McDonald, D. Astianti, Y.M. Hardiono, P. Siregar, I. Caniago, and E. Kasischke. 2004. Lowland forest loss in protected areas of Indonesian Borneo. Science 303:1000–1003. Fetherstonhaugh, A.H. 1940. Some notes on malayan bears. The Malayan Nature Journal 1:15–22. Feuilletau-de Bruyn, W. K. H. 1933. Bijdrage tot de kennis van de afdeeling Hoeloe Soengei (Z. en O.- afdeeling Borneo). Koloniale Studiën 17: 53-93 and 183-211(in Dutch). Fredriksson, G.M. 2002. Extinguishing the 1998 forest fires and subsequent coal fires in the Sungai Wain Protection Forest, East Kalimantan, Indonesia. Pages 74–81 in P. Moore, D. Ganz, L.C. Tan, T. Enters, and P.B. Durst, eds. Communities in flames: proceedings of an international conference on community involvement in fire management. FAO and FireFight SE Asia, Bangkok, Thailand. Fuller, D.O., T.C. Jessup, and A. Salim. 2004. Loss of forest cover in Kalimantan, Indonesia, since the 1997– 1998 El Nino. Conservation Biology 18:249–254. Garshelis, D.L. 1989. Nuisance bear activity and management in Minnesota. Bear-people conflicts: Pro- ceedings of a symposium on management strategies Northwest territories Dept. of Renew.

Res.:169–180. Garshelis, D.L., R.S. Sikes, D.E. Andersen, and E.C. Birney. 1999. Landowners’ perceptions of crop damage and management practices related to black bears in east-central Minnesota. Ursus 11:219–224. Gjertz, I., and E. Persen. 1987. Confrontations between humans and polar bears in Svalbard (Norway). Polar Research 5:253–256. Goldstein, I. 2002. Andean bear-cattle interactions and tree nest use in Bolivia and Venezuela. Ursus 13:369–372. Gustavson, C.R., J.C. Gustavson, and G.A. Holzer. 1983. Thiabendazole-based taste aversion in dingoes (Canis familiaris dingo) and New Guinea wild dogs (Canis familiaris hallstromi). Applied Animal Ethol- ogy 10:385–388. Herrero, S. 1985. Bear attacks: Their causes and avoidance. Nick Lyons Books, New York, USA. Huygens, O.C., and H. Hayashi. 1999. Using electric fences to reduce Asiatic black bear depredation in Nagano prefecture, central Japan. Wildlife Society Bulletin 27:959–964. 116 Chapter 5

Iswariah, V. 1984. Status survey report and recommendations for conservation of the sloth bear in Ramna- garam Taluk, Karnataka. World Wildlife Fund-India, Bangalore. Jonker, S.A., J.A. Parkhurst, R. Field, and T.K. Fuller. 1998. Black bear depredation on agricultural commodities in Massachusetts. Wildlife Society Bulletin 26:318–324. Knight, R.R., and S.L. Judd. 1983. Grizzly bears that kill livestock. Int. Conf. Bear Res. and Manage. 5:186–190. Mattson, D.J., and T. Merrill. 2002. Extirpation of Grizzly bears in the contiguous United States, 1850–2000. Conservation Biology 16:1123–1136. Medway, L. 1972. Phenology of a tropical rain forest in Malaya. Biological Journal of the Linnean Society 4:117–146. Meijaard, E. 1999. Human-imposed threats to sun bears in Borneo. Ursus 11:185–192. Naughton-Treves, L., A. Treves, C. Chapman, and R. Wrangham. 1998. Temporal patterns of crop-raiding by primates: linking food availability in croplands and adjacent forest. Journal of Applied Ecology 35:596–606. Nederlandsch-Indische Vereeniging tot Natuurbescherming. 1939. Natuur in Zuid- en Oost- Borneo. Fauna, ﬂora en natuurbescherming in de Zuider- en Ooster-Afdeeling van Borneo. 3 Jaren Indisch natuur leven. Opstellen over landschappen, dieren en planten. Elfde verslag (1936-1938), Batavia, Indone- sia. Pages 334-411 (in Dutch). Nomura F. 2003. Seasonal trends of habitat usage in two ursine species, a destructive beast Ursus arctos yesoensis and an agricutural pest Helarctos malayanus. PhD thesis. University of Hokkaido: 1-46. O-Viri. 1925. Tusschen de beren. de Tropische Natuur XIV:81–88. Rajpurohit, K.S., and P.R. Krausman. 2000. Human-sloth bear conﬂicts in Madhya Pradesh, India. Wildlife

Society Bulletin 28:393–399. Rijksen, H.D., and E. Meijaard. 1999. Our vanishing relative. The status of orangutans at the close of the centrury. Tropenbos, Kluwer Academic Publishers, Dordrecht, Boston, London. Siegert, F., G. Ruecker, A. Hinrichs, and A.A. Hoffmann. 2001. Increased damage from fires in logged forests during droughts caused by El Niño. Nature 414:437–440. Siex, K.S., and T.T. Struhsaker. 1999. Colobus monkeys and coconuts: a study of perceived human-wildlife conflicts. Journal of Applied Ecology 36:1009–1020. Sillero-Zubiri, C., and K. Laurenson. 2001. Interactions between carnivores and local communities: conflict or co-existence? Pages 282–312 in J. L. Gittleman, S. M. Funk, D. W. MacDonald, and R. K. Wayne, eds. Carnivore conservation. University of Cambridge Press, Cambridge, UK. Stewart, W.B., G.W. Witmer, and G.M. Koehler. 1999. Black bear damage to forest stands in Western Wash- ington. Western Journal of Applied Forestry 14:128–131. Human-sun bear conflicts 117

Swenson, J.E., F. Sandegren, A. Bjarvall, A. Soderberg, P. Wabakken, and R. Franzen. 1994. Size, trend, distribution and conservation of the brown bear Ursus arctos population in Sweden. Biological Conserva- tion 70:9–17. Ternent, M.A., and D.L. Garshelis. 1999. Taste-aversion conditioning to reduce nuisance activity by black bears in a Minnesota military reservation. Wildlife Society Bulletin 27:720–728. Treves, A., and K.U. Karanth. 2003. Human-carnivore conflict and perspectives on carnivore management worldwide. Conservation Biology 17:1491–1499. van Balen, J. H. 1914. De dierenwereld van Insulinde in woord en beeld. I. De Zoogdieren. W.J. Thieme & Cie, Zutphen, the Netherlands (in Dutch). van Nieuwstadt, M.G.L. 2002. Trial by fire-Postfire development of a dipterocarp forest. PhD thesis. Uni- versity of Utrecht, Utrecht. Wagner, K.K., R.H. Schmidt, and M.R. Conover. 1997. Compensation programs for wildlife damage in North America. Wildlife Society Bulletin 25:312–319. Witmer, G.W., and D.G. Whittaker. 2001. Dealing with nuisance and depredating black bears. Pages 73–82 in D. Immell, editor. 7th Western Black Bear Workshop. Wong, S.T., C. Servheen, and L. Ambu. 2002. Food habits of Malayan sun bears in lowland tropical forests of Borneo. Ursus 13:127-136. Woodroffe, R., and J.R. Ginsberg. 1998. Edge effects and the extinction of populations inside protected areas. Science 280:2126–2128. Chapter 6

Predation on sun bears by reticulated python in East Kalimantan, Indonesian Borneo

The Raffles Bulletin of Zoology (2005), 53(1): 165-168 120 Chapter 6

Abstract

The Malayan sun bear is the largest member of the order Carnivora on the island of Borneo. Few records exist of predation on this species beside humans, whereas accurate re- cordings of natural predation events can teach us about the ecology of the prey species. Here I report on an attempted and a successful predation of Malayan sun bears by a reticulated python, both in a lowland dipterocarp forest in East Kalimantan. The successful predation was accomplished by a ~7 m reticulated python. The python preyed and swallowed an adult female sun bear, possibly weakened at the time due to a fruiting failure and nursing of a cub. Both predation events occurred at night, with the python probably surprising the bears during their sleep. Sun bear predation by reticulated python 121

Introduction

Throughout the world bears are at the top of the food chain with few natural predators, apart from man and congeners, which are capable of killing and eating an adult individual (e.g. Garshelis, 2004). The Malayan sun bear is the smallest bear species, with the Bornean subspecies (Helarctos malayanus euryspilus) weighing between 20-40 kg for females and 30-60 kg for males in the wild (Fredriksson unpubl. data; F. Nomura pers. comm.; Wong et al., 2004). The distribution range of the sun bear covers most of tropical mainland Southeast Asia’s forests as well as the islands of Sumatra and Borneo (Servheen, 1999). Sun bears are primarily diurnal (Wong et al., 2004; Fredriksson, in prep.), though reportedly more nocturnal in forests with much human trafﬁc (Grifﬁths and van Schaik, 1993). Usually sun bears are encountered solitary and the most common social grouping is a female with cub(s), though infrequently 3 bears have been sighted together (Fredriksson, unpubl. data.). Sun bears are mainly terrestrial though expert tree climbers, with a diet that primarily comprises of insects and fruits (Wong et al., 2002; Fredriksson, in prep.). The main predator of sun bears throughout its range is by far man (Meijaard, 1999; Fredriksson, in review). Tigers and other large felines are also potential predators (e.g. Kawani- shi and Sunquist, 2004). Here I report on a predation of a wild female sun bear, radio-collared at the time for an ecological study, by a reticulated python, and another unsuccessful predation attempt, in East Kalimantan, Indonesian Borneo. The reticulated python (Python reticulatus, Pythonidae), is thought to be the world’s largest or second largest snake (Shine et al., 1999), or at least the world’s longest snake (Mur- phy and Henderson, 1997). They inhabit tropical rainforests of Southeast Asia from Myanmar (Burma) to most of the islands of the Philippines and Indonesia (Auliya and Abel, 2000a; Auliya, 2003a). The longest recorded length of a python comes from the island of Sulawesi, Indonesia, where a reticulated python measuring 10.05 m was caught (Raven, 1946; Murphy and Hender- son, 1997). Females can attain much larger body sizes than males when mature (Shine et al., 1998), but sexual dimorphism varies between areas (Shine et al., 1999). Large (>6 m) pythons are rarely encountered and during Shine et al’s (1999) study of reptile slaughterhouses few were recorded. Very large specimens may be relatively scarce as well in other natural python populations (Bhupathy, 1990). Huge pythons can attain a body mass of up to 150 kg (Pope, 1975). The reticulated python has a broad head and a huge gape enabling it to swallow large prey. Long curved teeth ensure that once the snake has caught its prey, it rarely loses grip. The snake is a powerful constrictor and when adult it can overpower and kill animals as large 122 Chapter 6

as pigs, deer, and dogs (Auliya, 2003a, 2003b). Prey are killed by asphyxiation, looping its body around the victim, squeezing until it stops breathing. The reticulated python is a nocturnal hunter that heavily relies on ambush tactics to catch prey. The thick and heavy body is usually firmly anchored when the python strikes. Due to their good camouflage prey may come close without detecting them. Pythons have sensory organs (infrared-sensitive labial pits) in order to locate prey accurately, even in total darkness. Pythons allegedly often wait in ambush at a spot where wildlife frequently passes by (Slip and Shine, 1988), although it might be possible that prey are traced during inactivity (Auliya, 2003a). Despite their massive size, a reticulated python is surprisingly hard to spot due to its excellent camouflage and partly because it is less active during the day and remains extremely silent. Prey size increases with larger body size (Shine et al., 1998, 1999; Auliya, 2003a), with small pythons mainly feeding on rats but shifting to larger mammals at 3-4 m body length, depending on prey availability (Shine et al., 1998; Auliya, 2003a). Prey recorded for pythons in Indonesia range from small mammals like rats (Muridae) and shrews (Soricidae), to larger animals like civets (Viverridae), pangolins (Manidae), porcupines (Hystricidae), binturong (Arctictis binturong), primates, wild pigs (Suidae), as well as domesticated prey like chickens, dogs, cats in agricultural/urban areas (Shine et al., 1998; Auliya and Abel, 2000b). It is one of the few snakes in the world known to eat humans occasionally. Kopstein (1927) reports on a 14-year old boy eaten by a 5.17 m reticulated python, and Schmidt (1998 in Auliya, 2003a) reports of a 32 year-old man eaten by a 7 m python.

Study Site

The reported case of sun bear predation took place in the Sungai Wain Protection For- est, near Balikpapan, East Kalimantan, Indonesian Borneo (1º 05’ S and 116º 49’ E). The reserve covers a watercatchment area of circa 10,000 ha. The topography of the reserve consists of gentle to sometimes steep hills, and is intersected by many small rivers. The area varies in altitude from 30 to 150 metres a.s.l. The most common tree families above 10 cm dbh (diameter at breast height) are Euphorbiaceae, Dipterocarpaceae, Sapotaceae and Myrtaceae. Poaching is rare and primarily restricted to the borders of the reserve (Fredriksson and de Kam, 1999). The reserve contains relatively large populations of potential prey species for pythons. Bearded pigs (Sus barbatus), barking deer (Muntiacus spp.), and mouse deer (Tragulus spp.) are common (pers. obs.), as well a primarily ground-dwelling pig-tailed macaques (Macaca nemestrina). Civets, pangolins (Manis javanica) and other ground dwelling mammals like porcupines are also present. Sun bear predation by reticulated python 123

Predation attempt

On June 21, 1999, a probable python attack occurred on a 3-year-old, 31-kg female sun bear. This bear had been conﬁscated as a cub and subsequently taken to the forest where I had raised her for several months before releasing her in the study area with a radiocollar, 1.5 years prior to the predation event. The bear’s activity was being monitored continuously that day, and it had been sleeping for several hours when, at 01:30, 4 loud barks were heard, suggesting that it was involved in an agonistic encounter. Afterwards, the bear was heard growling for 30 min up in a tree. The next morning the bear was observed in a tree nest, with blood drops covering nearby undergrowth. At 18:00 it climbed down and could be examined more closely (having been raised in captivity, the bear could be touched). The left side of the bear’s face was swollen with a row of small superﬁcial scabs found on the skin. On both back feet 2 claws each were torn out, causing the blood that was encountered on the undergrowth below the nest. Other- wise no obvious physical wounds could be found, nor were any signs of other animals (e.g. feline, pig) encountered at the place of attack. From the nature of the wounds and by eliminating all potential predators a python appeared the most likely attacker. The subsequent successful predation event occurred only 100 m from the site of this predation attempt.

Predation event

A wild adult female sun bear weighing 23 kg, with a small cub (1–2 months old), was trapped on July 6, 1999. This bear was in a poor physical condition (probably >10 kg under- weight), due to a prolonged fruit shortage at the time compounded with nutritional stress from nursing a small cub. The bear’s canines were worn down to stumps and her age was estimated to be >10 years. She was ﬁtted with an activity and mortality-sensing radiocollar (ATS, Isanti, Minnesota, U.S.A.). The bear was monitored on a daily basis, with locations obtained by triangulation, in the morning, midday and afternoon. A 24-hour activity monitoring was conducted once a week by listening and recording the pulse rate of the bear’s radio transmitter every 10 minutes. On the afternoon of July 30, the bear was located and found to be active. The next morning she was located 550m from the previous location, but her radio signal indicated that she had not moved for ≥4 hours, which was indicative of either mortality or a dropped radio 124 Chapter 6

Figure 1. Reticulated python (~7 m) lying in a small stream after having swallowed an adult female sun bear; the bulge on the left hand side in the python contains the sun bear.

collar. I tracked the collar to a swamp, and ultimately found that the signal was being emitted from the stomach of a large python, which was curled under a thicket. After being poked with a stick, it ﬂed into a nearby stream, producing, as it twisted, the sound of breaking bones (Fig 1). The bulge of the bear was ~2 m from the python’s head. No traces of a struggle were found near the site where the python was encountered nor any signs of the sun bear’s cub. The python remained resting in the stream until August 3, when it moved into a large hollow log (Shorea laevis), 120 m from the site where it was initially encountered. The log had an outer diameter of 153 cm and was hollow for approximately 8 m. The python was back about 5 m from the entrance. A few metres from the log entrance a large skin shedding of a python was found. The radiocollar remained functioning and hence the python’s movements could be monitored on a regular basis. The signal of the collar remained fully inactive for 26 days, while the python remained inside the log. On August 30 the python left the tree hollow and entered the swamp again. As the radiocollar contains batteries in a metal casing it was decided to monitor closely whether the python would regurgitate remains of the bear, including the radiocollar. Sun bear predation by reticulated python 125

Figure 2. Reticulated python being pulled out of an underground stream 1.5 month after it swallowed a radio-collared adult female Malay sun bear (photo courtesy M. van Nieuwstadt).

On September 8, I decided to catch the python and hold it in captivity to facilitate collection of regurgitated remains. Ten villagers were recruited and the python was caught manually and put in a steel-barred cage. That same night the python escaped by dislodging the bars of the cage through squeezing. The next day the python was caught again (Fig. 2), now encountered in a small underground stream. This time it was placed inside a barrel trap, designed for catching the sun bears. After holding the python for a month, it only passed remains of bones. Thus, on Oc- tober 29 I retrieved the collar surgically. The snake was sedated with zolazepam–tiletamine (Telazol® 3.75 mg/kg for an estimated weight of 80 kg). Full sedation took several hours due to initial low dosage of drugs (recommended dosage 20mg/kg, Kreeger et al., 2002). Once fully anaesthetized the python was palpated to ﬁnd the radiocollar and an incision was made in the skin and large intestine and the radiocollar removed. The python measured 6.95 m and weighed 59 kg, after having not eaten for nearly 3 months. Sex was not determined. The snake was held in captivity and monitored for another 3 weeks, twice being fed a chicken, and then released on November 20. 126 Chapter 6

Discussion

This is the first detailed record of predation of bears by a python, and could only have been recorded due to the fact that the bear was radiocollared. By its very nature, predation is a once in a life time experience, hence difficult to observe. Although anecdotal, accurate record- ings of predation events can teach us about the ecology of the prey species. The only other published case of a radio-collared mammal being eaten by a python was reported by Martin (1995), involving an amethystine python (Morelia amethystina) that preyed on a Bennett’s treekangaroo (Dendrogalus bennettianus) in North Queensland, Australia. Few published records exist of bear predation by other animals. Kurt and Jayasurya (1968) report of a sloth bear (Melursus ursinus) eaten by a leopard (Panthera pardus). Sloth bears, due to their less arboreal and more aggressive nature, might be less inclined than sun bears to climb a tree when threatened by a leopard; however, leopards are also expert tree climbers (Laurie and Seidensticker, 1977). A number of records exist of bears being predated upon by conspecifics (Ursus arctos, S. Brunberg. pers.comm.), or congenerics (Garshelis, 2004). In terms of animal predation on sun bears, few published records exist. Kawanishi and Sunquist (2004) report on 3 tiger (Panthera tigris) scats containing sun bear remains from peninsular Malaysia. Other predators on mainland Southeast Asia and Sumatra could potentially be the common leopard, and the clouded leopard (Neofelis nebulosa), which occur sympatrically with sun bears. The moon bear (Ursus thibethanus), also occurring sympatrically throughout parts of the sun bears range, being substantially larger than the sun bear, could potentially pose a risk to sun bears. A sun bear, with its long claws and powerful jaws with large canines, would be a formi- dable prey for a python. In the cases reported here, it appeared that the sun bears were probably caught by surprise while they were sleeping. It is possible that a python had come across the sleeping bear, which usually spend the night resting on a log or on the forest floor. Although it has been described that pythons attack from ambush, it is possible that the python followed a scent trail (M. Auliya, pers. comm.), came across the bear, bit it in the face (indicated by the scars from the predation attempt) while trying to constrict it. Moreover, in the successful predation event, the python was >3 times the mass of the bear. Animals much larger than bears, such as >60-kg wild pigs have been swallowed by pythons (Shine et al., 1998; Auliya, 2003b). Although little is known about what body size a python would need to attain in order to successfully overpower a sun bear, it seems likely that it will need to be a fairly large specimen. The python that predated on the sun bear during this study was large (~7 m), and Sun bear predation by reticulated python 127

densities of such large pythons are probably low, even more so as large pythons are becoming extremely rare due to unsustainable harvesting (Shine et al., 1999). Of >1000 python stomach contents examined in Sumatra, none contained remains of a sun bear (Shine et al. 1998), although predation of a sun bear by a python in the interior of Borneo was reported by a local Dayak (Domalian 1991 in Auliya and Abel, 2000b). It remains unclear whether the incidents reported here were truly rare events. During the course of my study, only 6 sun bears were radiocollared, 3 wild caught and 3 partially reared in captivity. One of the captive-reared bears was killed by a person, and one of the wild-caught bears apparently starved after a prolonged fruiting failure, subsequent to the 1997-1998 ENSO event. The fruiting failure possibly contributed to the death of the bear that was captured by the python, due to her weakened physical condition. Perhaps these predation events are not uncommon during such periods of food scarcity, which on Borneo occur at irregular intervals.

Acknowledgements I thank the Indonesian Institute of Sciences (LIPI) for granting me permission to carry out research on sun bears. Gratitude is expressed as well to my ﬁeld assistants: Adi, Nono, Moudy, Bahronni, Lukas and Damy and the villagers from Sungai Wain who assisted with catching the python. Financial assistance from the Tropenbos Foundation is gratefully acknowledged. The Conservation Endowment Fund of the American Zoo and Aquarium Association provided ﬁnancial assistance for purchasing the radiocollars. M. Auliya, V. Nijman, S. Menken and D. Garshelis made comments on an earlier draft. An anonymous reviewer is also thanked for valuable comments. M. van Nieuwstadt is thanked for the picture of the python capture.

Literature cited Auliya, M., and F. Abel, 2000a. Zur Taxonomie, geographischen Verbreitung und Nahrungsökologie des Netz- pythons (Python reticulatus) - Teil 1 (geographische Verbreitung). Herpetofauna 22(127): 5-18. Auliya, M., and F. Abel, 2000b. Zur Taxonomie, geographischen Verbreitung und Nahrungsökologie des Netz- pythons (Python reticulatus) - Teil 2 (Nahrungsökologie). Herpetofauna 22(128): 19-28. Auliya, M., 2003a. Taxonomy, Life History and Conservation of Giant Reptiles in West Kalimantan (Indonesian Borneo). Ph.D thesis, Universität Bonn, 513 pp. Auliya, M., 2003b. A reticulated python (Python reticulatus) preys on an adult Sulawesi pig (Sus celebensis). Asian Wild Pig News 3(1): 11-12. 128 Chapter 6

Bhupathy, S., 1990. Blotch structure in individual identification of the Indian python (Python molurus molurus Linn.) and its possible usage in population estimation. Journal Bombay Natural History Society 87:399-404. Fredriksson G. M., and M. de Kam, 1999. Strategic plan for the conservation of the Sungai Wain Protection For- est, East Kalimantan, Indonesia. The International Ministry of Forestry and Estate Crops-Tropenbos Kalimantan Project. Pp. 1-46. Fredriksson, G. M., In press. Human-sun bear conflicts in East Kalimantan, Indonesian Borneo. Ursus. Garshelis, D.L., 2004. Variation in ursid life histories - is there an outlier ? In: Lindburg D. and K. Baragona. (eds). Giant pandas. Biology and conservation. University California Press, Berkeley. Pp. 53-73. Griffiths, M., and C. P. van Schaik, 1993. Camera trapping: a new tool for the study of elusive rain forest mammals. Tropical Biodiversity 1:131-135. Kawanishi, K., and M. E. Sunquist, 2004. Conservation status of tigers in a primary rainforest of Peninsular Malaysia. Biological Conservation 120:329–344. Kopstein, F., 1927. Over het verslinden van Menschen door Python reticulatus. Tropische Natuur 16:65-67. Kreeger, T. J., J. M. Arnemo, and J. P. Raath, 2002. Handbook of wildlife chemical immobilization. Wildlife Pharma- ceuticals, Inc., Fort Collins, Colorado. Pp.1-402. Kurt, F., and A. Jayasuriya, 1968. Notes on a dead bear. Loris: 182-183. Laurie, A., and J. Seidensticker, 1977. Behavioural ecology of the sloth bear (Melursus ursinus). J. Zool., Lond 182:187-204. Martin, R. W., 1995. Field observation of predation on bennetts treekangaroo (Dendrolagus bennettianus) by an amethystine python (Morelia amethistina). Herpetological Review 26:74-76.

Meijaard, E., 1999. Human-imposed threats to sun bears in Borneo. Ursus 11:185-192. Murphy, J.C. and R.W. Henderson, 1997. Tales of Giant snakes – A historical natural history of Anacondas and Pythons. Krieger Publ. Comp., Malabar, Florida. Pp. 1-221. Pope, C. H., 1975. The giant snakes. Albert A. Knopf, New York. Pp.1-221. Raven, H. C., 1946. Adventures in python country. Natural History 55: 38-41. Servheen, C., 1999. Sun bear conservation action plan. In: Servheen, C., S. Herrero and B. Peyton (eds.), Bears. Status survey and conservation action plan. International Union for Conservation of Nature and Natural Resources, Gland, Switzerland and Cambridge, United Kingdom. Pp. 219-224. Shine, R., P. Ambariyanto, S. Harlow, and Mumpuni, 1999. Reticulated pythons in Sumatra: biology, harvesting and sustainability. Biological Conservation 87:349-357. Shine, R., P. S. Harlow, J. S. Keogh, and Boeadi, 1998. The inﬂuence of sex and body size on food habits of a giant tropical snake, Python reticulatus. Functional Ecology 12:248-258. Sun bear predation by reticulated python 129

Slip D.J., and R. Shine, 1988. Feeding habits of the diamond python, Morelia s. spilota: ambush predation by a boid snake. Journal of Herpetology 22: 323-330. Wong, S. T., C. Servheen, and L. Ambu, 2002. Food habits of Malayan sun bears in lowland tropical forests of Borneo. Ursus 13:127-136. Wong, S. T., C. Servheen, and L. Ambu, 2004. Home range, movement and activity patterns, and bedding sites of Malayan sun bears, Helarctos malayanus in the rainforest of Borneo. Biological Conservation. 119: 169-181. Chapter 7

Exploring the use of sign surveys for monitoring relative abundance of sun bears: a case study in burned and unburned forests in East Kalimantan, Indonesian Borneo

With D.L. Garshelis, W.J. Sastramidjaja and S.B.J. Menken 132 Chapter 7

Abstract

Sun bears, Helarctos malayanus, the largest carnivore on the island of Borneo, are threatened by habitat loss. Besides direct cutting and forest conversion, forest fires and drought, related to recurring El Niño Southern Oscillation (ENSO) events, have degraded vast areas of sun bear habitat on Borneo. In this study we explore the use of sign (e.g., claw-marked trees, raided termite nests) surveys as a way to monitor changes in relative abundance of sun bears following an ENSO-related forest fire in 1998. We compared sign densities to an adjacent lowland dipterocarp forest that was not burned. We searched for sign in 92 1-ha strip transects (45 in unburned forest and 47 in burned forest) in 2000, 2001, 2002, 2005 and 2010. Densities of sun bear sign were close to zero in burned forest areas for several years after the fire, but by 2010 had increased to 28.0 ± 12.9 (SD) sign/ha, indicating re-colonization by sun bears. Sign densities in adjacent unburned forest remained stable over the years, averaging 41.6 ± 11.6 (SD) sign/ha. Densities of sun bear sign were still 35% lower in burned forest, which showed a paucity of certain types of sun bear foods, especially fruiting trees and above-ground termite colonies. Nevertheless, it is evident that these once-burned forest areas can still play an important role in the conservation of sun bears, as long as source populations for re-colonization are present in adjacent undisturbed forest. We discuss the potential and challenges to further develop sign transects as a tool for monitoring population trends of this species. Exploring the use of sign surveys for monitoring relative abundance of sun bears 133

Introduction

Monitoring population abundance and trends is essential for the management and conservation of many wildlife species, including large carnivores (Kendall et al. 1992, Mowat and Strobeck 2000). These parameters are used to assess population status, evaluate the effects of management measures or decisions (Wilson and Delahay 2001), decide quotas for harvested populations (Kindberg et al. 2011), or to obtain parameters for conservation principles, including IUCN criteria for Red-listing evaluations (Kühl et al. 2008, Vié et al. 2009). For animals that are cryptic, occur at low densities but over extensive areas, and have large home ranges, it might be more practical to monitor indices of population size or trends in key areas, rather than obtain actual estimates of abundance. Bears represent one such taxon in which population estimation is expensive, impractical in some remote areas, and frequently lacks the precision necessary to monitor population trends (Garshelis and Hristienko 2006). Bears habitually leave signs of characteristic appearance that indicate their presence. These range from distinct claw marks on trees, various feeding sign, scats, footprints, and even feeding or sleeping platforms in trees. Relative abundance can be judged from the frequency of signs per standardized unit of sampling effort (Wilson and Delahay 2001), which is assumed to be related to densities of the species depositing that sign (Kühl et al. 2008). However, various other factors may affect sign density as well, including: sign production rates and seasonal movement of animals (e.g. Kutilek et al. 1983, Stanley and Bart 1991), persistence of sign (Bider 1968), and variation in detection of sign (Kühl et al. 2008). Sun bears, Helarctos malayanus, have not been surveyed systematically in Southeast Asian forests, but are thought to be in widespread decline (Fredriksson et al. 2008). It appears that sun bears have recently been extirpated in Bangladesh (Islam et al. 2010) and possibly China (Wang Dajun unpubl. report) and are declining in several protected areas throughout their range (Mei- jaard 1999, Steinmetz et al. 2006, Harrison 2011). The sun bear is a forest-dependent species and is threatened by high rates of habitat loss (Miettinen et al. 2011), fragmentation and degradation of remaining habitat, as well as poaching for body parts throughout much of its range (Foley et al. 2011). Indonesia, covering some 45% of the sun bear distribution range (IUCN/SSC Bear Specialist Group 2006), has the highest rates of deforestation in the region (Hansen et al. 2009). Sun bears have been protected by Indonesian law since 1973, but most of their habitat is not. Fires associated with El Niño/Southern Oscillation (ENSO) are recurring events in Indonesia. Within the past two decades, major ﬁres have occurred during ﬁve ENSO events 134 Chapter 7

(Meijaard and Dennis 1997, Barber and Schweithelm 2000). In Indonesia, the scale of ENSO- related forest fires has been extensive with more than 3.6 million ha of forest and scrub burned in 1982–83. The 1997–98 fires damaged at least 2.6 million ha of concession and conservation forests in the Indonesian province of East Kalimantan alone (Hoffman et al. 1999, Siegert et al. 2001, Tacconi 2003, Fuller et al. 2004). These fires caused loss of primary forest, large-scale habitat change and resource availability as well as loss of biodiversity (Singer et al. 1989, Kinnaird and O’Brien 1998, Barber and Schweithelm 2000). However, changes to wildlife communities, including patterns of recovery or extinction of specific species, resulting from fires remain poorly understood (Whelan 1995, Kinnaird and O’Brien 1998, O’Brien et al. 2003, Barlow and Peres 2006). The objective of this study was to explore the potential of using sign surveys as a monitoring tool to assess trends in the relative abundance of sun bears in a burned lowland dipterocarp forest versus an adjacent unburned area in East Kalimantan, over a period of 11 years.

Methods

Study area The study was carried out in lowland dipterocarp forest of the Sungai Wain Protection Forest (SWPF), a reserve near Balikpapan, East Kalimantan, Indonesian Borneo (1°16’ S and 116° 54’ E) at the eastern edge of the sun bear distribution range (Fig. 1). The reserve covers a watercatchment area of ca 100 km2. The topography of the reserve consists of gentle to sometimes steep hills, and is intersected by many small rivers. The area varies in altitude from 30 to 150 m above sea level. Trees with stems greater than 10 cm DBH (diameter at breast height) are dominated by the families Euphorbiaceae, Dipterocarpaceae, Sapotaceae, and Myrtaceae. The relative dominance of Dipterocarpaceae increased substantially in the larger size classes. Due to an altitudinal gradient, with rivers running in a north-south direction, the southern part of the reserve has more moist, swampy areas. Dipterocarpaceae (> 10 cm DBH), which are dominant in the northern part of the reserve decrease in abundance towards the moister south where Sapotaceae and Euphorbiaceae become more dominant. The 25 most common tree species form 40% of the total stem density (van Nieuwstadt 2002). Rainfall patterns in the region are not consistent on an annual basis, however a distinction could be made between a ‘wet’ season (Nov-April) and a ‘dry’ season (May-Oct) (t=4.96, df=6, p=0.002). Rainfall type falls within the category of ‘slightly seasonal’ according to Mohr’s Exploring the use of sign surveys for monitoring relative abundance of sun bears 135

Figure 1. Sun bear distribution range in Southeast Asia and location of the Sungai Wain Protection Forest, East Kalimantan, Indonesian Borneo.

index (Whitmore 1984), with average annual rainfall 2968 ± 510 mm y-1. Mean annual maximum and minimum temperature was 29.7 ± 0.7ºC and 23.2 ± 0.4ºC, respectively, measured in unburned forest on a daily basis during January 1998 – October 2002. Temperature in burned forest was not measured as regularly, but maximum temperatures (in the shade) of up to 34.6ºC were measured. The prolonged drought of the 1982-83 El Niño/Southern Oscillation (ENSO) had probably caused elevated levels of mortality among large trees, resulting in an irregular canopy cover. Consequently the primary forest became increasingly vulnerable to desiccation during droughts and subsequently more susceptible to ﬁres. Most of the reserve was unaffected 136 Chapter 7

by fires in 1997, when this study on sun bears commenced, except for a small area near the eastern border that had been burned in 1982-83. Fires entered the SWPF in March 1998, initially from a neighbouring state-owned logging concession, but subsequently also from surrounding agricultural fields. Fires moved slowly through the leaf litter and remained mainly in the undergrowth. In burned areas, all leaf litter and surface soil humus was reduced to ash, and mortality of seedlings and saplings was close to 100% (GM Fredriksson pers. obs.). Fire-breaks were created over a period of 2 months but nevertheless approximately 50% of the reserve was affected by the fires (Fredriksson 2002), leaving an unburned central core of some 4,000 ha of primary forest (Fig. 1) with unburned gallery forest along rivers in the burned-over forest areas covering between 7-18% of burned areas (van Nieuwstadt 2002, Fredriksson and Nijman 2004). Burned forest areas suffered elevated tree mortality up to several years after the fires with close to 80% mortality of sun bear fruit trees (>10 cm DBH) (Fredriksson et al. 2006b). Densities of various invertebrates (especially termites), which are food resources for sun bears, declined significantly after the fires in the burned areas (Fredriksson et al. unpublished data). In addition, burned forest areas became virtually inaccessible for almost 4 years after the fire event due to a thick groundcover of ferns (Dicranopteris linearis and Nephrolepis biserrata), which obstructed the understory, impeding movement for large ground-dwelling mammals.

Sign transects Bear signs were recorded along 10 x 1000m (1 ha) strip-transects. Transects were laid out along straight lines (primarily running across watersheds in a east-west direction, Fig. 2) and were surveyed by a team of three people, with one person moving slowly along the midline whilst recording data and distances with a Walktax Distance Measurer (Forestry Suppliers Inc., USA), and two people zigzagging on either side of the midline in search of sign throughout the 5-m strip of forest on either side of the midline. Transects were spread widely (at least 250 m apart) throughout the unburned and burned areas in the reserve (Fig. 2). Sign surveys were carried out over a period of >10 years starting in 2000, with follow-up surveys in 2001, 2002, 2005 and 2010. Sign surveys between 2000-2002 were carried out by GMF with 2 field assistants. Surveys in 2005 were carried out by two trained field assistants and a biology volunteer over a period of 3 months. In 2010, transects were again carried out by two trained field assistants and WJS over a period of approximately 5 months. A training session of 2 days was provided by GMF both in 2005 and 2010 in the forest, with several trial transects carried out with each Exploring the use of sign surveys for monitoring relative abundance of sun bears 137

Figure 2. Layout of transects in 2010 in the SWPF.

survey team in order to synchronize identiﬁcation and categorization of sun bear sign and ag- ing, and to reduce inter-observer differences. All sun bear sign (Table 1) encountered within the strip-transect was investigated in detail and recorded after consensus was achieved among the researchers. We recorded type of sign, estimated age of sign, and forest category in which sign was encountered (i.e. ridge, swamp, slope). All tree stems >5 cm DBH were investigated for claw marks, and the forest ﬂoor was searched for dug up or broken termite nests, soil excavations, or logs ripped by sun bears. On average 4.5 hours were used to record sign along one 1-km strip-transect. Within strip-transects, 138 Chapter 7

Table 1. Main categories of sun bear sign encountered in SWPF.

Sign category: Deposited during what activity: Broken stumps/rotten logs/toppled trees foraging for termites, ants/brood, beetles/beetle grubs, cockroaches Broken aboveground termite nests foraging for termite larvae/eggs/reproductives Dug up & destroyed underground termite nests foraging for termite larvae/eggs/reproductives Soil digs foraging for termite larvae/eggs/reproductives Claw mark, and associated sign climbing trees, primarily for obtaining fruit - Broken into stingless bee nests foraging for honey - Bite marks on trees marking behaviour (?) - Tree nests resting/feeding platforms Scat pile defecation Foot print walking live aboveground termite nests and figs (Ficus spp., both trees and hemi-epiphytes) were also recorded during surveys in 2005 and 2010, as termites and figs make up a large proportion of the sun bear diet (Wong et al. 2002, Fredriksson et al. 2006a). A small sample was inspected of each termite mound for signs of activity as well as species identification for those nest types that could contain different genera. Termite nests >5 m up tree trunks or in the canopy were not included. Figs were identified as such by means of checking latex with a small incision in the stem, as well as confirming the genus through its specific leaf venation.

Sign deposition by sun bears and recording of bear sign along strip-transects Identiﬁcation of sun bear sign was greatly assisted by behavioural observations of three orphaned sun bears that were rehabilitated between 1998-2002 and closely observed after their release in the forest. Continuous focal observations (Altman 1974) were made at least twice a week of each bear, during which we recorded each activity (i.e. breaking into termite mound, climbing tree, breaking into rotten log), duration of each activity, whether or not items were consumed (samples collected for species identiﬁcation), and if so, how many (for larger items like fruits or beetle grubs) and/or for how long (licking up termites or ant brood). We also occasionally made observations of other wild bears (including three that were radio-collared) while they foraged and produced sign. Sun bears in the study area, and during the study period, were found to predominantly feed on insect matter (88% of feeding time) Exploring the use of sign surveys for monitoring relative abundance of sun bears 139

supplemented by a large variety of fruit species (Fredriksson et al. 2006a). We classiﬁed sign into categories related to the type of activity of the bear and type of food sought (Table 1).

Decaying logs Sun bears searched for a variety of wood-boring and -living insects (beetle grubs, cockroaches, ants, termites) in dead wood. These ranged from small decaying broken branches (5 cm diameter) to massive rotten tree stumps or toppled trees (>200 cm) in varying degrees of decay. Along transects all decaying tree trunks with signs of sun bear damage were investigated in detail, but only recorded as sun bear sign when secondary proof was present at the site (e.g. hair, partial claw mark or dental impressions) or when all three researchers were in agreement that the sign could only have been produced by a sun bear. We were cautious in classifying ripped logs as sun bear sign because bearded pigs (Sus barbatus) also trampled logs in search of insects.

Termite nests Sun bears feed extensively on a variety of termite species. We categorized five main shapes of aboveground termite nests (Fig. 3a-e) plus one subterranean termite nest (Proham- itermes mirabilis; Fig. 3f) that sun bears were observed to consume. Sun bears located underground nests by smell, and dug up, and ripped apart the chambers to varying degrees. Nests of this species were occasionally clumped in distribution, with a number of nests connected underground within an area of several square meters. We observed one sun bear that excavated more than 40 nests of this species within 6 hours. Along transects, we recorded broken termite nests as sun bear sign when there was significant damage to nests, or when subterranean nests of P. mirabilis were excavated. Bear damage included large sections missing from termite nests glued on tree trunks between but- tresses (Hospitalitermes spp.), toppled and damaged termite nest pillars or drupes (Fig. 3b-c, Dicuspiditermes spp. or Termes spp.), or large sections missing from nests of Microcerotermes dubius usually constructed some 1-2 m off the ground onto tree trunks (Fig. 3d). We recorded the type of termite (assessed by opening a section of the remaining nest and identifying the species-specific soldiers), and estimated the time since it was broken into by a sun bear. Sympatrically occurring mammal species at the study site that are reported to feed on termites include pangolins (Manis javanica), orangutans (Pongo pygmaeus) and bearded pigs. Although we never directly observed any of these species feeding on termite nests, care was taken to look for characteristic signs of sun bear foraging whenever a broken or excavated termite nest was encountered. 140 Chapter 7

Figure 3a-e. Main termite mound shapes in SWPF. Constructed by species of the genera Dicuspiditermes and Termes (3a-c); Microcerotermes (2d); Bulbitermes (2e).

Figure 3f. Detail of an opened underground termite mound of the species Prohamitermes mirabilis and the typical posture of a sun bear breaking into a nest of this species.

Soil digs Sun bears also dug up soil in search of certain underground termites species that do not produce solid nest structures (primarily Macrotermes spp. and Odontotermes spp.). Species within the latter genus produce small (<10 cm), extremely brittle nests, but with extensive underground tunnels systems (Darlington 1997). Sun bears excavated nests in search of their winged reproductive forms (alates). Holes dug in the soil in search of these termites could be relatively deep (up to 30 cm) and frequently clustered, with several locations dug up in a small area. Along transects, soil digs were identiﬁed as produced by sun bears when associated sun bear sign was encountered (e.g. claw marks on large roots, or thick roots ripped apart during digging) or when excavation was of characteristic size and with backward piled excavated soil (similar to a dog digging in soil). Other mammals that produced soil excavations in this area Exploring the use of sign surveys for monitoring relative abundance of sun bears 141

include bearded pigs who till the soil shallowly with their noses, or potentially pangolins and porcupines (Hystrix brachyura) although their digging behavior was never directly observed.

Claw marks and other tree-related sign Sun bears climbed trees to feed on fruits or stingless bees nests, and occasionally to rest. Sun bear claw marks were easily distinguished from claw marks of other large tree-climbing species (e.g., monitor lizard Varanus salvator, clouded leopard Neofelis diardi), which leave finer, wavy marks. Along transects, all trees with sun bear claw marks were measured (diameter, height), identified to species whenever possible, and highest point of claw marks recorded. Each trunk with claw marks was investigated in detail to see whether multiple-age claw marks could be found, indicating separate climbing events. Ripped open tree trunks, or less frequent extensive diggings at the base of trees under the root system, to predate nests of stingless bees in search of honey and larvae, were one of the most characteristic sun bear signs. This sign was frequently associated with claw marks when recent. Large holes in trees, with sun bear bite marks, could remain identifiable as sun bear sign for many years, even when claw marks disappeared. Sun bears also left bite marks on trees, frequently with a section of the trunk debarked. These were usually associated with claw marks (the bear holding onto the tree with its front paws whilst biting) or associated dental indentations left in the bark.

Figure 3g. A sun bear ripping into a tree trunk to access a stingless bees nest and the aftermath of such a foraging session. 142 Chapter 7

Tree nests or feeding platforms were another characteristic sign of sun bears. These were constructed close to tree trunks, distinctive ‘V’ shaped, and usually associated with claw marks, quite easily differentiated from orangutan nests.

Other sign Sun bear scats were easily identiﬁed as such due to their size, neutral smell and contents. Footprints, though easily identiﬁed (as no other bear species occurs on Borneo), were rarely encountered due to the thick layer of leaf litter.

Ageing of sign and decay rates We assigned an age category to each sign encountered along transects: extremely fresh (<1 week); fresh (<1 month); relatively fresh (2-6 months); recent (6-12 months); intermediate (1-2 years); old (≥ 2 years). These categories were developed by examining decay rates of sign deposited during observations of habituated rehabilitated sun bears that were released in the forest. For subsequent analysis of sign we grouped all sign <1 year as ‘fresh’, sign aged at 1-2 years as ‘intermediate’, and sign >2 years as ‘old’. After watching a sun bear break into a termite nest, we monitored approximately 25 nests of each main nest type (Fig. 3a-e) and 45 excavated P. mirabilis nests (Fig. 3f) regularly for patterns of change. Decay rates of termite nests were affected by the level of destruction caused by sun bears, with nests that were heavily destroyed becoming difﬁcult to identify as sun bear sign within 12 months. Likewise, decay of soil digs was affected primarily by the size and depth of excavation and weather conditions. Most soil digs were not identiﬁable as bear excavations after 12 months. Claw marks and chewed tree trunks (from predated stingless bees nest) persisted from several months up to several years. Ageing of claw marks was monitored after observation of climbing events by human-habituated sun bears; trees were marked and checked on regular intervals. Ageing was similar to that described for decay trials carried out by Steinmetz and Garshelis (2010). Trees with multiple-age claw marks were treated as separate sign deposition events. Sun bear scats were primarily aged as very fresh or fresh, though most scats containing insect remains disintegrated in <36 hours. Sun bear scats containing seeds could be recognized as such for 6-43 days. Seedling clumps growing from fruit scats could be recognized up to 3 years after deposition. Footprints were recognizable for hours to 1 week, depending primarily on weather conditions or physical disturbance. Exploring the use of sign surveys for monitoring relative abundance of sun bears 143

Table 2. Decay rates of main categories of sun bear sign encountered in SWPF.

Sign category Subtype of sign Decay range Logs rotten logs 1 mo-1 yr ﬁrm dead wood 3 mo-2 yrs Termite nests dug up-but not broken 3 yrs partially destroyed 2 yrs entirely destroyed 1 yr Soil digs small 1 mo-1 yr large 3 mo-1 yr Claw marks - 6 mo-8+ yrs Stingless bee nests under tree bases 6 mo- 2 yrs cavity chewed in stem 3 yrs-13+ yrs Bite marks small 6 mo-8+ yrs large 2 yrs Scats termites 12-36 hours ﬁgs 6-17 days various hard seeds 6-43 days Seedling clump from scat - 3 yrs Foot prints 1 hour-1 week

Clusters of similarly aged sign Sun bear soil excavations, as well as excavations of P. mirabilis nests, estimated to be of the same age and clustered within 25 m of each other along a transects, were treated as a single feeding event based on our observations of foraging bears.

Ad-hoc searches of fresh sign after foraging activity by radio-collared sun bears We searched for fresh foraging sign of radio-collared wild sun bears on several occasions, when a bear was found (through telemetry readings) to have been active in a conﬁned area for an extended period of time.

Statistical analysis Differences in sign density between burned and unburned forest were ﬁrst compared using a two-way analysis of variance (ANOVA), with burned/unburned and years (n= 5) as 144 Chapter 7

independent variables. When significant differences were found, a series of one-way ANOVA’s in conjunction with Bonferroni post-hoc tests were used to evaluate sign types and year effects within the two sampled forest types (i.e., burned and unburned forest). Differences in the ages of sign between years in burned and unburned forest were evaluated using log likelihood ratio tests (i.e., G tests), and P values were estimated via Monte Carlo randomization (using SPSS’s default 10,000 permutations). In the case of a significant G test, implying that there was a significant amount of variation between variables, the adjusted standardized cell residuals were evaluated in order to identify which cells were most responsible for the observed variation. Following Agresti (2002), adjusted standardized cell residuals ≥ |2.0| were considered significant. Between year differences of fig and aboveground termite nest densities (2005 and 2010) were compared with paired sampled t-tests. Analyses were carried out with SPSS and R.

Results

We surveyed 92 strip transects, 45in unburned forest and 47 in burned areas, during 5 years over a period of a more than 10 years. Differences in sign densities were found between unburned and burned forest (two-way ANOVA, F=157.9, df=1,90, p<0.001) (Fig. 4), with a significant interaction effect in sign frequencies between years and treatment (burned/unburned) (F=5.1, df = 4, 91, P=0.001). Sign density in unburned forest remained stable across years, averaging 41.6 ± 11.6 (SD) sign/ha (Table 3, Fig. 4) (one-way ANOVA, F=0.28, df=4,40, p=0.891). We found some sun bear sign (range 9-71 sign/ha) in all transects in unburned forest. Lowest sign density encountered was on one transect that crossed an open swamp area, rarely used by sun bears. We cata- logued 1,871 sign events in 45 ha of unburned forest, the most common of which was broken termite nests (63.7%, n= 1,192), of which 24.3% were above ground and the remainder excavated underground nests (n=738). Claw marks (22.9%, n=429), and broken logs (7.9%, n=148) were next most common. Soil digs, broken logs and footprints all showed significant changes over the years (one-way ANOVA, soil: F=3.07, df=4,40, p=0.027; broken logs: F=3.12, df=4,40, p=0.025; prints: F=5.43, df=4,40, p=0.001) although only differences in soil digs between the years 2005-2010 remained significant (p=0.032) after post hoc Bonferroni tests. Sign densities in burned forest increased significantly over the years (one-way ANOVA, F=21.51, df=4,42, p<0.001), with 2005 and 2010 each having higher sign densities than all previous years (Bonferroni post hoc test, 2005: p<0.05 for all tests, 2010: p<0.01 for all tests). Exploring the use of sign surveys for monitoring relative abundance of sun bears 145

Figure 4. Average number of sun bear sign encountered along strip transects in burned and unburned forest 2000-2010.

Sign densities in burned forest areas were extremely low in 2000-2002 ranging from 0-4 sign/ha (Table 3). Sign density increased by 2005 (16.1 ± 6.3 (SD) sign/ha) but was still 60% lower than in adjacent unburned forest (40.8 ± 10.9 (SD) sign/ha) (t=7.10 df=24, p<0.001). By 2010 sign densities in burned forest areas had increased to 28.0 ± 12.9 (SD) sign/ha, though still 35% less than in adjacent unburned forest (44.1 ± 16.1 (SD) sign/ha) (t=3.04, df=28, p<0.005). Underground termite nest sign was signiﬁcantly higher in 2010 compared to all previous years (Bonferroni post hoc tests, p<0.001). No between-year differences were found in aboveground termite nest sign. By 2010, the most common sign categories encountered in burned forest (16 transects, n=439 sign) were broken termite nests (51.0%, n=224) consisting mostly of dug-up underground nests (40.5%, n=178), followed by claw marks (40.1%, n=176). Broken logs only made up 3.6% of sign (n=16). The three main categories of sign (broken termite nests, claw marks, broken/ripped logs) made up 90-97% of sign encountered in burned and unburned forest. In unburned forest, the percentage of sign related to termite diggings (62-71%), claw marks (23-25%) and dead wood (4-13%) remained stable over time (Fig. 5). In burned forest, by contrast, types of sign varied across years. The proportion of dead wood sign decreased over the years, whereas 146 Chapter 7

7.0 6.1 2.4 1.5 5.9 5.0 14.0 14.5 47 Total (0-27) (0-13) (0-22) (0-49) Burned forest

0 (0-1) (1-2) (1-18) (1-27) 0 0.5 0.5 4.6 8.1 0 0.3 1.3 7.6 11.1 0 (0-1) (0-1) (0-4) (0-13) 0 0.4 0.5 1.3 3.5 0 0.2 0.3 1.5 2.9 0 0 (0-1) (0-11) (2-22) 0 0 0.5 3.0 5.6 0 0 0.3 4.2 11.1 0.5 0.6 1.2 6.3 12.9 0.3 1.0 2.8 16.1 28.0 6 6 6 13 16 (0-1) (0-2) (1-4) (7-28) (5-49) 2000 2001 2002 2005 2010

5.0 9.5 6.8 7.2 10.1 16.4 11.6 41.6 45 Total (1-19) (0-36) (3-31) (9-71) Unburned forest

6 6 6 13 14 2000 2001 2002 2005 2010 SD 2.7 4.3 6.1 4.8 5.8 SD 5.8 4.2 3.1 4.1 10.0 SD 5.3 4.9 2.6 9.2 8.1 SD 12.5 5.5 3.4 10.9 16.1 mean 7.3 8.7 12.0 8.7 10.6 mean 9.0 6.7 8.3 9.5 13.4 mean 17.7 14.2 13.7 17.3 17.1 mean 39.3 39.2 42.0 40.8 44.1 range (3-10) (4-16) (5-22) (2-17) (1-19) range (5-20) (4-15) (3-12) (5-18) (0-36) range (12-26) (7-20) (10-18) (4-31) (3-31) range (29-63) (33-48) (38-46) (28-60) (9-71) Clawmarks Broken aboveground aboveground Broken termite mound Broken underground underground Broken termite mound Total sign (all categories combined) sign (all categories Total No. of 1 km belt transects ( 1ha) No. Table 3. Sun bear sign encountered along strip-transects in unburned and burned 2000-2010. forest Table Exploring the use of sign surveys for monitoring relative abundance of sun bears 147 0.1 0 0.8 0.4 1.0 0.8 1.2 0.7 (0-1) (0-3) (0-4) (0-4) Total

Burned forest 0 0 0 0 (0-1) 0 0 0 0 0.3 0 0 0 0 0.1 0 0 (0-1) (0-3) (0-3) 0 0 0.4 0.9 1.1 0 0 0.2 0.4 0.7 0.5 0.5 0.5 1.1 1.3 0.3 0.5 0.7 1.1 0.9 0 0 0 (0-4) (0-4) 0 0 0 1.4 1.2 0 0 0 1.2 1.2 (0-1) (0-1) (0-1) (0-3) (0-4) 2000 2001 2002 2005 2010 0.4 0.1 0.7 0.5 3.4 3.3 1.8 1.6 (0-2) (0-3) (0-6) Total (0-13)

Unburned forest 2000 2001 2002 2005 2010 SD 0 0.5 0.8 0 0 SD 0.5 0.4 1.2 0.5 0.8 SD 4.9 2.8 2.7 3.5 1.8 SD 2.1 2.3 1.1 1.9 0.9 mean 0 0.3 0.7 0 0 mean 0.3 0.8 0.7 0.3 0.6 mean 3.0 6.2 5.7 2.5 1.9 mean 2.0 2.3 1.0 2.5 0.6 range 0 (0-1) (0-2) 0 0 range (0-1) (0-1) (0-3) (0-1) (0-2) range (0-13) (1-9) (3-9) (0-13) (0-5) range (0-5) (0-5) (0-3) (0-6) (0-2) Print/Scat Broken bees nest Broken Broken log Broken Soil dig Table 3. Continued. Table 148 Chapter 7

Figure 5. Ratio of main sign categories in unburned and burned forest areas.

proportion of termite sign increased steadily. Termite sign (above- and underground nests combined) made up 35.9% of sign in burned forest in 2005 whereas in unburned forest this was 65.5% of total sign (Fig. 5). By 2010 termite sign in burned forest areas rose to 50.2% of total sign. Most of this rise was related to excavation of underground termites; density of that sign increased from 4.2 ± 3.0 nests/ha in 2005 to 11.1 ± 5.6 (SD) nests/ha in 2010 (compared to ~17 nests/ha in unburned forest; Table 4). Claw marks dominated the sign in burned areas in 2005, making up 47.4% of encountered sign compared to 21.3% in unburned forest, with equivalent densities between the two areas (Table 3). Trees with claw marks that belonged to taxa known to occur in the diet of sun bears (Fredriksson et al. 2006a) made up 77% of identiﬁed climbed trees in 2005 (both in burned and unburned forest), and 68-82% of trees in 2010 in burned and unburned forest, respectively. The proportions of different types of sun bear sign encountered along transects in unburned forest (all years combined) differed substantially from either the types of sign deposited during observations of bear foraging activity or the fresh sign encountered during searches of known bear foraging locations (Table 4). Direct observations of three rehabilitated sun bears between 1998-2002 (n=4,964 hours) provided a record of some 26,131 visually-conﬁrmed sign. More than half of this sign was ripped decaying logs, which was similar to what we found when searching areas where wild-caught radio-collared bears had previously fed for an extended Exploring the use of sign surveys for monitoring relative abundance of sun bears 149

Table 4. Percentage of sign allocated to different categories from 1) direct observations; 2) searches of sign left by radio-collared bears after prolonged activity in a restricted area; 3) sign encountered along transects in unburned forest.

Observations Sign searches of Transects (45 km) in (n=26,131 sign) known bear location unburned forest (n=147 sign) (n=1,871 sign) Ripped logs 54.4 (n=14,224) 55.8 (n=82) 7.9 (n=148) Broken termite nests 21.1 (n=5,515) 21.2 (n=31) 63.7 (n=1,192) Claw marks 10.3 (n=2,701) 8.8 (n=13) 22.9 (n=429) Soil digs 7.5 (n=1,964) 8.8 (n=13) 3.9 (n=73) Bite marks 3.1 (n=808) 0 0 Foot prints n/a 4.8 (n=7) 0 Scats 2.2 (n=584) 0.7 (n=1) 0.3 (n=6) Stingless bee nests 0.2 (n=52) 0 1.2 (n=23) Tree nests 0 0 0 Others 1.1 (n=283) 0 0 period (Table 4). In contrast, ripped logs represented only 8% of sign along transects in unburned forest. Whereas broken termite nests comprised >60% of the sign along transects, they represented only 21% of the sign observed associated with foraging by radio-collared bears

Densities of active termite nests and figs along transects Densities of active aboveground termite nests (all nest shapes combined) counted along strip transects were significantly lower in burned areas, both in 2005 and 2010 (2005, t=9.73, df=24, p<0.001; 2010, t=4.58, df=28, p<0.001) with densities reaching only 33% of those encountered in unburned forest those years. Densities of hemi-epiphytic figs did not differ significantly between burned and unburned areas, both in 2005 (t=-0.78, df=24, p=0.44) and 2010 (t=-0.32, df=28, p=0.75).

Changes in density of sign in different age categories over years The ratio of age categories, to which each sign was allocated, changed considerably in both burned (G statistic = 52.96, df=4, P<0.001) and unburned (G statistic = 90.45, df=4, P<0.001) areas over the sampling period. Density of sign allocated to the fresh (<1 year) category decreased in unburned forest during 2000-2010, from 27.6±11.4 (SD) sign/ha in 2000 150 Chapter 7

Figure 6. Densities of sign/hectare of different age categories in burned and unburned forest for all survey years. Exploring the use of sign surveys for monitoring relative abundance of sun bears 151

to16.9±8.5 (SD) in 2010 (Fig. 6). In burned forest the density of both fresh and old sign increased over the years, the latter from 2.6 ± 2.3 (SD) sign/ha in 2005 to 9.9 ±6.9 (SD) in 2010, whilst density of fresh sign only increased slightly from 10.2±5.1 (SD) sign/ha in 2005 to 11.1 ± 6.5 (SD) in 2010.

Discussion

Interpretation of sun bear sign surveys in the context of the SWPF Our comparison of sun bear sign densities between adjacent unburned and burned areas, through replicated sampling over time, showed that sun bears are slowly re-colonizing these once-burned areas, although they did not yet recover to pre-fire levels after more than a decade of forest regeneration. Sign densities increased from close to zero 2 years after the fire event, to 65% of that encountered in adjacent unburned forest habitat 12 years after the fire event. Sign densities in unburned forest remained relatively stable (44.1 ± 16.1 (SD) sign/ ha) over the decade of monitoring. The increase in sun bear sign observed in burned areas must be related to a slow recovery of food resources, as habitat use by bears has been found to be highly influenced by food availability (Steinmetz et al. 2011). The main increase of sign in burned areas was that of excavated underground termite nests (P. mirabilis) and claw marks on fruit-producing trees. Fruit resources play an important part of the sun bear diet, but availability varies drastically over time and is heavily influenced by aseasonal and rare mast fruiting events (Fredriksson et al. 2006a). In burned forest, close to 80% of sun bear fruit trees died, 3 years after the fire (Fredriksson et al. 2006b); however densities of claw marks in burned forest areas in 2010 were similar to those recorded in unburned forest (Table 4). This would indicate that sun bears were climbing remaining trees in these burned areas at a higher rate than in unburned forest. Close to 70% of trees with claw marks in burned areas in 2010 were taxa important in the diet of sun bears, and it might be that trees were more frequently climbed even if they had small fruit crops. Figs are one of the main fruit resources for sun bears during the prolonged inter- mast periods (Wong et al. 2002, Fredriksson et al. 2006a), and densities of hemi-epiphytic figs counted along transects, both in 2005 and 2010, were similar in burned and unburned forest. Putz and Susilo (1994) also reported quick recovery of root connections of hemi-epiphytic figs, 10 years after the extensive fire event of 1982-83 in East Kalimantan. 152 Chapter 7

Density of foraging sign of aboveground termite nests in burned forest was still very low 12 years after the fire event. Densities of active epigeal nests were still 60% lower than in adjacent unburned forest, indicating that recovery of aboveground termite species after fire disturbance is a slow process. Jones et al. (2003) reported the collapse of termite assemblages along a disturbance gradient in Sumatra related to changes in forest canopies, which normally buffer the forest interior from extreme variations in microclimate. Temperature, relative humidity, wind speed and solar radiation at ground level are all highly sensitive to changes in canopy depth and structure (Chen et al. 1999).In burned forest areas the buffering capacity of the canopy was much reduced as almost 80% of trees(>10 cm DBH) died after the fire event (Fre- driksson et al. 2006b). Gathorne-Hardy et al. (2002) hypothesized that termite recovery after heavy disturbance might take 60 years, if source populations of forest termites are found nearby. The reduction in density of fresh sign in unburned forest in 2010 (Fig. 6) could be in- terpreted in a number of ways. Sign transects only commenced after the fire event and hence any increase in sign densities from an influx of bears displaced from the burned area would not have been discernable. Bears fed mainly on insect resources in the remaining unburned area after the forest fires because fruits, even Ficus spp., were virtually absent. This fruiting low was possibly caused by disrupted phenologies after the prolonged El Niño drought (Harrison 2000, 2001), combined with exhausted energy reserves after the mast fruiting event (van Schaik 1986, Sork et al. 1993). Termites and ants, on the other hand, have been found consistently common on the forest floor in tropical forests (Kikkawa and Dwyer 1992), with highest termite densities found in lowland forests (Collins 1989). The diffuse but abundant distribution of this food resource, as opposed to highly sought-after, but low density fruit resources, might have enabled some bears from burned areas to squeeze into or between the homeranges of resident bears in unburned forest. The subsequent decrease in density of fresh sign might indicate that some bears moved back into regenerating burned areas, as substantiated by increasing sign densities there. Unexpectedly, data collected in 2010 in unburned forest showed an increase in intermediate and old sign, whereas in earlier years fresh and intermediate sign showed a decline. If fresh sign declined, old sign also should have declined, but with a time lag. The 5-year gap between 2005 and 2010 should have been a sufficient time lag to witness a decline in old sign (since most was <5 years old). That this was not the case might indicate that age estimation of sign during surveys in 2010 was inconsistent with previous surveys (i.e., some sign allocated to an older age category than it was). That is, inter-observer variability could have affected the results, despite our training course. Exploring the use of sign surveys for monitoring relative abundance of sun bears 153

Our results also suggest that temporal changes in regeneration of the forest and re- colonization by sun bears in SWPF are still in flux; this emphasizes the importance of long-term studies. Had this study only been carried out within the first 4 years post-fires, the conclusion would have been that there was virtually no use of the burned areas by sun bears. While true at the time, this would not have documented the potential value of these areas for re-colonization. Few studies so far have looked at the recovery of large mammals after forest fires in Indonesia (O’Brien et al. 2003) despite the immense scale of the impacted area: the fire event of 1997–1998 affected 5.2 million ha of land in the province of East Kalimantan alone, 2.6 million ha of which were forest including several protected lowland reserves (Hoffman et al. 1999, Siegert et al. 2001, Fuller et al. 2004).

Post-fire regeneration potential and conservation implications The re-colonization of sun bears into these once-burned forest areas implies that these forest areas can still be important for sun bear conservation, if there is undisturbed forest nearby that supports a source population. Sun bears are known to persist in logged-over areas (Wong et al. 2004, Linkie et al. 2007), but fire causes much more severe damage to forest integrity than selective logging (Slik et al. 2002). Most sun bear fruit resources are tree-borne (versus shrubs or understory plants) and regeneration of this food resource after a fire probably takes several decades, if not longer. Primary rainforest in Indonesia can rapidly be converted to savanna grasslands if affected by repeated fires (Goldammer 1999).Fire-return intervals of less than 90 years can eliminate rainforest species, whereas intervals of less than 20 years may eradicate trees entirely (Cochrane et al. 1999). Many of these fire-disturbed areas have seen further fire damage during smaller drought events in subsequent years (2002, 2004), or have been converted to other uses altogether. Dennis and Colfer (2006) calculated that 70% of forest initially damaged by fire and drought during the 1982–83 El Niño event was classified as non-forest by 2000, following a degradation pathway starting with commercial logging, followed by government sponsored transmigration schemes, then by government-licensed timber and oil palm plantations and, finally, the devastating fires of 1998. A similar, though faster, trajectory of conversion of degraded forests to plantations was evident for many fire-affected forest areas after 1997-1998 (Fuller et al. 2004, Miettinen et al. 2011). Thus remaining sun bear habitat has become increasingly fragmented, impairing postfire immigration of bears from other habitat blocks. Recovery of populations in forest areas 154 Chapter 7

disturbed by severe ecological catastrophes, such as the combined effects of fires, fragmentation and fruiting failure discussed in this paper, will now in many cases have to rely on recruitment from within remaining patches of less-disturbed habitat. Unfortunately, with the current minimal capacity to prevent or control forest fires in Indonesia, it seems unlikely in the near future that fire threat to forest areas will be reduced.

Further development of sign surveys as a monitoring tool Sun bears are currently listed as ‘Vulnerable’ by the IUCN (Fredriksson et al. 2008) primarily based on high rates of habitat loss over the last 30 years and increasing hunting pressures throughout mainland Southeast Asia (Foley et al. 2011). There is a pressing need to develop monitoring tools for this species, especially as poaching rates are on the rise and sun bears have already gone quietly extinct in several countries. Our current use of sign transects, as described here, to monitor changes in sun bear use of burned forest areas, with adjacent unburned forest as our control site, has proven to be a useful method. The implementation of these transects was relatively cheap and time efficient, although inter-observer variability in ageing sign might have been a problem. However, several challenges still remain in further developing this method as a tool to monitor trends in relative abundance of sun bears in key conservation areas. Steinmetz and Garshelis (2008) stressed that caution is required in inferring differences in population size over time or space from sign abundance as sign production can vary not only with changes in the number of bears but also their foraging behaviour. A number of issues related to the classification and interpretation of sign data need further attention. Different decay rates of different types of sign can cause potential problems. Certain sign categories can remain identifiable as sun bear sign for more than a decade (e.g., big holes ripped in tree trunks to predate stingless bees) whereas other types of sign disappear or become difficult to identify within months (e.g., broken logs or soil digs). This large variation in decay rates of different sign categories implies that it will be more prudent to focus trend data analysis on fresh sign only, where variance in decay is less pronounced. During our observations of rehabilitated sun bears we documented sign creation rates for different sign categories; these results were consistent with what we observed in areas where a wild sun bear had been foraging shortly before, but differed considerably from the ratio of sign categories encountered along transects (Table 4). The disparity, we believe, is primarily related to variation of decay rates of different sign types: sign types that decayed quickly were still evident a short time after Exploring the use of sign surveys for monitoring relative abundance of sun bears 155

a sun bear foraging bout, but had decomposed before our transect sampling. Thus, long-lasting sign (e.g., clawmarks and broken termite nests) were over-represented on transects. Other sign, like ripped-open logs, had decayed to the extent that they could no longer be positively identified as sun bear sign. One other important issue that will need further scrutiny is the definition of a single sign ‘event’. This is currently open to varied interpretation. Based on our feeding observations of bears in this study area, we considered similarly-aged dug-up underground termite nests, as well as soil digs, encountered within 25 m of each other as a single sign event, as bears would frequently create such sign during one feeding bout. In some cases this could mean that 10 dug up termite nests, all of similar age, would be recorded as 1 sign. We did not however group similarly-aged sign of different types, nor did we group two nearby trees with claw marks of similar age. To some extent, such decisions about grouping sign may affect results. Seasonality in the deposition of certain sign types related to seasonal differences in food availability might also be an issue. Sun bears foraged for termites year-round but certain mounds were more severely destroyed, and probably more sought after, when alates were ready to emerge. In the tropics, mass emergence of alates (swarming) is triggered by seasonal phenomena, initially thought to be rainfall (Nutting 1969), though species can swarm over prolonged periods of time (Neoh and Lee 2009a) with triggers for swarming linked to atmos- pheric pressure and temperature (Neoh and Lee 2009b). Sun bears in Borneo were found to feed heavily on fruits during the short and irregular mast fruiting season that occurs once every 2-10 years (Medway 1972, Ashton et al. 1988, Curran and Leighton 2000). During our study a significant mast fruiting event occurred toward the end of 1997-early 1998 and lasted for 2 months, 2 years prior to the start of sign transect work; no large fruiting event occurred during subsequent years but larger quantity of fruiting trees might lead to higher rates of claw marked trees. A final topic that would be most useful to further investigate is the feasibility of con- verting sign densities into estimates of bear abundance. During our observations of rehabilitated sun bears we documented sign creation rates and, on some occasions, measured real distances traveled by one bear (Fredriksson et al. in prep). These data, in combination with decay rates of different sign and sign detection along transects in the same forest area, might produce a workable relationship between sign densities and bear abundance. As sun bear behaviour is influenced by the distribution and abundance of different food resources, their sign deposition rates and the proportion of different sign types will differ amongst different ecosystems, making the conversion of sign densities to bear abundance a site-specific issue. 156 Chapter 7

Acknowledgements We thank the Indonesian Institute of Sciences (LIPI) for granting GMF permission to carry out research on sun bears in East Kalimantan. GMF would like to express gratitude to the Forest Research Institute in Samarinda for their assistance and the Wanariset Herbarium. Homathevi Rahman is thanked for identifying termite samples during a visit to Sungai Wain. Timbull is thanked for all illustrations. We would like to thank Claire Palmer for supervising the transect work in 2005. The Tropenbos Foundation, the Borneo Orangutan Foundation, and the Conservation Endowment Fund of the American Zoo and Aquarium Association provided ﬁnancial assistance. Field assistants Trisno, Lukas, and Damy are thanked for assistance in the forest throughout the years, as well as many people from the Sungai Wain village who helped with ferrying logistical supplies. WJS would like to thank Ristek for granting permission to carry our research in 2010, as well as ﬁeld assistants Ulir, Fitri, Misran and Anto. E. Rood and M. Nowak are thanked for assistance with statistics. Graham Usher is thanked for all help, as well as GIS analysis and maps for this paper.

References Agresti, A. (2002). Categorical data analysis. Wiley & Sons, New Jersey. Altmann, J. (1974). Observational study of behavior: sampling methods. Behaviour 49:227-265. Ashton, P.S., Givnish, T.J., and Appanah, S. (1988). Staggered Flowering in the Dipterocarpaceae - New In- sights into Floral Induction and the Evolution of Mast Fruiting in the Aseasonal Tropics. American Naturalist 132, 44-66. Barber, C.V., and Schweithelm, J. (2000). Trial by fire. Forest fires and forestry policy in Indonesia’s era of crisis and reform (Washington D.C, USA, World Resources Institute (WRI), Forest Frontiers Ini- tiative. In collaboration with WWF-Indonesia and Telapak Indonesia Foundation). Barlow, J., and Peres, C. (2006). Effects of single and recurrent wildfires on fruit production and large vertebrate abundance in a Central Amazonian forest. Biodiversity and Conservation 15, 985-1012. Bider, J.R. (1968). Animal activity in uncontrolled terrestrial communities as determined by a sand transect technique. Ecological Monographs 38, 269-308. Chen, J., Saunders, S.C., Crow, T.R., Naiman, R.J., Brosofske, K.D., Mroz, G.D., Brookshire, B.L., and Franklin, J.F. (1999). Microclimate in forest ecosystem and landscape ecology. Bioscience 49, 288–297. Cochrane, M. A., A. Alencar, M. D. Schulze, C. M. J. Souza, D. C. Nepstad, P. Lefebvre, and E. A. Davidson. (1999). Positive feedbacks in the fire dynamic of closed canopy tropical forests. Science 284:1832-1835. Exploring the use of sign surveys for monitoring relative abundance of sun bears 157

Collins, N.M. (1989). Termites. In Tropical Rain Forest Ecosystems Biogeographical and Ecological Studies, H. Lieth, and M.J.A. Werger, eds. (Amsterdam, Elsevier), pp. 455-471. Curran, L.M., and Leighton, M. (2000). Vertebrate responses to spatiotemporal variation in seed production of mast-fruiting dipterocarpaceae. Ecological Monographs 70, 101-128. Darlington, J.P.E.C. (1997). Comparison of nest structure and caste parameters of sympatric species of Odontotermes (Termitidae, Macrotermitinae) in Kenya. Insectes Sociaux 44, 393-408. Dennis, R.A., and Colfer, C.P. (2006). Impacts of land use and fire on the loss and degradation of lowland forest in 1983–2000 in East Kutai District, East Kalimantan, Indonesia. Singapore Journal of Tropical Geography 27, 30-48. Foley, K.E., Stengel, C.J., and Shepherd, C.R. (2011). Pills, Powders, Vials and Flakes: the bear bile trade in Asia (Petaling Jaya, Selangor, Malaysia, TRAFFIC Southeast Asia). Fredriksson, G.M. (2002). Extinguishing the 1998 forest fires and subsequent coal fires in the Sungai Wain Protection Forest, East Kalimantan, Indonesia. In: Moore, P., Ganz, D., Tan, L.C., Enters, T. and Durst, P.B. (eds.), Communities in flames: Proceedings of an international conference on community involvement in fire management. FAO and FireFight SE Asia, pp. 74-81. Fredriksson, G.M., and Nijman, V. (2004). Habitat-use and conservation of two elusive ground birds (Carpo- coccyx radiatus and Polyplectron schleiermacheri) in Sungai Wain protection forest, East Kalimantan, Indonesian Borneo. Oryx 38, 297-303. Fredriksson, G.M., Wich, S.A., and Trisno (2006a). Frugivory in sun bears (Helarctos malayanus) is linked to El Niño-related fluctuations in fruiting phenology, East Kalimantan, Indonesia. Biological Journal of the Linnean Society 89, 489-508.

Fredriksson, G.M., Danielsen, L.S., and Swenson, J.E. (2006b). Impacts of El Niño related drought and forest ﬁres on sun bear fruit resources in lowland dipterocarp forest of East Borneo. Biodiversity and Conservation 15, 1271-1301. Fredriksson, G., Steinmetz, R., Wong, S. & Garshelis, D.L. (2008). Helarctos malayanus. In: IUCN 2011. IUCN Red List of Threatened Species. Version 2011.2. . Downloaded on 27 April 2012. Fuller, D.O., Jessup, T.C., and Salim, A. (2004). Loss of forest cover in Kalimantan, Indonesia, since the 1997- 1998 El Niño. Conservation Biology 18, 249-254. Garshelis, D.L., and Hristienko, H. (2006). State and Provincial Estimates of American Black Bear Numbers versus Assessments of Population Trend. Ursus 17, 1-7. Gathorne-Hardy, F.J., Jones, D.T., and Syaukani (2002). A regional perspective on the effects of human disturbance on the termites of Sundaland. Biodiversity and Conservation 11, 1991-2006. 158 Chapter 7

Goldammer, J. G. (1999). Forest on ﬁre. Science 284:1782-1783. Hansen, M.C., Stehman, S.V., Potapov, P.V., Arunarwati, B., Stolle, F., and Pittman, K. (2009). Quantifying changes in the rates of forest clearing in Indonesia from 1990 to 2005 using remotely sensed data sets. Environmental Research Letters 4, 034001. Harrison, R. D. (2000). Repercussions of El Niño: drought causes extinction and the breakdown of mutualism in Borneo. Proceedings of the Royal Society of London Series B-Biological Sciences 267:911-915. Harrison, R. D. (2001). Drought and the consequences of El Niño in Borneo: a case study of ﬁgs. Population Ecology 43:63-75. Harrison, R.D. (2011). Emptying the Forest: Hunting and the Extirpation of Wildlife from Tropical Nature Reserves. BioScience 61, 919-924. Hoffmann, A.A., Hinrichs, A., and Siegert, F. (1999). Fire damage in East Kalimantan in 1997/98 related to land use and vegetation classes: Satellite radar inventory results and proposal for further actions. IFFM-SFMP Report No.1a (Samarinda, East Kalimantan, MOFEC, GTZ and KfW). Islam, M.A., Muzaffar, S.B., Aziz, M.A., Kabir, M.M., Uddin, M., Chakma, S., Chowdhury, S.U., Rashid, M.A., Chowdhury, G.W., Mohsanin, S., et al. (2010). Baseline survey of Bears in Bangladesh, 2008-2010 (Wildlife Trust of Bangladesh). IUCN/SSC Bear Specialist Group (2006), Workshop on Rangewide Mapping of Asian Bears, 17th Inter- national Conference on Bear Research and Management, International Bear Association (IBA), Karuizawa, Japan. Jones, D.T., Susilo, F.X., Bignell, D.E., Hardiwinoto, S., Gillison, A.N., and Eggleton, P. (2003). Termite assem-

blage collapse along a land-use intensiﬁcation gradient in lowland central Sumatra, Indonesia. Jour- nal of Applied Ecology 40, 380-391. Kendall, K.C., Metzgar, L.H., Patterson, D.A., and Steele, B.M. (1992). Power of sign surveys to monitor population trends. Ecological Applications 2, 422-430. Kikkawa, J., and Dwyer, P.D. (1992). Use of scattered resources in rain forest of humid tropical lowlands. Biotropica 24, 293-308. Kindberg, J., Swenson, J.E., Ericsson, G.r., Bellemain, E., Miquel, C., and Taberlet, P. (2011). Estimating population size and trends of the Swedish brown bear Ursus arctos population. Wildlife Biology 17, 114-123. Kinnaird, M.F., and O’Brien, T.G. (1998). Ecological effects of wildﬁre on lowland rainforest in Sumatra. Conservation Biology 12, 954-956. Kühl, H., F. Maisels, M. Ancrenaz, and E. A. Williamson. (2008). Best Practice Guidelines for Surveys and Monitoring of Great Ape Populations. IUCN, Gland, Switzerland. Exploring the use of sign surveys for monitoring relative abundance of sun bears 159

Kutilek, M.J., Hopkins, R.A., Clinite, E.W., and Smith, T.E. (1983). Monitoring population trends of large carnivores using track transects. In Renewable resource inventories for monitoring changes’ and trends, J. F. Bell, and T. Atterbury, eds. (Corvallis, Oregon State University), pp. 104-106. Linkie, M., Dinata, Y., Nugroho, A., and Haidir, I.A. (2007). Estimating occupancy of a data deficient mammalian species living in tropical rainforests: Sun bears in the Kerinci Seblat region, Sumatra. Biological Conservation 137, 20-27. Medway, L. (1972). Phenology of a tropical rain forest in Malaya. Biological Journal of the Linnean Society 4, 117-146. Meijaard, E., and Dennis, R.A. (1997). Forest fires in Indonesia; Bibliography and background information (Zeist, WWF Netherlands), pp. 38. Meijaard, E. (1999). Human-imposed threats to sun bears in Borneo. Ursus 11, 185-192. Miettinen, J., Shi, C., and Liew, S.C. (2011). Deforestation rates in insular Southeast Asia between 2000 and 2010. Global Change Biology 17, 2261-2270. Mowat, G., and Strobeck, C. (2000). Estimating population size of grizzly bears using hair capture, DNA profiling, and mark-recapture analysis. Journal of Wildlife Management 64, 183-193. Neoh, K.-B., and Lee, C.-Y. (2009). Flight Activity of Two Sympatric Termite Species, Macrotermes gilvus and Macrotermes carbonarius (Termitidae: Macrotermitinae). Environmental Entomology 38, 1697-1706. Neoh, K.-B., and Lee, C.-Y. (2009). Flight activity and flight phenology of the Asian subterranean termite, Coptotermes gestroi (Isoptera: Rhinotermitidae). Sociobiology 54, 521-530. Nutting, W.L. (1969). Flight and colony foundation. In Biology of Termites, K. Krishna, and F.M. Weesner, eds. (New York, Academic Press), pp. 233-282.

O’Brien, T.G., Kinnaird, M.F., Nurcahyo, A., Prasetyaningrum, M., and Iqbal, M. (2003). Fire, demography and the persistence of siamang (Symphalangus syndactylus: Hylobatidae) in a Sumatran rainforest. Ani- mal Conservation 6, 115-121. Putz, F.E., and Susilo, A. (1994). Figs and fire. Biotropica 26, 468-469. Siegert, F., Ruecker, G., Hinrichs, A., and Hoffmann, A.A. (2001). Increased damage from fires in logged forests during droughts caused by El Niño. Nature 414, 437-440. Singer, F. J., W. Schreier, J. Oppenheim, and E. O. Garton. (1989). Drought, Fires, and Large Mammals. BioSci- ence 39:716-722. Slik, J.W.F., Verburg, R.W., and Kessler, P.J.A. (2002). Effects of fire and selective logging on the tree species composition of lowland dipterocarp forest in East Kalimantan, Indonesia. Biodiversity & Conser- vation 11, 85-98. Sork, V. L. (1993). Evolutionary ecology of mast-seeding in temperate and tropical oaks (Quercus spp.). Vegetatio 107/108:133-147. 160 Chapter 7

Stanley, T.R.J., and Bart, J. (1991). Effects of roadside habitat and fox density on a snow track survey for foxes in Ohio. Ohio Journal of Science 91, 186-190. Steinmetz, R., Chutipong, W., and Seuaturien, N. (2006). Collaborating to Conserve Large Mammals in Southeast Asia. Conservation Biology 20, 1391-1401. Steinmetz, R., and Garshelis, D.L. (2008). Distinguishing Asiatic Black Bears and Sun Bears by Claw Marks on Climbed Trees. Journal of Wildlife Management 72, 814-821. Steinmetz, R., and Garshelis, D.L. (2010). Estimating ages of bear claw marks in Southeast Asian tropical forests as an aid to population monitoring. Ursus 21, 143-153. Steinmetz, R., Garshelis, D.L., Chutipong, W., and Seuaturien, N. (2011). The Shared Preference Niche of Sympatric Asiatic Black Bears and Sun Bears in a Tropical Forest Mosaic. PLoS One 6, e14509. Tacconi, L. (2003). Fires in Indonesia: Causes, costs and policy implications (Bogor, CIFOR), pp. 1-34. van Nieuwstadt, M.G.L. (2002). Trial by fire- Postfire development of a dipterocarp forest. Ph.D. thesis. University of Utrecht, Netherlands. van Schaik, C. P. (1986). Phenological changes in a Sumatran rain forest. Journal of Tropical Ecology 2:327-347. Vié, J-C., Hilton-Taylor, C. and Stuart, S.N. (2009). Wildlife in a changing world - An analysis of the 2008 IUCN red list of threatened species. IUCN, Gland, Switzerland, 180 pp. Whelan, R.J. (1995). The ecology of fire (Cambridge, Cambridge University Press). Whitmore, T.C. (1984). Tropical rain forests of the Far East, 2nd edn. (Oxford, Clarendon Press). Wilson, G. J. and Delahay, R. J. (2001). A review of methods to estimate the abundance of terrestrial carnivores using field signs and observation. Wildl. Res. 28: 151–164. Wong, S.T., Servheen, C., and Ambu, L. (2002). Food habits of Malayan sun bears in lowland tropical forest

of Borneo. Ursus 13, 127-136. Wong, S.T., Servheen, C., and Ambu, L. (2004). Home range, movement and activity patterns, and bedding sites of Malayan sun bears, Helarctos malayanus in the rainforest of Borneo. Biological Conserva- tion 119, 168-181. Chapter 8

Synthesis and Conclusions Synthesis and conclusions 165

Synthesis and Conclusions

The Malayan sun bear, Helarctos malayanus, is the smallest of the eight extant bear species, and is distributed throughout Southeast Asia including the large islands of Sumatra and Borneo. Research on sun bears in the wild, including those in Kalimantan (Indonesian Borneo), was non-existent before the work that is presented in this thesis commenced in 1997, in a small lowland reserve in East Kalimantan. Apart from a few anecdotal records, virtually nothing was known about basic aspects of sun bear ecology and they were listed as Data Deﬁcient by the IUCN Red list at the time. The results of this study have ﬁlled some of the bigger gaps in our knowledge on various aspects of sun bear behaviour, ecology, and conservation.

Setting the stage: El Niño-Southern Oscillation and forest fires The start of this research in the Sungai Wain Protection Forest (SWPF) coincided with a strong El Niño-Southern Oscillation (ENSO) episode. Global climatic cycles, such as ENSO, affect diverse ecological processes including community dynamics and landscape disturbances in tropical regions (Webster and Palmer 1997). The onset of ENSO events in Asia is marked by a reduction in rainfall, resulting in drought conditions (Walsh 1996, Walsh and Newbery 1999). Furthermore, climatic conditions associated with ENSO provide an irregular supra-annual, regional trigger which initiates asynchronous, widespread flowering in Bornean Dipterocar- paceae (Curran et al. 1999). Interestingly, a variety of other taxonomic groups flower and fruit synchronously with these dipterocarps (Appanah 1985, Sakai et al. 1999, Cannon et al. 2007). One of these rare community mast fruiting events [supra-annual production of large seed crops interspersed by irregular periods of low seed production] triggered by the 1997-98 ENSO took place at our study site towards the end of 1997 (Chapter 2). This same ENSO event also provided the enabling conditions, through prolonged drought, for the catastrophic forest fires of 1997-98. This fire event affected circa 5 million hectares (Mha) of land in the province of East Kalimantan (Indonesian Borneo) alone, including 2.6Mha of forest (Hoffmann et al. 1999, Siegert et al. 2001). In March 1998, fires burned half of the forest in the reserve where this research project had started. Between 1997-2006, a total of 16.2 Mha of Borneo’s landmass (21% of total land surface area) was affected by fires (Langner and Siegert 2009). Habitat degradation due to large-scale forest fires, combined with a prolonged fruiting low after the ENSO event, put large strains on the sun bear population in the area. 166 Chapter 8

Sun bear responses to large-scale fluctuations in food availability Sun bears, like most bear species, are omnivores. During rare mast-fruiting events in East Kalimantan, sun bears opportunistically switched their diet to become frugivores (Chap- ter 2). During these short periods, when calorie-rich fruits are available, sun bears gorged themselves on large quantities of large-seeded fruits, for which they are effective seed dispersers (McConkey and Galetti 1999, Chapter 2). In SWPF this brief glut only lasted two months, followed by almost four years with very little fruit available (Chapter 2). This fruiting low was probably caused by exhausted energy reserves after the mast fruiting event (van Schaik 1986, Sork et al. 1993), but might have been aggravated by disrupted phenologies after the prolonged El Niño drought (Harrison 2000, 2001). The extent of frugivory in sun bears will vary considerably throughout their distribution range as patterns of fruit production differ substantially among tropical forests (van Schaik et al. 1993, Wich and van Schaik 2000, Clark et al. 2001, Cannon et al. 2007). Bornean forests, despite their food-rich appearance, display a significantly lower productivity than Sumatran forests (Wich et al. 2011). Fruit productivity is generally thought to set carrying capacity for rainforest vertebrates, which in turn may affect plant life history, seed dispersal, and vertebrate population dynamics (Janzen 1974, Curran and Leighton 2000, Wich et al. 2004). During the prolonged inter-mast period, sun bears shifted their diet to insects, primarily termites and ants, complemented by a large variety of other saproxylic insects like beetle grubs, cockroaches, and adult beetles. Although food resources in lowland tropical forests in general are considered to be widely scattered and scarce, termites and ants are found to be relatively abundant (Collins 1980, Kikkawa and Dwyer 1992). However, a large amount of energy is expended in obtaining these small quantities of food. A small number of plant taxa were still producing fruit during the inter-mast period, some of which were highly sought after by sun bears (Chapter 2). Sun bear ranging patterns during these inter-mast periods were influenced by the number of trees in fruit, with bears ranging over larger areas in search for these widely dispersed but highly sought after fruit resources (Chapter 3). Female bears captured in the unburned forest core, 16 months after the fire event and well into the inter-mast period, showed signs of serious undernourishment and weighed 30- 40% less than female bears caught simultaneously 700 km to the north in the Tabin Reserve, Sa- bah, Malaysian Borneo. Bears at Tabin were found to be foraging on oil palm fruits in plantations adjacent to the reserve (Nomura et al. 2004) during this period of fruit scarcity. At my study Synthesis and conclusions 167

site, few external food resources were available for bears, except for bears living near the forest edge, some of which were augmenting their diet with fruits grown in orchards neighbouring the forest edge (Chapter 5). The emaciated appearance of the female sun bears caught in SWPF, without access to additional food supplies, suggests that prolonged survival solely on insect resources was insufficient and that energy requirements were not met. Increased competition due to immigration of displaced bears from the fires might have been an additional strain factor. The predation of one of the bears by a large python might have been related to the poor nutritional condition of the bear, which might have rendered the bear less effective in fighting off an attack or having a lowered vigilance (Chapter 6). Simultaneous to the period of fruit scarcity in SWPF, a prolonged fruiting low was also reported from the Danum Valley Conservation Area in Sabah, Malaysian Borneo: emaciated bears and bearded pigs (Sus barbatus) were encountered at the same time in primary forest there (Wong et al. 2005). In Danum, bears also had no access to human-related food resources like oil palm fruits, unlike bears living near the forest edges in Tabin Reserve (Nomura et al. 2004). Curran and Leighton (2000) reported on emaciated and dead pigs after a prolonged fruiting low during the same inter-mast period in West Kalimantan, further indicating that this famine was possibly an island-wide occurrence. They speculated that the length of the inter-mast cycle was long enough to suppress populations of bearded pigs, with piglets only seen during mast fruiting events. Survival during periods of fruit scarcity is viewed as the critical determinant of frugivore vertebrate densities in Southeast Asian forests (van Schaik et al. 1993). Famine following fruiting failures has also been reported from other parts of the world. Following a year of extremely low fruit production, mass mortality of Japanese macaques was reported by Hanya et al. (2004), when 56% of individual primates of several well-studied groups died. They hypothesized that the shortage of high-quality foods caused poor nutritional conditions for the primates, making them highly vulnerable to diseases, with death as the ultimate result. Wright et al. (1999) reported that on Barro Colorado Island, Panama, over a period of 49-years, four episodes of famine occurred every time an El Niño event was followed by a mild dry season with very low fruit production. Reproductive behavior of Bornean sun bears has been found to be aseasonal and poly- estrous (Schwarzenberger et al. 2004, Frederick et al. 2012). These breeding habits of sun bears might be an adaptation to the unpredictable availability of fruit, both in space and time, allowing female sun bears to come back into estrous soon after having lost a cub. In SWPF female sun bears were observed throughout the period of extreme fruit low as well as during mast fruiting events with small cubs. Two undernourished female sun bears caught during the peak of fruit 168 Chapter 8

scarcity in SWPF for telemetry studies, both had small cubs. However, both female bears died subsequently leaving the fate of the cubs unknown. This indicates that these bears were still able to find sufficient food resources to give birth, but it seems that the extreme food scarcity following the ENSO event and fires, and possibly competition for food, meant that there was insufficient food to sustain both lactation and maintenance of themselves. For American black bears, which are seasonal breeders, a clear link between mast fruiting failure and a subsequent decline in reproduction has been shown in several regions (Jonkel and Cowen 1970, Elowe and Dodge 1989, Costello et al. 2003).

Impacts of fire on sun bear populations The fires in 1998 burned 40% of SWPF in six weeks, depriving the sun bears living there of their habitat. Forest fires had a destructive effect on the main sun bear food resources (tree- borne fruits and social insects), with a decrease of > 80% in the density of sun bear fruit trees (Chapter 4), as well as a significant reduction of densities of above- and underground termite nests. Besides a decrease in food resources, reduced canopy cover due to high post-fire tree mortality caused large changes in the micro-climate of the burned forest areas. Temperatures during daytime were several degrees higher than in unburned forest, and a smothering thicket of ferns dominated the understory in once burned areas (Slik and Eichhorn 2003), creating a physical barrier for large ground-dwelling mammals. While it is possible that some bears could have become trapped and burned, there are a number of possibilities as to what happened to bears displaced from burned forest areas. During the period of active fires, a high rate of sun bear roaring calls were heard near the fire front indicating an increase in antagonistic encounters between bears, suggesting that bears were displaced by fires and moved into home ranges of bears living in unburned forest. It is possible that some bears subsequently were able to fit into the remaining unburned forest, thus increasing densities of bears in the central core of the reserve. Sign transects, developed in order to monitor relative abundance of sun bears (Chapter 7), only commenced after the fire event and hence any increase in sign densities that would have occurred could not be detected. The diffuse but abundant distribution of termites and ants, the main inter-mast food resource, as opposed to highly sought after but low density fruit resources, might have allowed bears from burned areas to fit into the homeranges of resident bears in unburned forest. Crop-raiding activity increased in gardens adjacent to the forest edge after the forest fires (Chapter 5), suggesting that bears were being displaced either from the core remaining forest or from burned areas. In response to the crop-raiding certain farmers laid out poisoned Synthesis and conclusions 169

baits and snares, and it is possible that bears were killed due to this increased encounter rate with humans. Another possibility is that certain bears survived in unburned forest islands within the burned forest matrix. Between 7-18% of the burned forest remained unscathed by the ﬁres as unburned patches, primarily near rivers or swamps (van Nieuwstadt 2002, Fredriksson and Nijman2004). The use of such unburned patches and avoidance of burned areas has also been reported for black bears in Arizona after ﬁres (Cunningham et al. 2003).

Ranging and Activity Annual ranging patterns for female sun bears measured during this study were 4-5 km2 (Chapter 3), with ranging patterns of collared bears entirely encompassed within the reserve boundaries. Home range sizes of bears are related to body size (McNab 1963, Gittleman and Harvey 1982, Gompper and Gittleman 1991), metabolic needs (McNab 1983) and distribution and abundance of food (McLoughlin et al. 1999, 2000, Gompper and Gittleman 1991), resulting in large inter- and intraspeciﬁc variation. Insectivorous carnivores usually have relatively small home ranges (Gittleman and Harvey 1982) seemingly because protein-rich insect prey, especially ants and termites, are abundant and omnipresent over relatively small areas (e.g. Wood and Sands 1978, Redford 1987). As such, the sloth bear (Melursus ursinus), the most insectivorous ursid, appears to have the smallest home ranges of any species of bear (Ratnayake et al. 2007). During our study, sun bears had a comparable diet to sloth bears, spending up to 90% of their foraging time consuming insects. It is possible that ranging patterns were slightly enlarged during the inter-mast period due to the bears searching over a wider area for the few available, scattered fruit trees, as opposed to smaller ranging patterns which would have been expected on a predominantly insect diet (Gittleman and Harvey 1982). Sun bears were found to display diurnal activity patterns within the SWPF (Chapter 3), although more nocturnal behavior was reported for bears living near the forest edges (Chap- ter 5). Consistent with their temporal response to humans, sun bears living in the core of the reserve also showed spatial avoidance of human activity. Bears living near forest edges are attracted to crops, especially when natural food supplies are low, making border areas potential demographic sinks (Woodroffe and Ginsberg 1998).

Re-colonization of burned forest areas Our comparison of sun bear sign densities between adjacent unburned and burned areas showed that sun bears are slowly re-colonizing these once-burned areas, although they did not yet recover to pre-ﬁre levels after more than a decade of forest regeneration. Sign 170 Chapter 8

densities increased from close to zero 2 years after the fire event, to 65% of that encountered in adjacent unburned forest habitat 12 years after the fire event. Sign densities in unburned forest remained relatively stable (44.1 ± 16.1 (SD) sign/ha) over the decade of monitoring (Chapter 7). In burned forest areas, claw marks and dug up underground termite nests dominated sign density, but density of foraging sign of aboveground termite nests was still very low 12 years after the fire event. Densities of active epigeal nests were still 60% lower than in adjacent unburned forest, indicating that recovery of aboveground termite species after fire disturbance is a slow process. Jones et al. (2003) reported the collapse of termite assemblages along a disturbance gradient in Sumatra related to changes in forest canopies, which normally buffer the forest interior from extreme variations in microclimate. Temperature, relative humidity, wind speed and solar radiation at ground level are all highly sensitive to changes in canopy depth and structure (Chen et al. 1999). In burned forest areas the buffering capacity of the canopy was much reduced as almost 80% of trees (>10 cm DBH) died after the fire event (van Nieuwstadt 2002, Chapter 4). Gathorne-Hardy et al. (2002) hypothesized that termite recovery after heavy disturbance might take 60 years, if source populations of forest termites are found nearby. Temporal changes in regeneration of the forest and re-colonization by sun bears in SWPF are still in flux, which emphasizes the importance of long-term studies. Had this study been carried out only within the first 4 years post-fires, the conclusion would have been that there was virtually no use of the burned areas by sun bears. While true at the time, this would not have documented the potential value of these areas for re-colonization.

Conservation implications Few studies so far have looked at the recovery of large mammals after forest fires in Indonesia (O’Brien et al. 2003) despite the immense scale of the impacted area. The re- colonization of sun bears into these once-burned forest areas implies that these can still be important for sun bear conservation, if there is undisturbed forest nearby that supports a source population. Sun bears are known to persist in logged-over areas (Wong et al. 2004, Linkie et al. 2007), but fire causes much more severe damage to forest integrity than selective logging (Slik et al. 2002). Most sun bear fruit resources are tree-borne (versus shrubs or understory plants) and regeneration of this food resource after a fire probably takes several decades, if not longer. Primary rainforest in Indonesia can rapidly be converted to savanna grasslands if affected by repeated fires (Goldammer 1999). Fire-return intervals of less than 90 years can eliminate rainforest species, whereas intervals of less than 20 years may eradicate trees entirely Synthesis and conclusions 171

(Cochrane et al. 1999). Many of these fire-disturbed areas have seen further fire damage during smaller drought events in subsequent years (2002, 2004), or have been converted to other uses altogether. Dennis and Colfer (2006) calculated that 70% of forest initially damaged by fire and drought during the 1982–83 El Niño event was classified as non-forest by 2000, following a degradation pathway driven by successive government policies starting with commercial logging, followed by government sponsored transmigration schemes, then by timber and oil palm plantations and, finally, the devastating fires of 1998. A similar, though faster, trajectory of conversion of degraded forests to plantations or wasteland was evident for many fire-affected forest areas after 1997-1998 (Fuller et al. 2004, Miettinen et al. 2011). Thus remaining sun bear habitat has become increasingly fragmented, impairing postfire immigration of bears from other habitat blocks. Recovery of populations in forest areas disturbed by severe ecological catastrophes, such as the combined effects of fires, fragmentation and fruiting failure discussed in this thesis, will now in many cases have to rely on recruitment from within remaining patches of less-disturbed habitat. Unfortunately, with the current minimal capacity to prevent or control forest fires in Indonesia, it seems unlikely in the near future that fire threat to forest areas will be reduced.

Importance of monitoring population trends in the face of increasing conservation threats Sun bears are now listed as Vulnerable on the IUCN Red List (Fredriksson et al. 2008). Poaching for bears and their body parts is reaching alarming rates in mainland Southeast Asia (Foley et al. 2011). Although current poaching rates are lower in Sumatra and Borneo, where the main threat to sun bear populations comes from habitat loss, it is probable that the continuing economic growth in the primary markets for bear products, namely China, will lead to increased poaching in Indonesia, particularly as mainland populations become increasingly de- pleted. In this scenario, the track record of the conservation institutions in Indonesia responsible for species protection and control of wildlife trade does not provide much reassurance that they would be able to mitigate this threat. Monitoring of sun bear population trends in key conservation areas will become the next imperative. Results from our research show that changes in sun bear relative abundance can be monitored using relatively simple and inexpensive sign surveys. At the moment, six medium sized to large (1,000-10,000 km2) protected areas have been designated in Kalimantan, still consisting of good quality habitat. Kalimantan still holds the largest contiguous expanse of forest in the Indo-Malayan realm, primarily in the mountainous interior of the island. These 172 Chapter 8

remote, inaccessible, large protected areas provide the most secure future prospect for sun bears and other wildlife. Indonesia, with its remaining forests on Sumatra and Kalimantan, currently covers some 45% of the sun bears distribution range, making it the most important country for the conservation of this bear species.

References Appanah, S. (1985). General flowering in the climax rain forests of South-east Asia. Journal of Tropical Ecol- ogy 1:225-240. Cannon, C. H., L. M. Curran, A. J. Marshall, and M. Leighton. (2007). Long-term reproductive behaviour of woody plants across seven Bornean forest types in the Gunung Palung National Park (Indonesia): suprannual synchrony, temporal productivity and fruiting diversity. Pages 956-969. Wiley-Blackwell. Chen, J., Saunders, S.C., Crow, T.R., Naiman, R.J., Brosofske, K.D., Mroz, G.D., Brookshire, B.L., and Franklin, J.F. (1999). Microclimate in forest ecosystem and landscape ecology. Bioscience 49, 288–297. Clark D.A., Brown S., Kicklighter D., Thommlinon J.R., and N. J. (2001). Net primary production in tropical forests: an evaluation and synthesis of existing field data. Ecological Applications 11:371–384. Cochrane, M. A., A. Alencar, M. D. Schulze, C. M. J. Souza, D. C. Nepstad, P. Lefebvre, and E. A. Davidson. (1999). Positive feedbacks in the fire dynamic of closed canopy tropical forests. Science 284:1832-1835. Collins, N. M. (1980). The effect of logging on termite (Isoptera) diversity and decomposition processes in lowland dipterocarp forest. Tropical Ecology and Development:113-121. Costello, C., D. E. Jones, R. M. Inman, K. H. Inman, B. Thompson, and H. B. Quigley. (2003). Relationship of variable mast production to American black bear reproductive parameters in New Mexico. Ursus14:1-16. Cunningham, S. C., W. B. Ballard, L. M. Monroe, M. J. Rabe, and K. D. Bristow. (2003). Black bear habitat use in burned and unburned areas, central Arizona. Wildlife Society Bulletin 31:786-792. Curran, L.M., and Leighton, M. (2000). Vertebrate responses to spatiotemporal variation in seed production of mast-fruiting dipterocarpaceae. Ecological Monographs 70, 101-128. Curran, L. M., I. Caniago, G. D. Paoli, D. Astianti, M. Kusenti, C. E. Nirarita, and H. Haeruman. (1999). Impact of El Nino and logging on canopy tree recruitment in Borneo. Science 286:2184-2188. Dennis, R.A., and Colfer, C.P. (2006). Impacts of land use and fire on the loss and degradation of lowland forest in 1983–2000 in East Kutai District, East Kalimantan, Indonesia. Singapore Journal of Tropical Geography 27, 30-48. Elowe, K. D. and W. E. Dodge. (1989). Factors affecting black bear reproductive success and cub survival. Journal of Wildlife Management 53:962-968. Synthesis and conclusions 173

Foley, K.E., Stengel, C.J., and Shepherd, C.R. (2011). Pills, Powders, Vials and Flakes: the bear bile trade in Asia (Petaling Jaya, Selangor, Malaysia, TRAFFIC Southeast Asia). Frederick, C., R. Kyes, K. Hunt, D. Collins, and S. K. Wasser. (2012). Reproductive timing and aseasonality in the sun bear (Helarctos malayanus). Journal of Mammalogy 93. Fredriksson, G.M., and Nijman, V. (2004). Habitat-use and conservation of two elusive ground birds (Carpo- coccyx radiatus and Polyplectron schleiermacheri) in Sungai Wain protection forest, East Kalimantan, Indonesian Borneo. Oryx 38, 297-303. Fuller, D.O., Jessup, T.C., and Salim, A. (2004). Loss of forest cover in Kalimantan, Indonesia, since the 1997- 1998 El Niño. Conservation Biology 18, 249-254. Gathorne-Hardy, F.J., Jones, D.T., and Syaukani (2002). A regional perspective on the effects of human disturbance on the termites of Sundaland. Biodiversity and Conservation 11, 1991-2006. Gittleman, J.L., and Harvey, P.H. (1982). Carnivore home-range size, metabolic needs and ecology. Behavio- ral Ecology and Sociobiology 10, 57-63. Goldammer, J. G. (1999). Forest on ﬁre.Science 284:1782-1783. Gompper, M.E., and Gittleman, J.L. (1991). Home range scaling: intraspeciﬁc and comparative trends. Oe- cologia 87, 343-348. Hanya, G., M. Matsubara, H. Sugihara, S. Hayakawa, S. Goto, T. Tanaka, J. Soltis, and N. Noma. (2004). Mass mortality of Japanese macaques in a western coastal forest of Yakushima. Ecological Research 19:179-188. Harrison, R. D. (2000). Repercussions of El Niño: drought causes extinction and the breakdown of mutualism in Borneo. Proceedings of the Royal Society of London Series B-Biological Sciences

267:911-915. Harrison, R. D. (2001). Drought and the consequences of El Niño in Borneo: a case study of ﬁgs. Population Ecology 43:63-75. Hoffmann, A.A., Hinrichs, A., and Siegert, F. (1999). Fire damage in East Kalimantan in 1997/98 related to land use and vegetation classes: Satellite radar inventory results and proposal for further actions. IFFM-SFMP Report No.1a (Samarinda, East Kalimantan, MOFEC, GTZ and KfW). Janzen, D. H. (1974). Tropical blackwater rivers, animals, and mast fruiting by the dipterocarpaceae. Bio- tropica 6:69-103. Jones, D.T., Susilo, F.X., Bignell, D.E., Hardiwinoto, S., Gillison, A.N., and Eggleton, P. (2003). Termite assemblage collapse along a land-use intensiﬁcation gradient in lowland central Sumatra, Indonesia. Jour- nal of Applied Ecology 40, 380-391. Jonkel, C. J. and I. Mc.Cowan. (1970). The Black Bear in the Spruce-Fir forest. Wildlife Monographs 27:1-57. 174 Chapter 8

Kikkawa, J., and Dwyer, P.D. (1992). Use of scattered resources in rain forest of humid tropical lowlands. Biotropica 24, 293-308. Langner, A. and F. Siegert. (2009). Spatiotemporal fire occurrence in Borneo over a period of 10 years. Global Change Biology 15:48-62. Linkie, M., Dinata, Y., Nugroho, A., and Haidir, I.A. (2007). Estimating occupancy of a data deficient mammalian species living in tropical rainforests: Sun bears in the Kerinci Seblat region, Sumatra. Biological Conservation 137, 20-27. McConkey, K. and M. Galetti. (1999). Seed dispersal by the sun bear (Helarctosmalayanus) in Central Borneo. Journal of Tropical Ecology 15:237-241. McLoughlin, P.D., Case, R.L., Gau, R.J., Ferguson, S.H., and Messier, F. (1999). Annual and seasonal movement patterns of barren-ground grizzly bears in the central Northwest Territories. Ursus 11, 79-86. McLoughlin, P.D., Ferguson, S.H., and Messier, F. (2000). Intraspecific Variation in Home Range Overlap with Habitat Quality: A Comparison among Brown Bear Populations. Evolutionary Ecology 14, 39-60. McNab, B. K. (1963). Bioenergetics and the determination of home range size. American Naturalist 97:133-140. McNab, B. K. (1983). Ecological and behavioral consequences of adaptation to various food resources. Pages 664-697 in J. F. Eisenberg and D. G. Kleiman, editors. Advances in the study of mammalian behavior. American Society of Mammalogists Special Publication 7. Miettinen, J., Shi, C., and Liew, S.C. (2011). Deforestation rates in insular Southeast Asia between 2000 and 2010. Global Change Biology 17, 2261-2270. Nomura, F., Higashi, S., Ambu, L., and Mohamed, M. (2004). Notes on oil palm plantation use and seasonal

spatial relationships of sun bears in Sabah, Malaysia. Ursus 15, 227–231. O’Brien, T.G., Kinnaird, M.F., Nurcahyo, A., Prasetyaningrum, M., and Iqbal, M. (2003). Fire, demography and the persistence of siamang (Symphalangus syndactylus: Hylobatidae) in a Sumatran rainforest. Ani- mal Conservation 6, 115-121. Ratnayeke, S., van Manen, F.T., and Padmalal, U.K.G.K. (2007). Home Ranges and Habitat Use of Sloth Bears Melursus ursinus inornatus in Wasgomuwa National Park, Sri Lanka. Wildlife Biology 13, 272-284. Redford, K. H. (1987). Ants and termites as food. Patterns of Mammalian Myrmecophagy.Pages 349-397 in H. H. Genoways, editor. Current Mammalogy. Plenum press, Nebraska. Sakai, S., K. Momose, T. Yumoto, T. Nagamitsu, H. Nagamasu, A. A. Hamid, and T. Nakashizuka. (1999). Plant reproductive phenology over four years including an episode of general ﬂowering in a lowland dipterocarp forest, Sarawak, Malaysia. American Journal of Botany 86:1414-1436. Schaik, C. P. v., (1986). Phenological changes in a Sumatran rain forest. Journal of Tropical Ecology 2:327-347. Synthesis and conclusions 175

Schaik, C. P. v., J. W. Terborgh, and S. J. Wright .(1993). The phenology of tropical forests: adaptive significance and consequences for primary consumers. Annual Review of Ecology and Systematics 24:353-377. Schwarzenberger, F., G. M. Fredriksson, K. Schaller, and L. Kolter. (2004). Fecal steroid analysis for monitoring reproduction in the sun bear (Helarctos malayanus). Theriogenology 62:1677-1692. Siegert, F., Ruecker, G., Hinrichs, A., and Hoffmann, A.A. (2001). Increased damage from fires in logged forests during droughts caused by El Niño. Nature 414, 437-440. Slik, J.W.F., Verburg, R.W., and Kessler, P.J.A. (2002). Effects of fire and selective logging on the tree species composition of lowland dipterocarp forest in East Kalimantan, Indonesia. Biodiversity & Conser- vation 11, 85-98. Slik, J. W. F. and K. A. O. Eichhorn. (2003). Fire survival of lowland tropical rain forest trees in relation to stem diameter and topographic position. Oecologia137:446-455. Sork, V. L. (1993). Evolutionary ecology of mast-seeding in temperate and tropical oaks (Quercus spp.). Vegetatio107/108:133-147. van Nieuwstadt, M.G.L. (2002). Trial by fire- Postfire development of a dipterocarp forest. Ph.D. thesis. University of Utrecht, Netherlands. Walsh, R. P. D. (1996). Drought frequency changes in Sabah and adjacent parts of northern Borneo since the late nineteenth century and possible implications for tropical rain forest dynamics. Journal of Tropical Ecology 12:385-407. Walsh, R. P. D. and D. M. Newbery. (1999). The ecoclimatology of Danum, Sabah, in the context of the world’s rainforest regions, with particular reference to dry periods and their impact. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences 354:1869-1883.

Webster, P. J. and T. N. Palmer. (1997). The past and the future of El Niño. Nature 390:544-564. Wich, S. A. and C. P. van Schaik. (2000). The impact of El Niño on mast fruiting in Sumatra and elsewhere in Malesia. Journal of Tropical Ecology 16:563-577. Wich, S., R. Buij, and C. van Schaik. (2004). Determinants of orangutan density in the dryland forests of the Leuser Ecosystem. Primates 45:177-182. Wich, S. A., E. R. Vogel, M. D. Larsen, G. M. Fredriksson, M. Leighton, C. P. Yeager, F. Q. Brearley, C. P. van Schaik, and A. J. Marshall. (2011). Forest Fruit Production Is Higher on Sumatra than on Borneo. PLoS ONE 6: Published online 2011 June 2028.doi: 2010.1371/ journal.pone.0021278. Wong, S.T., Servheen, C., and Ambu, L. (2004). Home range, movement and activity patterns, and bedding sites of Malayan sun bears, Helarctos malayanus in the rainforest of Borneo. Biological Conserva- tion 119, 168-181. 176 Chapter 8

Wong, S.T., Servheen, C., Ambu, L., and Norhayati, A. (2005). Impacts of fruit production cycles on Malayan sun bears and bearded pigs in lowland tropical forest of Sabah, Malaysian Borneo. Journal of Tropi- cal Ecology 21, 627–639. Wood, T. G. and W. A. Sands (1978). The role of termites in ecosystems. Pages 245-292 in M. V. Brian, editor. Production ecology of ants and termites. Cambridge University Press, Cambridge, UK. Woodroffe, R., and Ginsberg, J.R. (1998). Edge effects and the extinction of populations inside protected areas. Science 280, 2126-2128. Wright, S. J., C. Carrasco, O. Calderon, and S. Paton. (1999). The El Niño Southern Oscillation: variable fruit production, and famine in a tropical forest. Ecology 80:1632-1647. Acknowledgements

There are many people I would like to thank for helping, supporting and stimulating me throughout my research on sun bears and the conservation purpose this developed into. First of all are my parents, Einar and Zosia, who encouraged me during my initial biology endeavors and continued their enthusiastic support throughout the following 15 years, most of which I spent in Indonesia. Your ongoing interest and encouragement means a lot! Once my interest in tropical wildlife ecology had become rooted, Dr. Herman Rijksen is most to be thanked (or cursed...) for the route my life has taken. Herman initially encouraged me in 1994 to do my masters’ fieldwork on orangutans in Kalimantan, during which I had my first observation of two sun bears in the wild raiding a stingless bee’s nest high up in a giant Dipterocarp tree. Subsequently Herman supported and advised me through many different Indonesian conservation challenges, starting in Sungai Wain but later in North Sumatra and Aceh, and I am grateful for all his pep talks and advise. This fist sun bear encounter spurred my interest in this species, especially as I could not find any information on field studies carried out on this bear species anywhere in the wild. Once my interest in sun bear’s had been triggered, Dr. Dave Garshelis has been instrumental in many ways and I am grateful for his generous guidance and advice. Dave encouraged me, when I first contacted him in 1996 about sun bear research, to learn more about bears in general before embarking on a sun bear field study. Dave put me in touch with Prof. Jon Swenson, who then invited me to visit the Scandinavian Brown Bear project in Dalarna, Sweden, to learn the basics about various other aspects of bear fieldwork and telemetry. In Sweden I was welcomed by Sven Brunberg, who enthusiastically allowed me to join a variety of bear research activities, from aerial telemetry, moose follows, to sneeking up on a brown bears in the beautiful boreal forest. The months I spent with the Scandinavian Bear Project were invaluable training and preparation for studying sun bears in Indonesia. I would like to thank the Indonesian Institute of Sciences (LIPI) for granting me permission to study sun bears in Kalimantan, especially Ibu Ina Syarief and Ibu Krisbiwati for help with arranging the necessary permits and letters. I would like to thank Dr. Willie Smits for putting Sungai Wain on the map by releasing orangutans there and facilitating my initial research there, as well as building the nice research camps that became my home for many years. I would like to thank the Tropenbos Foundation, especially Tinus de Kam, for collaboration and support 180 Acknowledgements

with the initial Sungai Wain management plan we wrote in 1998, and for establishment of the permanent sample plots. I would like to thank the Wanariset Herbarium, especially Dr. Kade Sidiyasa, Ambriansyah, Arifin and Insyah for help with plant identification. Balai Penelitian Kehu- tanan (BPK), Bapak Boen Purnama and Bapak Selamat Gadaz, are gracefully thanked for their collaboration. I would also like to express my gratitude to the former Mayor of Balikpapan, Bapak Imdaad Hamid, and Bapak Saleh Basrie for their enthusiasm to establish a new management system for the Sungai Wain forest. I would gratefully like to acknowledge financial support for the research from the Tropenbos Foundation, the Gibbon Foundation, the American Zoo and Aquarium Associa- tion, the Balikpapan Orangutan Society, the World Society for Protection of Animals (Victor Watkins), the University of Amsterdam, the van Tienhoven Foundation, Virbac, the Artis Zoo (Peter Klaver), the International Bear Association, the Bevins Foundation, the Stichting Doctor Catharina van Tusschenbroek Fonds, and the Great Bear Foundation. During many years in the field several research assistants, especially Trisno, Lukas and Dami, provided invaluable help and I thank them for their dedication, hard fieldwork and com- panionship. I would like to thank the fire teams and subsequent coal fire extinguishing teams who helped between 1998-2001 to save the Sungai Wain forest, as well as Willie Smits and Al Whitehouse for financial and logistical support for the fire extinguishing efforts. I would especially like to thank Agusdin, Babam, Iman, Juahir and Parmin for all their help with the forest protection work, especially during the initial years. I would also like to thank a large number of villagers from Sungai Wain for helping as porters and bringing in logistical supplies to the forest, especially Almarhum Pak Yasin. In Balikpapan a number of people provided invaluable help and friendship, particularly Judy O’Dwyer, Sam and Yanti are thanked warmly for all their help with delivering fruit and food supplies for our bears during the fires and a comfortable place to sleep in Balikpapan where I could regain energy. I would also like to thank Jane Vincent who took care of sun bear cub Schizo for several months before we could take him to the forest, as well as for providing a welcome place for me to stay in Balikpapan. The Unocal School and Peter Karsono are thanked for looking after sun bear Ganja for some months until she got fed up being pestered by the gang of gibbons. Once my sun bear fieldwork had started to take shape, I want to thank Prof. Steph Menken for his offer to do a PhD at the University of Amsterdam. I would especially like to thank him for his patient reminders over the years to finish it, as I became caught up with more and more conservation work in Kalimantan and Sumatra. Acknowledgements 181

I would also like to thank various researchers for their collaboration over the many years working on sun bear ecology. Dr. Homathevi Rahman is thanked for the mammoth task of identifying more than 3000 termite samples whilst sitting in the forest. Dr. Decha Wiwatwittaya from Kasetsart University, Bangkok, Thailand is thanked for identification of ant samples. I would like to thank Live Solbrekken, Anders Endrestøl, and Karen Marie Matthiesen from the University in Ås, Norway, for their hard fieldwork in Sungai Wain in 2001. I want to thank Claire Palmer and Wiwit Sastramidjaja for their interest in sun bear research and for their help with coordinating and carrying out the sun sign transect work in 2005 and 2010, not always under easy conditions. I would like to thank Lydia Kolter and Rebecca Riese for collaboration on sun bear nutrition, as well as Franz Schwartzenberger for collaboration on sun bear reproductive patterns. I would like to thank Ente Rood and Matthew Nowak for assistance and advise on various statistical matters and Dandy Irawan and Jurgen van Gessel with translations of the summaries to Indonesian and Dutch, and Corien Smit for the lay-out of this thesis. I would also like to thank students from Universitas Mulawarman (UNMUL) in Sama- rinda in particular Wiwien Effendi, Joned, Eko and Ali Mustofa as well as their supervisors Dr. Chandradewana Boer and Dr. Agung Sardjono. I want to thank Nunuk Kasyanto for his ongoing friendship, visits to the forest, and the NGO LORIES for helping with various programmes in the forest and conservation awareness campaigns for the Sungai Wain forest in the early days. I am very grateful for all the support, advise and help Dave Garshelis provided over all the years, ranging from training on trapping bears to telemetry equipment and various visits to the forest, as well the Garshelis family for providing as a welcome home in Minnesota during a period of writing up, and to Judy, Daren and Brian for visiting the forest. Many thanks as well to Ruth and Laura Pierce for their warm hospitality in Grand Rapids and letting me stay at your home, and all the wonderful staff at the DNR especially Karen Noyce and Sharon, Nancy and July for showing me around Minnesota! I would like to thank Mark van Nieuwstadt for being such a nice ‘housemate’ in Camp Djamaludin where we lived together for quite some time, as well as for many stimulating discussions. I would also like to thank fellow researchers Martjan Lammertink, Vincent Nijman, Ferry Slik, Karl Eichhorn and Danny Cleary who spent some of their research time in Sun- gai Wain for interesting discussions and collaboration. Fellow Asian bear researchers Siew Te Wong, Rob Steinmetz, Mei-Hsiu Hwang, and Dave Augeri are also thanked for their nice visits to Sungai Wain, useful discussion and collaboration over the years, and sharing experiences and difficulties studying Asian bears. 182 Acknowledgements

I very much enjoyed the various visits of my friends and family over the years I lived in the forest in Sungai Wain, especially Vincent Nijman, Erik Meijaard and Graham Usher with whom I had many good talks and walks through the forest. The visits by Sven Brunberg, Darin Collins, Cheryl Fredericks, Anja Hoffman, Georg Buchholz, Frank Momberg, Resit Sözer, Danielle Kreb, Teguh, Rani, Daniel and Chris, Timbul, Nunuk, Cabok, Gigih, Serge Wich and Tine Geurts were much appreciated. My close friends Arlette and Jan Jaap, Bas Huijbregts and Jurgen van Gessel are thanked for their fun forest visit and subsequent holidays. The visit by the Fami- lie de Vos (Bunkie, Lucy, Claire and Fiona) to the forest in the early days is also fondly remem- bered. I would like to thank my parents Einar and Zosia for their visits, as well as my brother Hjalmar (and Martijn) who made me laugh so many times during his visits and hikes through the forest. Also many thanks to my friends in Amsterdam, with whom I have shared many of my bear stories and forest tribulations over the years and with whom I could revive socially, especially Eva and Frank, Arlette and Jan Jaap, Mark and Marchien, Jurgen, and Claire and family. Although no longer with us, I would also like to express my warm feelings and gratitude to my grandparents in Sweden and Poland, where I spent some time writing up intermittently during annual visits when back from the forest, especially my grandmothers Margit Fredriksson and Gabriela Biernacka. Bvll is thanked warmly for all his wonderful illustrations that have helped so much with the conservation awareness campaigns for the Sungai Wain forest and all our subsequent sun bear education work, as well as the wonderful cover and illustrations in this thesis! I would especially like to thank my wonderful companion and partner Graham Usher. Your many visits to the forest in the early years, and your ongoing support, encouragement, patience, enthusiasm throughout all aspects of my work with the bears, conservation of the forest, and subsequent years working on the education center, as well as many other conservation challenges are immensely appreciated! Without you I would probably not have achieved half! I would also like to thank you very much for all technical and GIS help you have provided, all the maps in this thesis, and for reading and commenting on many parts of this thesis. And last but certainly not least, immense gratefulness towards the wonderful sun bears that allowed me to get insight into their ecology and behavior, especially Ganja, Ucil, Schizo and si Kecil. Life in the forest never had a dull moment when they were around. The many years we had together in the forest include some of my happiest moments. I wish I could have done more for them and all other sun bears! Summary

The Malayan sun bear, Helarctos malayanus, smallest of the eight extant bear species, is distributed throughout Southeast Asia including the large islands of Sumatra and Borneo. Research on sun bears in the wild, including those in Kalimantan (Indonesian Borneo), was non-existent before the work that is presented in this thesis commenced in 1997, in a small lowland reserve in East Kalimantan. Apart from a few anecdotal records, virtually nothing was known about basic aspects of sun bear ecology and they were listed as ‘Data Deficient’ by the IUCN Red list at the time. In order to fill in some of the bigger gaps of knowledge on this species, data were collected on various aspects of sun bear behaviour, ecology, and conservation during 5 years of intensive research from mid 1997 until mid 2002, with further surveys carried out in 2005 and 2010. During the early stages of this study, large-scale forest fires, related to severe drought in the region triggered by recurring El Niño Southern Oscillation (ENSO) events, affected millions of hectares of forest in Kalimantan, including the reserve where this study had commenced. Between 1997 and 2006, a total of 16.2 Mha of Borneo’s landmass (21% of the total land surface area) have been affected by fires, a considerable percentage of this being forested areas. Prompted by the scale of this environmental disaster, part of my research focused on investigating the effects of these fires on the ecology and conservation of sun bears. ENSO events, besides causing drought conditions in the region, have also been linked to mast fruiting in Southeast Asia, especially the eastern parts of Borneo. The canopy and emergent trees of these forests are dominated by trees of the Dipterocarpaceae, which display a unique reproductive pattern. Flowering and fruiting occurs at intervals of 2-10 years with little or no reproductive activity in between, and a large proportion of other trees covering varied taxonomic groups and flowering syndromes, flower and fruit synchronously with these dipterocarps. These masting events provide a brief period of high fruit availability, aimed at sati- ating seed predators, but also providing a glut of nutritious fruits. Through monthly phenology surveys of more than 600 trees, we documented that sun bear fruit availability decreased from 13 trees ha-1 fruiting month-1 during the mast fruiting to only 1.6 trees ha-1 during the prolonged inter-mast period. Sun bears were found to become almost completely frugivorous during these short periods (approx. 2 months) of high fruit availability. During this study, sun bears were found 186 Summary

to feed on at least 115 fruit species, with the families Moraceae, Burseraceae and Myrtaceae contributing more than 50% to the sun bear fruit diet, whereas Ficus spp. (Moraceae) were the main fallback fruit during the prolonged intermast. Most sun bear fruit species only fruited during the rare masting events or displayed supra-annual fruiting patterns, indicating sporadic availability of favored fruits. During the prolonged intermast period, subsequent to the fire event, sun bears switched to an insectivorous diet, dominated by termites, ants, beetles and beetle grubs, and cockroaches. Two female bears captured for telemetry studies in the unburned forest core, 16 months after the fire event and well into the inter-mast period, showed signs of serious undernourishment. Both females had small cubs at the time, probably exacerbating their already weakened condition. The emaciated appearance of the female sun bears suggests that prolonged survival solely on insect resources was insufficient and that energy requirements were not met. The predation of one of the bears by a large python might have been related to the poor nutritional condition of the bear, possibly rendering the bear less effective in fighting off an attack or having a lowered vigilance. To gain insight into the movement and activity patterns of sun bears, three wild and two rehabilitated-released female bears were fitted with motion-sensitive VHF radio-collars. Three of these bears were tracked continuously for 1-3 years during 1999-2003. Home range sizes (12-month period) were 4-5 km2 (95% MCP and 95% fixed kernel) with considerable home range overlap. These relatively small home ranges may have been related to a diet composed predominantly of insects, as no mast fruiting event occurred during the period of telemetry study. Activity monitoring revealed that sun bears were predominantly diurnal and crepuscular throughout the year with activity commencing about 0.5 h before sunrise and lasting until 2.5h after sunset, and low levels (20-30%) of activity at night. Bears were active 61.6% of the day (14.8 ± 2.3 hours), on average, with higher levels of activity associated with insect feeding than fruit feeding. Forest fires and drought caused extreme mortality of sun bear fruit trees, with close to an 80% reduction in densities of fruit trees in burned areas from 167 ± 41 (SD) fruit trees ha-1 in unburned forest to 37 ± 18 fruit trees ha-1 in burned forest. In addition a 44% reduction in species diversity of sun bear fruit trees was measured some 3 years after the fire event. Densi- ties of hemi-epiphytic figs, an important fruit species during the prolonged intermast period, declined by 95% in burned forest. Sun bear use of burned forest areas was monitored through repeated sign surveys (i.e. claw marks, broken termite mounds, diggings) over a period of 10 years, with transects carried Summary 187

out in adjacent primary and burned forest areas. Densities of sun bear sign were close to zero in burned forest areas for several years after the fire, but by 2010 had increased to 28.0 ± 12.9 (SD) sign/ha, indicating re-colonization by sun bears. Sign densities in adjacent unburned forest remained stable over the years, averaging 41.6 ± 11.6 (SD) sign/ha. Densities of sun bear sign were still 35% lower in burned forest 12 years after the fire event, linked to a paucity of certain types of sun bear foods, especially fruiting trees and above-ground termite colonies. Steady recolonization of sun bears into these once-burned areas indicates the recovery potential of these areas as sun bear habitat, as long as source population of sun bears are present in adjacent unburned forest. Subsequent to the drought and fire event, we carried out repeated interviews with farmers in 5 communities situated along the edge of the Sungai Wain Protection Forest to investigate sun bear crop raiding. Results from these interviews indicated that crop damage caused by sun bears subsequent to the forest fires was higher than in years prior to the ENSO and fire event. The main source of antagonism toward bears resulted from the damage they caused to stands of old coconut trees (Cocos nucifera), frequently killing the trees. This prompted farmers to seek removal of the bears. Bear damage to annual crops generally spurred a less hostile reaction. Experiments with metal sheeting affixed to the trunks of coconut trees to deter climbing by bears were successful, at least in the short term (≥3 years). Inexpensive and easily applicable crop-protection devices such as this could help protect sun bears in the future, as increased human-bear conflicts are anticipated due to rapid human population growth, unabated forest destruction and fragmentation, and increased susceptibility of remaining forests to future fires. Habitat degradation due to large-scale forest fires, combined with a prolonged fruiting low after the ENSO event, put large strains on the sun bear population in the area. As sun bear habitat is continuing to be destroyed at an alarming rate in Borneo, even small reserves are becoming more important for the conservation of this species. Behavioural characteristics such as a relatively small home range size, bounded entirely within the small reserve where this research was carried out, and avoidance of areas with human disturbance (trails, burned forest) suggest that small reserves can still be valuable for the conservation of sun bears. Long-term persistence of sun bears in these areas will depend on prevention of poaching, forest degradation from fires, and maintaining connectivity to other forested areas. Samenvatting

De Maleise zonbeer, Helarctos malayanus, is de kleinste van de huidige acht beersoor- ten. De zonbeer komt voor door heel Zuidoost-Azië, ook op de grote eilanden van Sumatra en Borneo. Het onderzoek dat ik in dit proefschrift presenteer ging in 1997 van start in een klein laaglandreservaat in Oost-Kalimantan. Voor die tijd was er nog geen veldonderzoek gedaan naar zonberen in het wild; ook niet naar die in Kalimantan, het Indonesisch deel van Borneo. Behalve de informatie uit een paar anekdotische verslagen was er nagenoeg niets bekend van de fundamentele aspecten van de ecologie van de zonbeer. De soort stond toen nog in de categorie ‘Data Deﬁcient’ (‘onvoldoende gegevens’) op de Rode Lijst van de IUCN. Om een voornaam deel van de ontbrekende kennis over de zonbeer aan te vullen, hebben we in het kader van deze studie gedurende vijf jaar (van halverwege 1997 tot halverwege 2002) intensief data verzameld rond verschillende aspecten van het gedrag, de ecologie en het behoud van de zonbeer. Verder onderzoek is uitgevoerd in 2005 en in 2010. Ten tijde van de eerste fase van deze studie had de regio te maken met grootschalige bosbranden, veroorzaakt door ernstige droogte die gerelateerd was aan herhaaldelijke El Niño Southern Oscillation (ENSO)-verschijnselen. De branden tastten in Kalimantan miljoenen hectare bos aan, inclusief het reservaat waar het onderzoek plaatsvond. Tussen 1997 en 2006 is in totaal 16,2 miljoen hectare van Borneo’s landmassa getroffen door brand; dat is 21 procent van de totale landoppervlakte van het eiland. Een groot gedeelte hiervan bestond uit bosgebieden. Vanwege de grote schaal van deze milieuramp heb ik een deel van mijn studie gewijd aan het onderzoeken van de effecten van de branden op de ecologie en het behoud van zonberen. De ENSO-verschijnselen veroorzaken niet alleen droogte, maar spelen ook een rol bij het voorkomen van mast fruiting (fenomeen waarbij grote hoeveelheiden bomen gelijktijdig fruit produceren) in Zuidoost-Azië, in het bijzonder in de oostelijke delen van Borneo. Het bladerdak en de emergente bomen in de bossen hier worden gedomineerd door bomen uit de familie van de Dipterocarpaceae die een uniek regeneratief patroon laten zien. De bloei en het dragen van vruchten doen zich voor met intervallen van 2 tot 10 jaar; in de tussenperioden vertonen deze bomen geen of bijna geen voortplantingsactiviteit. De bloei en het dragen van fruit door een groot deel van de andere bomen – afkomstig uit verschillende taxonomische groepen met gevarieerde bloeiverschijnselen – verloopt hier synchroon met deze Dipterocar- paceae. Mast fruiting voorziet daardoor in een korte periode waarin fruit in zeer grote mate 190 Samenvatting

beschikbaar is; de functie hiervan is zaadpredatoren te verzadigen en voorziet ook in een overvloed aan voedzame vruchten. Door maandelijks fenologisch onderzoek aan meer dan 600 bomen hebben we kunnen vastleggen dat de beschikbaarheid van fruit voor de zonberen afnam van 13 fruitende bomen per hectare/maand tijdens de mast fruiting-periode tot slechts 1,6 bomen die fruit produceerden per hectare/maand tijdens de langdurige inter-mast-periode. Het bleek dat zonberen bijna compleet frugivoor worden tijdens deze korte perioden van ongeveer 2 maanden waarin het fruit zo rijkelijk aanwezig is. In deze studie stelden we vast dat zonberen zich voeden met minstens 115 verschillende soorten fruit. De families Moraceae, Burseraceae en Myrtaceae dragen voor meer dan 50 procent bij aan het dieet van de zonberen; de beren vielen in de langdurige inter-mast-periode vooral terug op het fruit van het genus Ficus spp. (Moraceae). De meeste boomsoorten waarvan de zonberen het fruit eten droegen alléén fruit tijdens de zeldzame mast fruiting-periodes óf vertoonden fruitpatronen die zich uitstrek- ken over meer jaren; dat duidt op sporadische beschikbaarheid van het favoriete fruit van de zonberen. Tijdens de langdurige inter-mast-periode, volgend op de bosbranden, schakelden de zonberen over op een dieet van insecten, vooral termieten, mieren, kevers en keverlarven, en kakkerlakken. Twee vrouwelijke beren die we vingen voor een telemetriestudie in het niet-verbrande hart van het bos vertoonden tekenen van ernstige ondervoeding, 16 maanden na de branden en diep in de inter-mast-periode. Beide vrouwtjes hadden op dat moment jonge welpen, waar- door ze waarschijnlijk verder verzwakt waren. De uitgemergelde verschijning van de beren suggereert dat een dieet van niets anders dan insecten gedurende een langere periode niet afdoende is en niet in de energiebehoefte van de beren voorziet. De predatie van een van deze vrouwtjesberen door een grote python zou gerelateerd kunnen zijn aan haar slechte conditie; mogelijk kon zij de aanval minder effectief weerstaan of was ze minder waakzaam. Om inzicht te krijgen de bewegings- en activiteitenpatronen van zonberen, zijn drie wilde en twee vrijgelaten gerehabiliteerde vrouwelijke beren uitgerust met VHF-radiokragen met bewegingssensors. Drie van deze beren zijn continu gevolgd gedurende 1 tot 3 jaar in de periode van 1999 tot 2003. Hun home range omvatte (in een periode van 12 maanden) 4 tot 5 km2 (95 procent Minimum Convex Polygon en 95 procent ﬁxed kernel) met een aanzienlijke onderlinge overlap van de home ranges. De relatief kleine omvang van de home ranges kan gerelateerd zijn aan een dieet dat voornamelijk bestond uit insecten, omdat zich in de periode van de telemetriestudie geen mast fruiting voordeed. Het monitoren van de activiteit toonde aan dat zonberen het hele jaar door hoofdzakelijk dag- en schemerdieren zijn. Ze worden ongeveer een half uur voor zonsopgang actief en blijven dat tot 2,5 uur na zonsondergang. Het Samenvatting 191

activiteitsniveau ’s nachts is laag (20 tot 30 procent). De beren waren gemiddeld 61,6 procent van de dag actief (14,8 ± 2,3 uur); hogere activiteitsniveaus hingen sterker samen met het eten van insecten dan met het eten van fruit. De bosbranden en de droogte veroorzaakten extreme sterfte van fruitbomen die ges- chikt zijn voor zonberen, met een reductie van bijna 80 procent in dichtheid van fruitbomen in verbrande gebieden: van 167 ± 41 (SD) fruitbomen per hectare in onaangetast bos tot 37 ± 18 fruitbomen per hectare in verbrand bos. Bovendien stelden we een reductie van 44 procent in de diversiteit van fruitbomen voor zonberen vast ongeveer 3 jaar na de brand. De dichtheden van hemi-epifytische vijgen – een belangrijke fruitsoort tijdens de inter-mast-periode – namen af met 95 procent in verbrande delen van het bos. De manier waarop de zonberen het verbrande bos benutten is gedurende 10 jaar lang gemonitord door regelmatig onderzoek naar sporen (zoals gebroken termietenheuvels, klau- wsporen en graafsporen). Transecten zijn uitgevoerd in aan elkaar grenzende stukken van verbrand en ongeschonden bos. In een aantal jaren na de branden waren de dichtheden van zonbeersporen bijna nul in de aangetaste delen van het bos. Maar in 2010 waren die dichtheden weer toegenomen tot 28,0 ± 12,9 (SD) sporen per hectare; de toename wijst op rekolonisatie van deze delen van het bos door de zonberen. De dichtheid van sporen in de aangrenzende ongeschonden delen van het bos bleef in de loop der jaren stabiel, gemiddeld 41,6 ± 11,6 (SD) sporen per hectare. De dichtheden van zonbeersporen waren 12 jaar na de bosbranden nog steeds 35 procent lager in de verbrande delen van het bos; dit had te maken met schaarste van sommige soorten voedsel voor zonberen, vooral fruitbomen en bovengrondse termietenkolo- nies. De gestage rekolonisatie van deze verbrande delen van het bos door de beren laat zien dat deze gebieden zich kunnen herstellen tot geschikte habitat voor de zonbeer; voorwaarde is wel dat er een bronpopulatie van zonberen aanwezig is in aangrenzende ongeschonden delen van het bos. Aansluitend op de periode van droogte en de bosbranden hebben wij boeren her- haaldelijk geïnterviewd in vijf gemeenschappen die langs de randen van het Sungai Wain Protec- tion Forest gesitueerd zijn. Het doel was te onderzoeken in welke mate de zonberen gewassen plunderen. De uitkomsten van de interviews gaven aan dat schade aan de gewassen groter was in de periode volgend op de bosbranden dan in de jaren vóór de ENSO-verschijnselen en de branden. Vijandigheid van de boeren ten opzichte van de beren was vooral het gevolg van de schade die ze toebrachten aan stammen van oude kokospalmen (Cocos nucifera); als gevolg daarvan gingen de bomen vaak dood. Daarom wilden de boeren de beren weg hebben. De schade die beren toebrachten aan de jaarlijkse gewassen leidde in het algemeen tot een minder 192 Samenvatting

vijandige reactie. Een experiment met het vastmaken van metalen platen aan de stammen van de kokospalmen om beren het klimmen onmogelijk te maken was succesvol, in elk geval op de korte termijn (≥ 3 jaar). Goedkope en gemakkelijk toe te passen gewasbeschermingsmiddelen als deze zouden in de toekomst kunnen helpen zonberen te beschermen; een toename in de conﬂicten tussen mensen en beren wordt verwacht door de groei van het aantal mensen, de niet-aﬂatende vernietiging en fragmentatie van het bos, en de toegenomen vatbaarheid voor brand van de stukken bos die overblijven. De achteruitgang van de habitat door de grootschalige bosbranden in combinatie met een langere periode met weinig fruit na de ENSO-verschijnselen, zette de zonberenpopulatie in het gebied zwaar onder druk. Omdat de habitat van de zonberen op Borneo nog steeds in een schrikbarend hoog tempo wordt vernietigd, worden ook kleine reservaten steeds belang- rijker voor het behoud van deze soort. Ook gedragskenmerken van de beren wijzen erop dat kleine reservaten van waarde kunnen zijn voor het behoud van de zonberen: onder andere de relatief kleine home ranges – volledig begrensd binnen het kleine reservaat waar dit onderzoek is uitgevoerd – en het vermijden van gebieden met verstoring door mensen, zoals paden en verbrande delen van het bos. Het overleven van zonberen in deze gebieden op de lange termijn zal afhangen van de preventie van stropen en de instandhouding van verbindingen met andere bosgebieden. Ringkasan

Beruang Madu (Helarctos malayanus), jenis terkecil dari 8 jenis beruang yang ada di dunia, tersebar di seluruh negara di daratan Asia Tenggara serta di Pulau Sumatera dan Kalimantan (Borneo). Penelitian terhadap beruang madu di alam bebas, termasuk Kalimantan, tidak pernah ada sebelum penelitian yang ditampilkan dalam tesis ini yang dimulai pada tahun 1997 di suatu hamparan kecil hutan dataran rendah di Kalimantan Timur. Terlepas dari beberapa catatan singkat, aspek dasar ekologi beruang madu hampir tidak diketahui pada saat itu dan beruang madu didaftar sebagai ‘Data Deﬁcient’ (Kekurangan Data) dalam Daftar Merah IUCN (IUCN Red List). Untuk mengurangi kesenjangan pengetahuan atas jenis ini, saya kumpulkan data mengenai berbagai aspek beruang madu mulai dari perilaku, ekologi dan konservasi selama 5 tahun penelitian intensif (mendalam) dari pertengahan 1997 sampai dengan pertengahan 2002, dengan penelitian lebih lanjut yang dilakukan pada tahun 2005 dan 2010. Pada awal penelitian ini, kebakaran hutan besar yang dikarenakan kemarau berkepanjangan di daerah tersebut terkait dengan terjadinya El Niño Southern Oscillation (ENSO), telah mempengaruhi jutaan hektar hutan di Kalimantan, termasuk hutan lindung di mana penelitian ini berlangsung. Antara tahun 1997-2006, sekitar 16.2 juta hektar dari luas wilayah Borneo (21% dari luas total daratannya) telah terpengaruhi kebakaran, dan persentase yang besar dari luasan terebut merupakan hutan. Merujuk kepada skala bencana alam ini, sebagian dari penelitian saya fokuskan pada dampak dari kebakaran tersebut terhadap ekologi dan konservasi beruang madu. Peristiwa ENSO, selain menyebabkan kondisi kekeringan di daerah tersebut, juga berkaitan dengan proses musim buah besar (panen raya, mast fruiting) di Asia Tenggara, khususnya di bagian timur Pulau Kalimantan. Kanopi dan pohon-pohon besar di dalam hutan-hutan ini didominasi oleh pohon dari marga Dipterocarpaceae, yang menampilkan pola reproduksi yang unik. Proses pembungaan dan pembuahan berlangsung dengan senjangan waktu 2-10 tahun dengan sedikit ataupun tidak adanya aktivitas reproduksi selama senjangan waktu tersebut, dan sebagian besar dari pohon dari marga lainnya di hutan, terdiri dari berbagai kelompok taksonomi dan sindrom pembungaan yang berbeda-beda, berbunga dan berbuah bersamaan dengan pohon dipterokarpa. Kejadian panen raya buah menyediakan buah dalam jumlah banyak dalam periode yang singkat, yang bertujuan untuk memuaskan para pemakan biji, namun juga menyediakan buah bernutrisi berlebihan. Dari penelitian fenologi selama beberapa tahun terhadap lebih dari 196 Ringkasan

600 pohon, kami mencatat bahwa ketersediaan buah makanan beruang madu menurun dari 13 pohon ha-1 bulan-1 selama panen raya buah menjadi hanya 1.6 pohon ha-1 bulan-1 selama periode panjang antar musim panen raya buah. Ditemukan bahwa beruang madu hampir tidak menkomsumsi makanan selain buah selama masa-masa periode singkat (kira-kira 2 bulan) tersedianya buah. Selama penelitian ini, beruang madu diketahui mengkonsumsi setidaknya 115 spesies buah, dengan lebih dari 50% buah yang dimakan berasal dari marga Moraceae, Burseraceae dan Myrtaceae, sedangkan Ficus spp. (Moraceae) merupakan kelompok buah utama yang dikonsumsi beruang madu selama period antar musim buah besar. Kebanyakan jenis buah yang dimakan oleh beruang madu hanya berbuah selama masa panen raya buah atau menampilkan pola berbuah yang ‘supra-annual’ (tidak setiap tahun). Hal ini menunkukkan ketersediaan buah makanan beruang yang sangat sporadis. Selama masa antar musim panen raya buah, yang mengikuti peristiwa kebakaran hutan, beruang madu berubah menjadi pemakan serangga, yang didominasi oleh rayap, semut, kumbang, anak kumbang (larva) dan kecoa. Dua beruang madu betina yang ditangkap untuk penelitian radio-telemetry (radio pemantauan jarak) di wilayah inti hutan yang tidak terbakar, 16 bulan setelah peristiwa kebakaran dan musim panen raya buah, menunjukkan tanda kekurangan nutrisi yang serius. Kedua betina tersebut mempunyai anak yang kecil, yang kemungkinan membuat kondisi fisik mereka lebih tertekan. Kondisi malnutrisi pada beruang madu betina menunjukkan bahwa bergantung pada memakan hanya serangga selama jangka waktu yang panjang tidak cukup, dan sumber makanan ini tidak mencukupi kebutuhan energi mereka. Pemangsaan salah satu beruang madu betina oleh ular sawah besar mungkin berkaitan dengan kondisi malnutrisi yang dialaminya, yang menyebabkan dia kurang mampu membela diri atau kurang wasapada. Agar dapat memahami gerak gerik dan pola aktifitas beruang madu, 3 ekor beruang madu betina liar dan dua ekor yang dilepasliarkan, dipasangkan kalung radio VHF yang peka gerak (motion-sensitive VHF radio-collars). Tiga dari beruang ini terpantau secara berkelanjutan selama 1-3 tahun selama 1999-2003. Wilayah jelajah (periode 12 bulan) adalah 4-5km2 (95% Poligon konvex minimal atau MCP dan 95% fixed kernel) dengan tumpang tindah antara wilayah jelajah masing-masing individu yang cukup besar. Wilayah jelajah yang cukup kecil ini kemungkinan berhubungan dengan pola makan yang didominasi oleh serangga, dan tidak adanya periode panen raya buah di hutan selama penelitian telemetry. Aktifitas monitoring menunjukkan bahwa beruang madu mempunyai pola aktifitas diurnal (aktif pada siang hari) dan crepuscular (aktif pada saat fajar dan senja) sepanjang tahun dengan aktifitas dimulai pada 0.5 jam sebelum matahari terbit dan berlangsung sampai 2.5 jam setelah matahari terbenam, dengan aktifitas rendah di Ringkasan 197

malam hari (20-30%). Beruang aktif selama 61.6% dalam kurun waktu 24 jam (14.8 ± 2.3 jam), dengan aktifitas lebih tinggi ketika memakan serangga dibanding ketika memakan buah. Kebakaran hutan dan kemarau panjang menyebabkan mortalitas ekstrim pada pohon- pohon buah yang dimakan beruang madu sehingga berkurang hampir 80% di daerah yang terbakar, dari 167 ± 41 (SD) pohon jenis yang dimakan oleh beruang ha-1di hutan yang tidak terbakar menjadi 37 ± 18 pohon buah ha-1 di hutan yang terbakar. Selain itu, juga diukur pengurangan 44% pada keragaman jenis pohon buah 3 tahun setelah peristiwa kebakaran. Kepadatan pohon ara setengah epifit (hemi-epiphytic figs), jenis buah yang penting selama musim paceklik, menurun sebanyak 95% di area yang terbakar. Pemanfaatan hutan yang terbakar oleh beruang madu dipantau melalui penelitian tanda/bekas (misalnya tanda cakaran kuku, bongkaran sarang rayap, penggalian) berkala selama periode 10 tahun, dengan peninjauan transek dilakukan di hutan primer dan hutan terbakar yang berdekatan. Tanda beruang madu hampir tidak ada di wilayah terbakar selama beberapa tahun setelah kebakaran, tapi pada tahun 2010 telah meningkat menjadi 28.0 ± 12.9 (SD) tanda ha-1. Hal ini mengindikasikan re-kolonisasi oleh beruang madu. Kepadatan tanda/bekas di hutan primer terlihat stabil dengan rata-rata 41.6 ± 11.6 (SD) tanda ha-1 sepanjang periode monitoring. Kepadatan tanda beruang madu masih lebih rendah 35% di wilayah yang terbakar 12 tahun setelah peristiwa kebakaran, terkait dengan degradasi/ pengurangan dari beberapa jenis makanan beruang madu, khususnya pohon buah dan sarang rayap di atas tanah. Re-koloniasasi beruang madu pada area yang dulunya terbakar mengindikasikan potensi pemulihan area ini sebagai habitat beruang madu, selama masih ada sumber populasi beruang madu di hutan yang tidak terbakar yang dekat. Setelah kemarau panjang dan peristiwa kebakaran, kami melakukan wawancara berkala dengan petani di 5 pemukiman yang terletak di sepanjang pinggir Hutan Lindung Sungai Wain untuk mengetahui gangguan beruang madu terhadap tanaman. Hasil dari wawancara tersebut mengindikasikan bahwa pengrusakan tanaman oleh beruang madu meningkat setelah terjadi kebakaran hutan dan ENSO. Kerusakan pada pohon kelapa tua yang seringkali mematikan pohon tersebut menjadi sumber konflik paling tinggi dan membuat para petani berkeinginan memusnahkan beruang. Pada umumnya, pengrusakan oleh beruang terhadap tanaman tahunan tidak menimbulkan reaksi berat dari para petani. Percobaan dengan menempelkan lapisan senk pada batang pohon kelapa berhasil menghalangi beruang memanjat pohon kelapa, setidaknya pada periode singkat (sekitar 3 tahun). Pengamanan tanaman petani dengan pola yang mudah dipraktekkan dan terjangkau seperti ini dapat membantu melindungi beruang ke depan, karena konflik antara manusia dan beruang kemungkinan akan meningkat seiring dengan pertumbuhan 198 Ringkasan

populasi manusia yang pesat, kerusakan dan fragmentasi hutan yang berlangsung terus-menerus serta meningkatnya ancaman persistiwa kebakaran pada hamparan hutan yang tersisa. Degradasi habitat yang disebabkan kebakaran hutan berskala besar, dikombinasikan dengan ketersediaan buah yang tidak memadai setelah peristiwa ENSO, membuat populasi beruang madu tertekan di wilayah ini. Dengan rusaknya habitat beruang madu dengan laju yang menkhawatirkan di Pulau Kalimantan, hamparan hutan yang kecil pun menjadi sangat penting untuk melindungi jenis beruang ini. Karakteristik perilaku seperti wilayah jelajah yang relatif kecil, yang dibatasi dalam hutan lindung seperti ditemukan dalam penelitian ini, dan sifat beruang untuk menghindari wilayah dimana terdapat gangguan manusia, menunjukkan bahwa hutan kecil pun masih sangat penting bagi perlindungan beruang madu. Keberadaan jangka panjang beruang madu di wilayah seperti ini akan bergantung pada pencegahan perburuan, kebakaran hutan, dan pemeliharaan hubungan langsung dengan daerah hutan lainnya. Curriculum Vitae

Gabriella Fredriksson was born on the 27th of June 1971 in Amsterdam, Netherlands. After finishing high school, she studied biology at the University of Amsterdam and carried out her MSc fieldwork in 1994 on reintroduced orangutans in East Kalimantan, Indonesia. In 1996 she did a short period of fieldwork in Gabon on mandrills, after which she returned in 1997 to East Kalimantan to start research on sun bears. During her work on sun bears she became involved in forest protection activities, starting with fighting fires and illegal loggers, followed by setting up an innovative collaborative management system for the Sungai Wain Protection Forest, which became formalized in local legislation in 2002. In the same year, the sun bear was proposed as the official mascot of Balikpapan. Starting in 2004 she initiated the development of an environmental education facility in East Kalimantan, focusing on sun bears and conservation, and though still further developing the facility currently receives over 50,000 local visitors per year. Gabriella was honoured for her conservation work by being knighted in the ‘Order of the Golden Ark’ in 2004 bestowed upon her by HRH Prince Bernhard. In 2006 Gabriella started working in North Sumatra to establish protection for an area of primary forest that is home to the southernmost population of Sumatran orangutans. Between 2008-2009 she became involved with a forest redesign team established by the Governor of Aceh. Currently her time is divided between conservation work in East Kalimantan and North Sumatra. Since 2004 she has co-chaired the Sun Bear Expert Team of the IUCN/SSC Bear Specialist Group. List of publications

Fredriksson, G. M. (1995). Reintroduction of orangutans: A new approach. Study on the behaviour and ecology of reintroduced orangutans in the Sungai Wain Nature Reserve, East Kalimantan, Indonesia. University of Amsterdam, Amsterdam, Netherlands. Fredriksson, G. M. (1996). Conservation status of the mandrill (Mandrillus sphinx) in Central Africa. Ecologie en Ontwikkeling 4. Fredriksson, G. M. and M. de Kam. (1999). Strategic plan for the conservation of the Sungai Wain Protec- tion Forest, East Kalimantan, Indonesia. The International Ministry of Forestry and Estate Crops- Tropenbos Kalimantan Project. Fredriksson, G. M. (2001). Conservation threats facing sun bears, Helarctos malayanus, and experiences with sun bear reintroductions in East Kalimantan, Indonesia. In: The evaluation of bear rehabilitation projects from a conservationist viewpoint, Ouwehands Zoo, Rhenen, Netherlands. Fredriksson, G. M. (2002). Extinguishing the 1998 forest fires and subsequent coal fires in the Sungai Wain Protection Forest, East Kalimantan, Indonesia. Pages 74-81 In Communities in flames: proceedings of an international conference on community involvement in fire management. FAO and FireFight SE Asia, Bangkok, Thailand. Wich, S. A., G. Fredriksson, and E. H. M. Sterck. (2002). Measuring fruit patch size availability for three sympatric Indonesian primates. Primates 43:19-27.

Fredriksson, G. and V. Nijman. (2004). Habitat use and conservation status of two elusive ground brids (Carpococcyx radiatus and Polyplectron scheiermacheri) in the Sungai Wain Protection Forest, East Kalimantan, Indonesian Borneo. Oryx 38:297-303. Schwarzenberger, F., G. M. Fredriksson, K. Schaller, and L. Kolter. (2004). Fecal steroid analysis for monitoring reproduction in the sun bear (Helarctos malayanus). Theriogenology 62:1677-1692. Fredriksson, G. M. (2005). Human sun bear conflicts in East Kalimantan, Indonesian Borneo. Ursus 16:130-137. Fredriksson, G. M. (2005). Predation on sun bears by reticulated python in East Kalimantan, Indonesian Borneo. The Raffles Bulletin of Zoology 53:197-200. Fredriksson, G. M., S. A. Wich, and Trisno. (2006). Frugivory in sun bears (Helarctos malayanus) is linked to El Niño-related fluctuations in fruiting phenology, East Kalimantan, Indonesia. Biological Journal of the Linnean Society 89:489-508. 204 Curriculum vitae and list of publications

Fredriksson, G. M., L. S. Danielsen, and J. E. Swenson. (2006). Impacts of El Niño related drought and forest ﬁres on sun bear fruit resources in lowland dipterocarp forest of East Borneo. Biodiversity and Conservation 15:1271-1301. Tumbelaka, L. and G. M. Fredriksson. (2006). The status of sun bears in Indonesia. Pages 73-78 In: Under- standing Asian bears to secure their future. Japan Bear Network, Ibaraki, Japan. Fredriksson, G., Steinmetz, R., Wong, S. & Garshelis, D.L. (IUCN SSC Bear Specialist Group) (2008). Helarc- tos malayanus. In: IUCN 2012. IUCN Red List of Threatened Species. Version 2012.1. . Downloaded on 01 September 2012. Marshall, A. J., M. Ancrenaz, F. Q. Brearley, G. Fredriksson, N. Ghaffar, M. Heydon, S. Husson, M. Leighton, K. R. McConkey, H. Morrogh-Bernard, J. Proctor, C. P. van Schaik, C. P. Yeager, and S. A. Wich. (2009). The effects of forest phenology and ﬂoristics on populations of Bornean and Sumatran orangutans: are Sumatran forests more productive than Bornean forests? Pages 97–117 in S. A. Wich, S. S. Utami- Atmoko, T. Mitra Setia, and C. P. van Schaik, editors. Orangutans: Geographic variation in behavioral ecology and conservation. Oxford University Press, Oxford. Slik, J. W. F., S.-I. Aiba, M. Bastian, F. Q. Brearley, C. H. Cannon, K. A. O. Eichhorn, G. Fredriksson, K. Kartawi- nata, Y. Laumonier, A. Mansor, A. Marjokorpi, E. Meijaard, R. J. Morley, H. Nagamasu, R. Nilus, E. Nurt- jahya, J. Payne, A. Permana, A. D. Poulsen, N. Raes, S. Riswan, C. P. van Schaik, D. Sheil, K. Sidiyasa, E. Suzuki, J. L. C. H. van Valkenburg, C. O. Webb, S. Wich, T. Yoneda, R. Zakaria, and N. Zweifel. (2011). Soils on exposed Sunda Shelf shaped biogeographic patterns in the equatorial forests of Southeast Asia. Proceedings of the National Academy of Sciences 108:12343-12347. Wich, S. A., E. R. Vogel, M. D. Larsen, G. M. Fredriksson, M. Leighton, C. P. Yeager, F. Q. Brearley, C. P. van Schaik,

and A. J. Marshall. (2011). Forest Fruit Production Is Higher on Sumatra Than on Borneo. PLoS ONE 6:Published online 2011 June 2028. doi: 2010.1371/journal.pone.0021278. Wich, S. A., G. M. Fredriksson, G. Usher, H. H. Peters, D. Priatna, F. Basalamah, W. Susanto, and H. Kühl. (2012). Hunting of Sumatran orang-utans and its importance in determining distribution and density. Biological Conservation 146:163-169. Effects of El Niño and forest ﬁ of sun bears,res and conservation on the ecology Indonesian Borneo Effects of El Niño and large-scale forest ﬁ res on the ecology and conservation of Malayan sun bears (Helarctos malayanus) in East Kalimantan, Indonesian Borneo G. Fredriksson

Gabriella Fredriksson