This paper contains supplementary material that can be found online at http://journals.cambridge.org
Introduction
Although poaching has been shown to be the most serious short-term threat to the tiger Panthera tigris (Chapron et al., Reference Chapron, Miquelle, Lambert, Goodrich, Legendre and Clobert2008), prey depletion may also be an extinction driver for this species (Karanth & Stith, Reference Karanth, Stith, Seidensticker, Christie and Jackson1999). In tiger range states such as Malaysia lobbying efforts have focused on amending outdated legislation to impose harsher penalties on poachers and consumers of tigers. However, improved protection of threatened ungulates to mitigate the threat of depletion of the tiger's prey has received comparatively less attention.
Across its range the tiger's prey base generally consists of various species of deer, wild pigs and wild cattle (Sunquist et al., Reference Sunquist, Karanth, Sunquist, Seidensticker, Christie and Jackson1999). Dietary studies on tigers in Peninsular Malaysia are lacking but the principal prey probably include large (> 20 kg) ungulates such as wild boar Sus scrofa, barking deer Muntiacus muntjac and sambar Rusa unicolor (DWNP, 2008). In southern Peninsular Malaysia the bearded pig Sus barbatus can be included in this list. However, it is uncertain whether the gaur Bos gaurus and Malayan tapir Tapirus indicus are preyed upon. Given the opportunity tigers may hunt tapirs, as indicated by a few records (Hislop, Reference Hislop1951; McClure, Reference McClure1963). Adult male gaurs, which can weigh up to ten times more than adult tigers, are not ideal tiger prey but calves could conceivably be hunted.
In Malaysia tiger prey populations appear to be declining. Baseline figures for prey populations have been difficult to establish because of the challenges of estimating inherently low densities of ungulates in dense rainforests. Furthermore, the sampling effort and resources (i.e. manpower and funding) required to establish baselines for populations of large mammals in developing countries such as Malaysia often exceed the capacities of both governments and NGOs. Among large mammals in Peninsular Malaysia, scientifically defensible density estimates are only available for tigers (e.g. Kawanishi & Sunquist, Reference Kawanishi and Sunquist2004; Darmaraj & Shariff, Reference Darmaraj and Shariff2009) and Asian elephants Elephas maximus (Hedges et al., Reference Hedges, Gumal and Ng2008; Gumal et al., Reference Gumal, Oziar, Wong, Chin, Ng and Hedges2009).
Poaching has long been regarded as the major threat to ungulates in Peninsular Malaysia (Mohd Khan, Reference Mohd Khan1968; Zaaba et al., Reference Zaaba, Mohd Tajuddin, Mustafa, Ebil and Kiew1991; DWNP, 1992; Misliah & Sahir, Reference Misliah, Sahir and Gaik1997; Abdul Kadir, Reference Abdul Kadir1998) and, along with habitat loss, is probably the main driver of declining tiger prey populations. Despite being considered totally protected or protected game species under the Protection of Wild Life Act 1972, some species can still be legally hunted (Table 1). Based on our recent camera-trapping data, however, certain tiger prey species such as sambar have become harder to detect outside protected areas. According to an interview-based survey to assess perceived changes in the abundance of wild ungulate populations (Goldthorpe & Neo, Reference Goldthorpe and Neo2011) respondents from all survey areas admitted that sambar were hunted illegally throughout the year, despite an 11-month closed hunting season. In discussions with the Malaysian government the regulation of legal hunting has also proven difficult with respect to species such as bearded pig that, together with the more common wild boar, continues to be indiscriminately hunted despite no licences having been issued for the former. Thus the conservation status of tiger prey species requires an evaluation using the best available data (BAD).
Summary of the protected status and information on hunting licences for the six large ungulate species in Peninsular Malaysia in 2008.
In 2008 we carried out an assessment on ‘the current status of the sambar deer, barking deer, bearded pig and wild boar based on existing information to justify the needs for better protection’ (Action 2.2.2 of the National Tiger Action Plan, DWNP, 2008). This assessment (Kawanishi, Reference Kawanishi2008) was timely given increasing reports of poaching in Malaysia's forests (e.g. Murali, Reference Murali2009; TRAFFIC South-east Asia, unpubl. data), and mounting evidence of significant range contractions of ungulates such as the bearded pig (Kawanishi et al., Reference Kawanishi, Richardson and Lazarus2006).
Ideally, policy changes for tiger prey management should be based on sound population data. However, the potential for tiger prey populations to be completely depleted in a relatively short time was seen as sufficient reason to use best available data to help conserve the tiger and associated prey species. There is precedence in adopting such a precautionary approach in conservation: Principle 15 of the Rio Declaration at the 1992 Earth Summit states ‘In order to protect the environment, the precautionary approach shall be widely applied by States according to their capabilities. Where there are threats of serious or irreversible damage, lack of full scientific certainty shall not be used as a reason for postponing cost-effective measures to prevent environmental degradation.’ Since then, the need for applying the precautionary principle has been reiterated in numerous wildlife management scenarios (e.g. Olea et al., Reference Olea, Sanchez-Barbudo, Vinuela, Barja, Mateo-Thomas and Pineiro2009; but see Jepson et al., Reference Jepson, Brickle and Chayadin2001).
Here, we present the synthesis of best available data from camera-trapping studies, to elucidate whether any of the six prey or potential prey species of the tiger (gaur, sambar, barking deer, wild boar, bearded pig and Malayan tapir) warrant improved legal protection in Peninsular Malaysia. We also explain how the results of this assessment were used by conservation NGOs and the government to improve legal protection for tiger prey.
Methods
To make crude comparisons of the abundance and distribution of our focal species, data from long-term and rapid assessment camera-trapping studies conducted at 23 sites in Peninsular Malaysia (Fig. 1) between 1997 and 2008 were compiled from published (Laidlaw et al., Reference Laidlaw, Rahman and Zainal2000; DWNP/DANCED, 2001; Kawanishi & Sunquist, Reference Kawanishi and Sunquist2004; Sharma & Ahmad Zafir, Reference Sharma, Ahmad Zafir and Latiff2005, Reference Sharma, Ahmad Zafir, Sharma, Ahmad Zafir, Nadzirah and Dayang Norwana2007; Ahmad Zafir et al., 2006; Ahmad Zafir & Sharma, Reference Ahmad Zafir, Sharma, Muda, Jaafar, Sabran, Som, Nizam and Latiff2006; Darmaraj, Reference Darmaraj2007; Lynam et al., Reference Lynam, Laidlaw, Wan Shaharuddin, Elagupillay and Bennett2007; Shariff et al., Reference Shariff, Darmaraj, Ahmad Zafir, Murray, Fong, Razani, Koh, Nik Mohd. Shah, Damanhuri and Latiff2010) and unpublished sources (WWF–Malaysia and Wildlife Conservation Society–Malaysia Programme/Johor National Parks Corporation). Most of these studies targeted tigers, with cameras set on active game trails to maximize detection probability, and so biases towards detection of tigers were expected. For the purposes of our assessment, however, detection probabilities were assumed to be constant temporally and spatially for our focal ungulate species. Detection (i.e. locations of confirmed presence) and non-detection (i.e. locality sampled but target species not detected) data were used to generate a distribution map for the six species. Non-detection either means (1) true absence of the species, or (2) false absence (i.e. failure to detect the species when present). Our assessment did not attempt to differentiate between the two, and thus the results were interpreted with caution and emphasis was placed on the pattern of sites with confirmed presence. For comparing relative abundance we assumed that true abundance has a positive and constant relationship with the rate of photographs taken (the detection rate; see Supplementary Information 1 for supporting information).
The 23 camera-trapping study sites (see Supplementary Table S1 for locality names) within protected areas and forest reserves across Peninsular Malaysia.
Results
The combined sampling effort of 23 camera-trapping studies spanned 40,303 trap-nights and gave a total of 10,145 photographs of species of wild mammals. Six of the studies were in protected areas where c. half of the trapping effort (n = 20,651) was spent. The areas sampled ranged from 6.3 to 400 km2 and total trap-nights per site were 120–5,970 nights. We used the collated data to construct a detection/non-detection map of ungulates from sampled sites in Peninsular Malaysia (Fig. 2). The details of our assessment are given in Supplementary Table S1.
Detection/non-detection of six tiger prey species between 1997 and 2008 in 23 camera-trapping study sites across Peninsular Malaysia (see Fig. 1 for numbered locations).
The three most commonly detected ungulate species (in order of increasing mean detection rates) were Malayan tapir, barking deer and wild boar (Fig. 3). However, the relative abundance of wild boar should be considered conservative as it is a gregarious species. These three species were also the most widely detected species based on the number of sites (i.e. 20, 22 and 23 sites, respectively; Fig. 2).
Mean detection rates (number of photographs per 100 trap-nights), based on 23 camera-trapping studies conducted in Peninsular Malaysia between 1997 and 2008 (see Fig. 1 for numbered locations), of six tiger prey species, with their protected status (P, protected; TP, totally protected). The error bars represent one standard deviation.
The three least detected ungulate species (in order of decreasing mean detection rates) were sambar, bearded pig and gaur (Fig. 3). These three species were also detected in the least number of sites (i.e. 8, 2 and 4 sites respectively; Fig. 2). Of 414 photos of sambar, 346 (84%) were from protected areas; this was higher than expected when the proportion of the sampling efforts in protected areas and non-protected areas were accounted for (X 2 = 41.9, df = 1, P < 0.001). This indicates their relative rarity in unprotected forests. Bearded pigs were only detected in the southern half of the Peninsula, whereas detections of gaur were restricted to two protected areas and none were detected in the southern peninsula (Fig. 2).
Discussion
Sambar and bearded pig require better legal protection
Based on the best available data, tiger prey species such as sambar and bearded pig appear to be as threatened as the totally protected gaur, with all three species having low relative abundance across Peninsular Malaysia. There is a paucity of detections and reliable information on sambar from sites outside protected areas. In two prolonged (> 9 months) camera-trapping studies (Ahmad Zafir et al., Reference Ahmad Zafir, Lee, Sharma and Jayasilan2006; Darmaraj, Reference Darmaraj2007) in two forest reserves (also known as selectively logged forests), sambar were not detected despite high sampling effort (7,631 and 2,664 trap-nights, respectively). An ongoing study in the largest forest reserve (1,948 km2), adjacent to the country's largest protected area (Taman Negara), has yet to detect this species (Kawanishi et al., Reference Kawanishi, Aik, John, Gumal and Sukor2011). Even within Taman Negara, sambar were not as abundant or widespread as other ungulates; an estimate of sambar density in a 200 km2 area was c. 0.16 individuals per km2, one order of magnitude lower than estimates of wild boar and barking deer densities (Kawanishi & Sunquist, Reference Kawanishi and Sunquist2004). These conclusions are corroborated by local resource users. An interview-based survey conducted in 18 villages around the country found a general consensus amongst hunters and wildlife dealers (n = 61) that both sambar and barking deer populations are either reduced or locally extinct, with many indicating overhunting as the primary cause (Goldthorpe & Neo, Reference Goldthorpe and Neo2011). It is surprising that the sambar appears to be becoming scarcer than the tapir, despite the latter's greater body mass, energetic needs and hence lower intrinsic rate of increase (Caughley & Krebs, Reference Caughley and Krebs1983). The implication is that external factors such as overhunting are probably responsible. Tapirs are relatively more abundant, probably because they are not highly prized as game animals (Kawanishi et al., Reference Kawanishi, Sunquist and Sahir2002). However, the contraction of the range of the sambar in Peninsular Malaysia cannot be demonstrated conclusively as its historical distribution has never been assessed. Nevertheless, its sparse distribution and low numbers in Malaysia have been documented in the country report for the IUCN Red List (IUCN, 2008).
Our results also corroborate the suspected extirpation of bearded pig populations in the northern half of Peninsular Malaysia. The bearded pig is known to be a migratory species, following mast fruiting of dipterocarp trees. This species used to be fairly abundant and widespread but has suffered severe population declines because of hunting (Caldecott et al., Reference Caldecott, Blouch, MacDonald and Oliver1993) and habitat loss over the past 60 years (Kawanishi et al., Reference Kawanishi, Richardson and Lazarus2006). Prior to the 1950s bearded pigs were widely reported from the northern, central and southern areas of Peninsular Malaysia (i.e. southern Perak, southern and eastern Pahang, Negri Sembilan and Johor; Allen, Reference Allen1948; Kempe, Reference Kempe1948; Hislop, Reference Hislop1949). This species now appears restricted to Endau Rompin State Park and neighbouring forest reserves from southern Pahang to eastern Johor. The impacts of hunting on bearded pigs remain difficult to quantify because of the lack of long-term data but hunters licensed to take wild boar are probably killing them, incidentally or intentionally, year-round.
The discussion and lobbying process
At the time of this study (Kawanishi, Reference Kawanishi2008) the Protection of Wild Life Act 1972 was considered by civil society to be outdated and its penalties were deemed ineffective in deterring illegal hunting (e.g. Damis, Reference Damis2009). The Tiger Action Plan addressed these faults to some degree but there was insufficient justification to effect changes in this law prior to our study. At the end of 2008 we presented our recommendations (Supplementary Information 2) for improved legal and regulatory protection of tiger prey species (except for gaur and tapir as they already had the highest legal protection status) to the local wildlife authority, the Department of Wildlife and National Parks. Subsequently the Department, with technical support from the Wildlife Conservation Society–Malaysia Programme, carried out a second assessment focusing on wildlife inventory and camera-trapping data for Taman Negara, and the results supported our findings. Concurrently, three of the country's main conservation NGOs (TRAFFIC South-east Asia, WWF–Malaysia and the Malaysian Nature Society) launched media campaigns designed to raise public awareness of the need to change the status of the sambar and bearded pig in Malaysia's wildlife legislation. Local and international pressure to amend the laws were also instrumental in the lobbying process and a petition, circulated by the three NGOs and signed by c. 56,000 people globally, was handed to the government.
Policy change and implementation of the Tiger Action Plan
In the face of overwhelming public concern the Department of Wildlife and National Parks initiated a 2-year moratorium in 2009 on the issuance of hunting licences for sambar and barking deer. Following our recommendation for a hunting ban of 5–8 years (Supplementary Information 2) the 2-year moratorium was extended to 6 years, which effectively implemented Action 2.2.1 of the National Tiger Action Plan for Malaysia (DWNP, 2008). Since then, NGOs and various researchers have commenced studies at various sites to establish scientifically defensible baselines for tiger prey species, to facilitate long-term monitoring of population trends and to assess the impact of the moratorium (Action 2.2.3, Tiger Action Plan; DWNP, 2008). The method involves sign surveys within an occupancy framework to account for imperfect detections (McKenzie et al., Reference MacKenzie, Nichols, Lachman, Droege, Royle and Langtimm2002).
Meanwhile, local and international pressure to improve the 30-year old Protection of Wild Life Act 1972 escalated in the millennium's first Year of the Tiger. A new and improved Wildlife Conservation Act 2010 was finally passed by the Cabinet in August 2010 and became fully enforced by December 2010 (Action 2.1.2, Tiger Action Plan; DWNP, 2008). Under this new law the bearded pig was afforded the highest protection status. Subsequently, the government and NGO partners publicized these changes in wildlife legislation and hunting regulation through the media, talks and community outreach programmes (Action 2.1.3 and 2.2.6, Tiger Action Plan; DWNP, 2008).
We have demonstrated that best available data can sometimes be used to improve legal protection of threatened mammals when baselines are absent. Furthermore, we have shown that it was such data that facilitated the implementation of certain actions in the Tiger Action Plan, the best blueprint for tiger conservation in Malaysia. Scientists should, of course, strive to collect data from peer-reviewed, long-term monitoring programmes, to improve conservation policies, but the precautionary principle should be invoked when there is new evidence suggesting the decline of populations of threatened species. Despite the shortcomings of our assessment, a combination of long-term biological information, cooperation, public awareness and determination eventually led to changes in the legal and regulatory framework to improve protection of threatened ungulates and, by association, the country's apex predator, the Malayan tiger. The greater long-term challenge, however, lies in the stringent enforcement of this new and improved legal framework, as well as in monitoring the recovery of tiger prey species that have been granted a reprieve from licensed hunting pressure.
Acknowledgements
We thank three anonymous reviewers, Arun Venkataraman, Loretta Ann Shepherd and Mark Ryan Darmaraj for useful discussions on this subject and for reviewing earlier drafts of this article, and Ahmad Zafir, Mark Rayan Darmaraj and Shariff Wan Mohamad of WWF–Malaysia and the Johor National Parks Corporation for making unpublished data available. The preparation of the original report was made possible with financial support from the National Fish and Wildlife Foundation's Save the Tiger Fund.
Biographical sketches
Kae Kawanishi is a wildlife biologist interested in carnivore ecology and conservation. Reuben Clements worked for WWF–Malaysia at the time of the study and is currently studying the impacts of roads on large mammals and indigenous peoples. Melvin Gumal's research interests include tigers, elephants, orang-utans and effects of land use change on wildlife. Gareth Goldthorpe previously worked for WWF–Malaysia. His research interests include carnivore conservation, human–wildlife conflict and protected area management. Mohd Nawayai Yasak was the Director of the Biodiversity and Conservation Division in the Department of Wildlife and National Parks Peninsular Malaysia at the time of the study. Dionysius S.K. Sharma has worked for over 20 years on a variety of mammals and reptiles in Peninsular Malaysia and Borneo.
