A Demi-Decade of Mammal Research: A Rapid Assessment within the Heart of Borneo in Sabah

Sabah contributes 4.2 million hectares to the total Heart of Borneo (HoB) areas. Some of the forest reserves in the HoB are newly gazetted as Totally Protected Forest. Hence, their mammal diversity has to be comprehensively documented. This study aims to record the presence of terrestrial mammal species, and assess the prevalence of poaching in selected forest reserves within the Sabah HoB area. A total of 15 forest reserves were surveyed within a 5-year timeframe which recorded 60 terrestrial mammal species, including 21 Bornean endemics. The variation in total enumerated mammal species in the study sites may be derived from unequal sampling efforts, geographical factors and anthropogenic influences. The intensity of poaching within the study sites is high. Though this study is a rapid assessment, it created baseline information for mammal diversity in some of the least studied forest reserves in Sabah, important for conservation of its terrestrial mammals.


INTRODUCTION
Protection of Bornean faunal diversity has been a major concern among conservationists (Payne & Davies 2013). Illegal hunting of animals (henceforth referred to as poaching) and deforestation are regarded as main threats to wildlife on the island (Phillipps & Phillipps 2018). The death of Malaysia's last Sumatran rhino delivers a strong signal that the conservation struggle is far from over, and that it is not an easy task -even with active efforts by Government agencies, foreign expertise or help from non-governmental organizations (NGOs). Fortunately, the Malaysian Government through the Sabah State Government's Forestry Department is still willing to invest much of its resources by implementing various strategies to conserve wildlife diversity in the state (SFD 2013).
The Heart of Borneo (HoB) is a conservation initiative encompassing approximately 22 million hectares (ha) of inter-connected rainforest, across three countries on the island of Borneo; Malaysia (Sabah and Sarawak), Indonesia (Kalimantan) and Brunei Darussalam. This initiative is a joint agreement by governments of respective countries to conserve the biodiversity of the said areas, through sustainable management of forests and other sustainable land uses (WWF 2020). The Strategic Plan for Sabah's HoB Initiative (2014)(2015)(2016)(2017)(2018)(2019)(2020) involves prioritising collective scientific studies, and enhancing wildlife protection (SFD 2013).
Sabah's HoB area is approximately 4.2 million ha, comprising at least 95 forest reserves. Overall, there are 333 forest reserves in the state. A number of these forest reserves or conservation areas have received substantial attention from wildlife biologists including Deramakot Forest Reserve (FR) (Samejima et al. 2012), Ulu Segama FR (Bernard et al. 2016), Danum Valley Conservation Area (Hanya et al. 2020) and Kinabalu National Park (Camacho-Sanchez et al. 2019). However, some forests are understudied, especially those that are newly classified as Protected Forests and small as well as isolated forest reserves. Prior to 2016, mammal diversity in some of these reserves has never even been scientifically documented. This highlights the urgency for a collective survey of these sites.
Since Sabah's HoB area comprises many forest reserves, reliable and cost-effective approaches for swiftly surveying mammal diversity are required. Rapid biodiversity assessment is one of the common solutions to meet a wide range of conservation needs (Larsen 2016). In Malaysia alone, mammal surveys have been carried out based on rapid assessment with a sampling period ranging between 7 to 12 days (Zahidin et al. 2016;William-Dee et al. 2019).
This study is part of the HoB Scientific Expedition series that was organised by the Forest Research Centre of Sabah Forestry Department from 2016 to 2020. Our main objective was to rapidly survey the presence of terrestrial mammal (including arboreal species) in selected forest reserves within the Sabah HoB area. We excluded Chiropteran in this study. Indirectly through this study, we were also able to assess the prevalence of poaching activities in each reserve. This study was designed to create baseline information aside from updating the current distribution data of terrestrial mammal species in Sabah, facilitating an effective collaboration between scientists and policy makers in formulating a comprehensive, data-driven forest management plan.

Study Sites
We surveyed 15 forest reserves between 2016 until 2020 ( Fig. 1). All forest reserves classified as Class I Protection Forest except the Agathis FR (Class VI Virgin Jungle Reserve). Both classes are meant for environmental protection, hence no timber extraction allowed. Aside from environmental protection, Virgin Jungle Reserve is also conserved for forestry research purposes. This study focused only on specific sites in each reserve. For example, the study site in Ulu Segama FR was on the upstream of Sungai Juak, whereas in Pensiangan FR, the fieldwork was conducted at Sungai Karilin and its surrounding areas.
Forest ecosystem types in each study site varied between mixed dipterocarp, montane, heath and ultramafic forests. Generally, all 15 forest reserves have been disturbed in the past but patches of old-growth forest are still remnant in the study sites. All sites are accessible, though sometimes different modes of transportation (e.g. boat) were utilised.

Data Collection of Terrestrial Mammal
We recorded the presence of terrestrial mammals using three methods, namely day transect, night spotlight and camera trapping surveys. We conducted day transect survey in all study sites but time spent and distance surveyed varied (Table 1). Both day transect and night spotlight surveys were carried out either along the old-logging roads, asphalt roads, old trails, newly-made trails and rivers. We included the use of a binocular (NIKON Monarch 7) and a white light spotlight in day survey and night survey, respectively. We used a Toyota 4WD Single Cab throughout the night spotlight surveys. We also used a boat to survey mammals along the Serudong river in Sungai Serudong FR. The day survey ranged from 07:30 to 17:30 hours, meanwhile the night spotlight surveys are often executed from 19:30 to 22:00 hours. Both direct (e.g., opportunistic sighting) or indirect (e.g. footprints, presence of dung or nests) sightings were identified by J. Miun and L. Tingkoi, whom have been continuously involved in Bornean wildlife research since 1996. The inclusion of using camera traps was intended to maximise mammal detection rates throughout this study. We deployed a total of 60 camera traps (Bushnell Trophy Cam HD Aggressor No Glow) in the selected study sites (  The distribution of camera traps in the study sites was purposively planned in relation to accessibility, safety and time. The location of two camera traps were kept at a minimum distance of 100 m apart. We standardised the position of the camera trap; secured at the base of a tree; 0.5 m above the ground. These passive infrared-activated camera traps are motion-sensitive in that they are triggered to take three shots every time movement is detected. Time interval between two images was set at 30 sec apart. We did not use baits, and set the camera traps to operate for 24 h on a daily basis. We calculated the camera trap success rate in each study site based on Ancrenaz et al. (2012). We followed the IUCN Red List (IUCN 2021) to determine the taxonomic names of each recorded mammal. We classified the conservation status of recorded species based on the IUCN Red List (IUCN 2021) as well as their local protection status in accordance to Sabah's Wildlife Conservation Enactment 1997 (Amendment 2017) (WCE 1997). In order to assess the prevalence of poaching, we selected eight elements as indicators of the illicit activity: presence of (1) gunshot sounds, (2) used bullet cartridges, (3) shelter or campfires, (4) snares, (5) hunting dogs, (6) game animal carcass, (7) poachers and (8) sale of bushmeat in the vicinity of the forest reserve. We recorded the presence of these elements throughout the rapid survey. We did not employ any interview technique for this study.

RESULTS
More than half of the surveyed forest reserves are situated in the interior and Tawau division of Sabah. Agathis, Pensiangan, Nuluhon Trusmadi (Extension), Tambulanan, Sungai Serudong, Silimpopon, Sungai Tiagau (Extension) and Mengilan forest reserves are at least 50 km away from any major towns. Journey into the study sites usually involved extensive driving and trekking. For instance, we had to walk for approximately 2.5 h to reach the Agathis FR in Tenom district. The other listed reserves are in close proximity to major towns.
Within the 62 days of field campaigns, we managed to survey a sum of 117.44 km long transects ( Table 1). The longest distance surveyed during both day transect and night spotlight methods were in Mengilan FR, while the shortest transect was established in Silimpopon FR. We attained 1,481 trap-nights from 60 camera traps that were deployed in nine study sites (Table 3). Highest mean of camera trap success rate was recorded in Nuluhon Trusmadi (Extension) FR. Duration of the combined survey methods was unequal across the study sites. The extended sampling duration in Mengilan FR is a result of the first Movement Control Order implemented during the COVID-19 pandemic. Throughout the rapid assessment, we recorded a total of 60 terrestrial mammal species belonging to nine orders, 23 families and 43 genera including 21 Bornean endemics (  A total of 43 species were detected through camera trapping (Fig. 2). The day transect survey as well as the night spotlight survey recorded 46 species and 19 species, respectively. There were ten species recorded by all three methods, namely the bearded pig, Asian elephant, leopard cat, common palm civet, greater mousedeer, lesser mousedeer, Malay civet, red muntjac, sambar deer and Western tarsier. We enumerated 36 mammal species in Mengilan FR which is the highest among all study sites (Table 4). Fieldwork in Silimpopon FR resulted in the discovery of six mammal species. On average, rapid assessment in each reserve recorded the presence of 13 species. In most reserves, we registered 14 to 23 species.
None of the surveyed forest reserves were devoid of poaching elements (Table 5; Fig. 3). All eight elements were present in Sungai Tiagau (Extension) and Pensiangan forest reserves. The presence of poachers was confirmed based on their footprints and vehicle tyre tracks. Occasionally, we even spotted the poachers during field campaigns. In Pensiangan FR, poachers were sighted on motorcycles without a license plate number while carrying dead bearded pigs. Other encounters occurred in Bukit Hampuan, Sungai Serudong, Kungkular, Nuluhon Trusmadi (Extension), Mengilan and Tambulanan forest reserves. Another common poaching element was the presence of snares. Most of these snares are designed to catch cervids, pigs, porcupines and pangolins. We also found empty bullet cartridges, abandoned shelters and game animal carcasses in most of the study sites.

Element
Forest Reserve

DISCUSSION
Borneo is a home to 147 terrestrial mammals, not including the order Chiroptera (Phillipps & Phillipps 2018). Therefore, this study has recorded about 40% of the total terrestrial species found on the island. The high number of unrecorded species is because of the nature of our sampling design -rapid assessment. Reporting species presence, many for the first time at certain sites, is important to verify current and historic distribution of mammals in Sabah (Kramer-Schadt et al. 2016). All forest reserves were under-sampled and hence, the list of recorded species from each reserve is not comprehensive. Indeed, more species are to be recorded if sampling efforts were increased. For instance, Ulu Segama and its surrounding areas are known to harbour at least 58 terrestrial mammal species (SFD 2016) as opposed to only 18 species during the rapid survey in the upstream area of Bole River in the same forest reserve.
In Mengilan FR, the highest number of recorded species might have been a result of the longer sampling period. However, we do not exclude other factors including location and impact of human activities. Severe anthropogenic disturbance in areas close to major towns has reduced mammal diversity in Peninsular Malaysia (William-Dee et al. 2019). Even though the duration of survey in Sungai Tiagau (Extension), Sungai Serudong and Tambulanan forest reserves were among the lowest, total recorded mammals from these reserves still ranged between 20 to 23 species. These aforementioned forest reserves are further from Tawau town but are closer to the Malaysia-Indonesia border. Fieldwork in Ulu Kalang FR which is just 3 km away from Tenom town yielded only 10 species.
Forest connectivity also led to the variation in total enumerated mammals across study sites. Fieldwork in Silimpopon FR focused only on its second block (477.37 ha). This block is completely isolated and is surrounded by oil palm plantation which may have contributed to its low species richness. Our finding corroborates with a study conducted in Lahad Datu where fragmented forests recorded low mammal species richness (Bernard et al. 2014). The majority of the study sites are connected to other forest reserves of different classes. Ulu Segama FR and Mengilan FR formed contiguous forest covers with Danum Valley Conservation Area as well as the vast North Kalimantan rainforest, respectively.
Other factors that influenced the detection of mammal species were potentially altitude and forest quality. It is reported that small mammal diversity decreases from low altitudes on both Mount Tambuyukon and Mount Kinabalu (Camacho-Sanchez et al. 2019). Similarly, our fieldwork in the montane environment (1040 m-1650 m) of Tenompok FR recorded 10 mammal species. At a much lower altitude, we detected seven species from Sungai Rayoh FR. It can be explained due to the recurrence of fire incidents in this reserve which formed the current extensive secondary vegetation subsequently impacting mammal richness. Detrimental impacts of forest fire on mammal population have also been reported in other parts of Borneo (Rijksen & Meijaard 1999).
Since the sampling efforts through camera trapping, day transects and night spotlight surveys were different across study sites, we could not confidently determine their sampling efficacies. However, wildlife ecologists favour camera trapping due to its effectiveness (Wearn & Glover-Kapfer 2019). We frequently detected more mammal species via day transect and night spotlight surveys. In some cases, our camera traps either malfunctioned or were stolen. Despite that, the combination of these three methods ultimately increased the mammal detection rates of this study.
Undeniably, mammal communities in Sabah are pressured by illegal hunting. Signs of poaching activities have become the norm throughout our field campaigns from 2016 to 2020. Illegal hunting of wildlife has also been reported outside the surveyed forest reserves (SFD 2016). In Sabah, poaching has been associated with indigenous culture (Wong et al. 2012), traditional medicine (Gomez et al. 2020), trade (Pantel & Anak 2010) and bushmeat (Kurz et al. 2020). Our encounters with poachers have so far never resulted in any undesirable situation. Many of the encounters happened while we were in a vehicle, travelling back and forth to the study sites. We however, tried to discourage some of them from proceeding with their illicit activities.
We reported the presence of poaching elements in just 15 out of 333 forest reserves. This leaves us to ponder -do we have enough manpower to deter poaching in Sabah? Sabah Forestry Department has implemented a number of initiatives to tackle the poaching issue. In 2016, SFD Protect Team was launched to strengthen the department's capability in combating forest crimes. In order to increase manpower capacity, local communities were encouraged to help through the Honorary Forest Rangers programme. Fortunately, various NGOs are also partnering with the Government to be actively involved in mitigating illegal harvest of forest resources in Sabah. For instance, Sabah Environmental Trust has been closely involved in patrolling Danum Valley, Maliau Basin and Imbak Canyon conservation areas since 2017 (SFD 2019). These efforts are bearing fruit because 860 cases of forest crimes were reported between 2016 to 2020 which resulted in the arrest of 433 individuals and 154 of them were convicted (SFD 2020).
In addition to that, the Sabah Forestry Department has continuously worked on instilling public conservation awareness. A series of environmental education programmes have been carried out by the department through its Rainforest Discovery Centre since the implementation of the HoB initiative. Such programmes create opportunities for local communities to support government efforts in conservation (Arpa et al. 2016). Because of the existence of local settlements nearby the study sites, we highly recommend recruiting the participation of these communities in the form of citizen science. In Sabah, citizen scientists have positively contributed to conservation efforts when scientific knowledge was expanded to them (Freitag et al. 2018;Araujo et al. 2020). Lastly, conservation programmes should also prioritise the improvement of forest connectivity of fragmented reserves. If left unchecked, it could possibly lead to the loss as a functionally connected habitat (Ocampo-Peñuela et al. 2020).

CONCLUSION
This study documented the presence of 60 species across 15 study sites, which indirectly updated the distribution of terrestrial mammal species in Sabah. We concluded that the actual number of species in each forest reserve is actually way higher, and hence, this study should not be used to confirm the absence of any terrestrial mammal species. This study also reiterated the rampant poaching issues in this state. In future, fieldwork should be carried out at a larger scale with focus on the understudied sites. In the long run, mammal rapid assessment could trigger a more concrete and targeted conservation program that emphasise protection of threatened species and ecosystems.

APPENDIX Appendix A
List of terrestrial species recorded throughout the mammal rapid assessment between 2016 until 2020.