Introduction
The mountains of south-west China, including the Hengduan, Nujiang Langcang Gorge and Qionglai-Minshan Mountains, are recognized as a global biodiversity hotspot (Myers et al., Reference Myers, Mittermeier, Mittermeier, da Fonseca and Kent2000; Mittermeier et al., Reference Mittermeier, Gil, Hoffmann, Pilgrim, Brooks and Mittermeier2004). The eastern margin of the Tibetan Plateau forms part of the Hengduan Mountains, and is a transitional zone between the subalpine conifer forests of the Hengduan Mountains and the alpine scrublands and meadows of the Tibetan Plateau. The region features a complex topography and a variety of vegetation types, including mixed conifer and broad-leaved forests, subalpine conifer forests, alpine scrublands, and meadows and tundra. It is believed to be rich in biodiversity, owing to species differentiation caused by the orographic uplift of the Tibetan Plateau (Zhang et al., Reference Zhang, Cheng, Li and Yao2002). However, human activities are threatening the biodiversity in the region. Although commercial logging and hunting have been banned following China's Natural Forest Conservation Programme in 1998, forests are still the primary source of wood for household use (e.g. for construction, and fuel for cooking and heating), and many areas in the region are intensively used for grazing and the gathering of food and traditional medicines (Brandt et al., Reference Brandt, Kuemmerle, Li, Ren, Zhu and Radeloff2012). In addition, despite the remoteness of these areas, rapid economic development and human population growth have resulted in an increasing demand for infrastructure development and a rapid expansion of extractive (e.g. hydropower and mining) and tourism industries (Brandt et al., Reference Brandt, Kuemmerle, Li, Ren, Zhu and Radeloff2012; Xinhua, 2018). These activities accelerate habitat loss and fragmentation, and place biodiversity under increasing threat throughout the region (Brandt et al., Reference Brandt, Kuemmerle, Li, Ren, Zhu and Radeloff2012; Zhao & Lu, Reference Zhao and Lu2017). The conservation of the region's unique biodiversity requires dedicated efforts in research, conservation planning and implementation.
Large and medium-sized mammals, particularly carnivores such as felids, are key indicators for biodiversity conservation and management (Dickman et al., Reference Dickman, Hinks, Macdonald, Burnham and Macdonald2015; Xiao et al., Reference Xiao, Li, Xiang, Li, Jiang and Zhang2017). Confirmed presence of large and/or medium-sized mammal species provides insights into the community composition and structure, and is an indicator of ecosystem health (Hedwig et al., Reference Hedwig, Kienast, Bonnet, Curran, Courage and Boesch2017). Mammals on the eastern edge of the Tibetan Plateau have been studied since the mid 19th century, when European and American explorers, missionaries and expeditions entered the region to collect specimens (Shi & Zhao, Reference Shi and Zhao1982). In the mid 20th century, Chinese researchers began to investigate mammals in the area (Shi & Zhao, Reference Shi and Zhao1982). In particular, after the founding of the People's Republic of China in 1949, surveys were conducted in the region (Wu et al., Reference Wu, Yu and Peng1989; Peng et al., Reference Peng, Zhou and Liu2006; Shen, Reference Shen2008), including two national surveys on terrestrial wildlife (State Forestry Administration, 2009; SCOL, 2013). Nevertheless, much of the region has yet to be surveyed and knowledge of the diversity of large and medium-sized mammals remains limited (Sichuan Observer, 2017; Chinanews, 2018; Hu et al., Reference Hu, Li, Luo, Zewengtaxu, Chen and Zhang2019; People.cn, 2019). Photographic records of mammals are particularly scarce (Xiao et al., Reference Xiao, Li, Xiang, Li, Jiang and Zhang2017). Other ecoregions in the hotspot, such as Qionglai-Minshan conifer forests, have been studied more thoroughly and are known to harbour flagship species for conservation, including the giant panda Ailuropoda melanoleuca and the golden monkey Cercopithecus kandti.
Here, we conducted camera-trap surveys of large and medium-sized mammals on the eastern edge of the Tibetan Plateau. Camera traps have become a valuable tool for wildlife surveys (Ancrenaz et al., Reference Ancrenaz, Hearn, Ross, Rahel and Wilting2012; Burton et al., Reference Burton, Neilson, Moreira, Ladle, Steenweg and Fisher2015) and are effective in monitoring rare or elusive animals such as felids (Carbone et al., Reference Carbone, Christie, Conforti, Coulson, Franklin and Ginsberg2001). Camera trapping can form the basis of effective long-term monitoring of wildlife and can help document population changes (Li et al., Reference Li, Wang, Xiao, Li, Wang, Feng and Wang2014; Burton et al., Reference Burton, Neilson, Moreira, Ladle, Steenweg and Fisher2015). Our objectives were to compile a species inventory and provide an initial assessment of the relative abundance of species, to ascertain the potential importance of this area for conservation.
Study area
We conducted the study in Xionglongxi Provincial Nature Reserve and adjacent areas, in Xinlong County, Garzê Tibetan Autonomous Prefecture, Sichuan Province, China. The area is located c. 600 km from Chengdu, the capital of Sichuan Province (Fig. 1a) and covers 1,872 km2 (582 km2 in the Reserve and 1,290 km2 in adjacent areas). It is characterized by complex topography around the mid Yalong River valleys. A monsoon-influenced humid continental climate and subarctic climate predominate, with cold and dry winters, and warm summers with frequent rain (Zhu & Xiang, Reference Zhu and Xiang2015). The vegetation comprises meadows or tundra (41.7%), conifer forests (27.8%), scrubs (28.3%) and broadleaf forests (2.3%; National Forestry and Grassland Data Center, Sichuan Subcenter, 2020). Potentially harmful human activities include timber extraction for household use (e.g. for construction, and fuel for cooking and heating), illegal hunting, livestock grazing, and the gathering of food and traditional medicines. In particular, with the implementation of the Natural Forest Conservation Programme, grazing may have become the most prevalent human disturbance in the area, as a result of incentives from local governments to develop alternative livelihoods to reduce timber harvesting.
(a) The study area in Xionglongxi Provincial Nature Reserve and adjacent areas, Xinlong County, Sichuan Province, China. The inset map illustrates the previously known range of the Chinese mountain cat Felis bieti. (b) Locations of camera traps. The locations of two camera traps are not shown because of errors in the recording of longitude and latitude data.
Methods
Camera trapping
We conducted camera-trap surveys during September 2016–May 2017. As the purpose of the study was to provide an initial species inventory, we prioritized capturing a high diversity of species over enabling statistical inference on populations (e.g. occupancy modelling). We installed camera traps (Ltl Acorn 6210/6310 series, Zhuhai Ltl Acorn Electronics Co., Zhuhai, China, and Loreda L510 series, Shenzhen Loreda Electronics Co., Shenzhen, China) in areas where high levels of wildlife activity had been reported by staff of the Reserve and the Forestry Bureau. In total, we installed 149 camera traps, placed on animal trails and ridge tops or near waterholes, at 3,100–4,860 m altitude. Of the 149 camera traps, 10.2% were placed in broadleaf forests, 44.2% in conifer forests, 21.8% in scrubs and 23.8% in meadows or tundra (Fig. 1b). The mean distance between cameras was 2.37 ± SD 1.33 km. The cameras used infrared lighting and operated continuously, with motion sensors set to trigger immediately and to take a sequence of three photographs when movement was detected. The camera traps did not use white flash that may have a greater risk of startling animals. We collected the photographic data every 3–4 months.
Data analysis
We defined large and medium-sized mammals as any mammal species with a mean body mass > 0.5 kg (Bruce et al., Reference Bruce, Amin, Wacher, Fankem, Ndjassi and Ngo Bata2018). We identified species based on Wang & Hu (Reference Wang and Hu1999), Peng & Zhong (Reference Peng and Zhong2005), Smith & Xie (Reference Smith and Xie2009) and Jiang et al. (Reference Jiang, Ma, Wu, Wang, Feng and Zhou2015, Reference Jiang, Liu, Wu, Jiang and Zhou2017). We assessed species’ conservation significance on a global and national level according to the threat categories assigned in the IUCN Red List (IUCN, 2019) and the Red List of China's Vertebrates (Jiang et al., Reference Jiang, Jiang, Wang, Zhang, Zhang and Li2016), respectively.
We calculated survey effort as the total number of camera-trap days. We considered a single capture event as any sequence for a given species taken > 30 minutes after the previous sequence of that species at the same location (O'Brien et al., Reference O'Brien, Kinnaird and Wibisono2003; Moo et al., Reference Moo, Froese and Gray2017). We calculated the total number of independent capture events of individual species (e.g. Moo et al., Reference Moo, Froese and Gray2017) and the per cent of camera-trap stations at which each species was detected (e.g. Rovero et al., Reference Rovero, Martin, Rosa, Ahumada and Spitale2014; Moo et al., Reference Moo, Froese and Gray2017). We did not calculate relative abundance indices suggested by Carbone et al. (Reference Carbone, Christie, Conforti, Coulson, Franklin and Ginsberg2001) and O'Brien et al. (Reference O'Brien, Kinnaird and Wibisono2003), because comparisons using this technique are not statistically valid given significant variation in detectability between species and areas (Sollmann et al., Reference Sollmann, Mohamed, Samejima and Wilting2013; Burton et al., Reference Burton, Neilson, Moreira, Ladle, Steenweg and Fisher2015; Anile & Devillard, Reference Anile and Devillard2016; Li et al., Reference Li, Wang, Bu, Liu and Jin2016).
Results
Of the 149 camera-trap locations, 27 failed to generate usable data because of camera malfunction or loss. Over a total of 15,455 camera-trap days, we obtained 5,752 independent captures and identified 29 species of large and medium-sized mammals belonging to five orders and 12 families. Of the 29 species detected, two are categorized as Endangered on the IUCN Red List, six as Vulnerable, and five as Near Threatened; four are categorized as Critically Endangered on the Red List of China's Vertebrates, five as Endangered, six as Vulnerable, and seven as Near Threatened (Table 1).
The large and medium-sized mammals camera-trapped in Xionglongxi Provincial Nature Reserve and adjacent areas, Xinlong County, Sichuan Province (Fig. 1), China, during September 2016–May 2017, with their status on the IUCN Red List (IUCN, 2019) and the Red List of China's Vertebrates (Jiang et al., Reference Jiang, Jiang, Wang, Zhang, Zhang and Li2016), previously known elevational limits, elevational range observed in the study area, per cent of camera-trap stations at which each species was detected, and the total number of independent capture events of each species in a total of 15,455 camera-trap days.
1 According to IUCN (2019).
2 B, broadleaf forests; C, conifer forests; S, scrubs; MT, meadows or tundra.
3 According to Chakraborty et al. (Reference Chakraborty, Nahmo, Upadhyay and Medhi2017).
Thirteen species occurred in all four habitats (broadleaf forest, conifer forest, scrubs, and meadow or tundra) over different elevations, and 16 occurred only in some habitats (Table 1). Of the latter, six species (the snow leopard Panthera uncia, Pallas's cat Otocolobus manul, the blue sheep Pseudois nayaur, the wapiti Cervus canadensis, the Himalayan marmot Marmota himalayana and the Altai weasel Mustela altaica) were recorded only in alpine habitat above 4,000 m (Table 1). We recorded new upper elevation limits for eight species: the Asiatic golden cat Catopuma temminckii (4,241 m), the grey wolf Canis lupus (4,843 m), the red fox Vulpes vulpes (4,865 m), the northern hog badger Arctonyx albogularis (4,824 m), the sambar Rusa unicolor (4,208 m), the forest musk deer Moschus berezovskii (4,179 m), the Malayan porcupine Hystrix brachyura (4,170 m) and the rhesus monkey Macaca mulatta (4,375 m).
The order with the largest number of species recorded was carnivores (16), followed by ungulates (9). Within carnivores, felids were the most diverse, with seven species recorded. The per cent of camera traps that detected a species, and the total number of independent capture events, were highest for two canines, the grey wolf and the red fox, followed by two felids, the leopard Panthera pardus and the leopard cat Prionailurus bengalensis, and one mustelid, the yellow-throated marten Martes flavigula (Table 1). Among the herbivorous species, the per cent of camera traps that detected a species was highest for the Chinese serow Capricornis milneedwardsii, but the woolly hare Lepus oiostolus had the highest number of independent capture events. Other herbivorous species that were detected by a high per cent of camera traps or had a high number of independent capture events were ungulates such as the sambar, the tufted deer Elaphodus cephalophus and the blue sheep (Table 1).
Discussion
Using camera trapping, we surveyed large and medium-sized mammals in Xionglongxi Provincial Nature Reserve and adjacent areas, on the eastern margin of the Tibetan Plateau. Our record of 29 large and medium-sized mammal species, including 13 globally threatened species and 22 species nationally threatened in China, is higher compared to other areas on the eastern edge of the Tibetan Plateau, highlighting the significant conservation value of this area. We recorded diverse carnivore species; of particular significance was our detection of seven felid species (58.3% of the felid species occurring in China). This is the highest documented presence of felid species at the county level in China (Smith & Xie, Reference Smith and Xie2009) and is high also compared with other areas of the world (Sunquist & Sunquist, Reference Sunquist and Sunquist2017), indicating a globally significant felid community.
Our detection of two cryptic, nocturnal species, the Chinese mountain cat Felis bieti and the Asian badger Meles leucurus, marks the first record of both species on the eastern edge of the Tibetan Plateau. The record of the Chinese mountain cat is particularly important, because it extends the previously known range of this species by c. 400 km to the south, from the north-eastern edge to the eastern edge of the Tibetan Plateau (Fig. 1; Riordan et al., Reference Riordan, Sanderson, Bao, Abdukadir and Shi2015). We recorded the Asiatic golden cat, the grey wolf, the sambar, the forest musk deer and the Malayan porcupine in alpine areas above 4,000 m, where they had not previously been observed (IUCN, 2019). In addition, we detected the northern hog badger, for which the previously recorded upper elevation limit was 4,300 m (Helgen & Chan, Reference Helgen and Chan2016), in tundra habitat at 4,824 m. The high-elevation records of these species suggest their versatility and adaptability across diverse habitats. Climate change may also push these species to higher elevations.
Leopards commonly live below the tree line (Roberts, Reference Roberts1977; Green, Reference Green1987), although they have been recorded at altitudes of up to 5,200 m in the Himalayas (Stein et al., Reference Stein, Athreya, Gerngross, Balme, Henschel and Karanth2019). In our survey, we obtained independent captures of leopards in rocky, mountainous areas at 4,200–4,400 m, where snow leopards are likely to occur (McCarthy et al., Reference McCarthy, Mallon, Jackson, Zahler and McCarthy2017), suggesting that the two leopard species share territory. However, we did not record snow leopards below 4,700 m. Competition may arise when the two leopard species occur sympatrically, because they are approximately the same size and have comparable food requirements (Lovari et al., Reference Lovari, Ventimiglia and Minder2013; Shao et al., Reference Shao, Song, Huang, Li and Yao2019). It is possible that, because of climate change, the upper forest limit has shifted to higher altitudes, where both leopard species occur. Consequently, the cold-adapted snow leopards could be confined to more barren, rocky habitats at higher altitudes as the more ecologically flexible leopards expand their territory (Lovari et al., Reference Lovari, Ventimiglia and Minder2013; Buzzard et al., Reference Buzzard, Li and Bleisch2017).
We did not record some of the species that, according to the literature (Wang & Hu, Reference Wang and Hu1999; Peng et al., Reference Peng, Zhou and Liu2006), occur in the study area, including the Tibetan fox Vulpes ferrilata, the dhole Cuon alpinus, the beech marten Martes foina, the Tibetan gazelle Procapra picticaudata, the argali Ovis ammon and Milne-Edwards' macaque Macaca thibetana. The Tibetan fox and the Tibetan gazelle prefer open grassland habitats. Given that our cameras were deployed in only some isolated patches of meadows or tundra around forests and scrubs (Fig. 1b), we were unable to confirm whether these two species occur in the study area. Local people reported that they had observed beech martens and argalis recently, but had not seen any dholes, or even traces of them, for many years. They also did not mention observing Milne-Edwards' macaque. Further camera trapping that covers a wider elevational range and a variety of microhabitats is needed to ascertain the presence of these species in the area.
The patterns for the per cent of camera traps that detected a species and the total number of independent capture events were similar for most species. When a species occurred at more locations, the capture rate was generally higher. The grey wolf and the leopard were the most dominant large predators. The brown bear Ursus arctos, however, although widespread across much of northern Eurasia (McLellan et al., Reference McLellan, Proctor, Huber and Michel2017), was detected relatively rarely, even compared to some threatened species such as the leopard. There are four subspecies of brown bear in China, and the one detected in our study is the Tibetan subspecies Ursus arctos pruinosus, which is endemic to the Tibetan Plateau (Ma et al., Reference Ma, Xu and Hu1998). This subspecies was previously estimated to have a small population (5,000–6,000 individuals; Ma et al., Reference Ma, Xu and Hu1998), but its current population status is unknown (McLellan et al., Reference McLellan, Proctor, Huber and Michel2017). Further studies are needed to examine the distribution and population density of this subspecies at a large scale.
The woolly hare and ungulates such as the sambar, the tufted deer, the Chinese serow and the blue sheep were the main prey species recorded in the study area. The woolly hare was the most dominant and is probably integral to the high-mountain food chains, consuming considerable quantities of plant matter and providing food for large carnivores such as the grey wolf and the leopard (Werhahn et al., Reference Werhahn, Kusi, Li, Chen, Zhi and Lázaro Martín2019). The prey base appears adequate to support large carnivores, but local people reported frequent livestock depredation by large predators such as the grey wolf and the leopard. Given the potential implications of negative human–wildlife interactions (Graham et al., Reference Graham, Beckerman and Thirgood2005; Xu et al., Reference Xu, Yang and Dou2015), we recommend that further studies are conducted in the area, to determine the level of livestock depredation and examine any conflicts.
Conclusion
Our survey provides important evidence to support conservation efforts in Xionglongxi Provincial Nature Reserve and adjacent areas. Further camera trapping should cover a wider elevational range and a variety of microhabitats, to increase the probability of detecting species that we did not record but that may occur in the area. Given that leopards may have expanded their range into that of snow leopards, we also recommend that interspecific relationships between these two species are examined further. Livestock depredation by leopards should be evaluated, and strategies for compensating herders for losses should be developed, to support the conservation of this species. In addition, we recommend that the protection status of the area should be upgraded. Xionglongxi is currently a provincial nature reserve aiming mainly to protect wetland ecosystems; we propose that a larger and national-level nature reserve should be established in the region of Xionglongxi Provincial Nature Reserve, to protect subalpine and alpine ecosystems, with a focus on felids of conservation concern and with the threatened leopard and snow leopard as flagship species.
Acknowledgements
This work was supported logistically and financially by Sichuan Station of Wildlife Survey and Management, Chengdu Leique Eco-environmental Protection Technology Co., Ltd. and Xinlong Forestry Bureau. Additional financial support was provided by the Joint Fund of the National Natural Science Foundation of China and the Karst Science Research Center of Guizhou Province (Grant No. U1812401), and the Doctoral Foundation of Guizhou Normal University (Grant No. 2016). We thank Yang Xiang and Li Zhou from Xinlong Forestry Bureau, and Dunwu Qi from Chengdu Research Base of Giant Panda Breeding, for support during fieldwork.
Author contributions
Study design: BY, YX, XG; fieldwork: DS, BY; data analysis, writing: YX, BW, DS, BY.
Conflicts of interest
None.
Ethical standards
This research was approved by Sichuan Station of Wildlife Survey and Management, and abided by the Oryx guidelines on ethical standards.
Open access
