Introduction
François' langur Trachypithecus francoisi, categorized as Endangered on the IUCN Red List (Bleisch et al., Reference Bleisch, Manh Ha, Khat Quyet and Yongcheng2008), is geographically widespread, ranging from the Red River in northern Vietnam to south-central China (Fooden, Reference Fooden1996; Groves, Reference Groves2001; Zhang et al., Reference Zhang, Chen, Xu and Chris2002). Wild populations of this species occur in fragmented forest habitats scattered across its range (Hu, Reference Hu2007). In China the available data indicate that François’ langurs were historically reported from 27 sites in 23 counties in Guangxi, 12 sites (nine counties) in Guizhou and two sites (three counties) in Chongqing, with a total population of c. 3,500 in the 1990s (Wu et al., Reference Wu, Wei and He1987; Li et al., Reference Li, Lei and Luo1994; Tang & Zhang, Reference Tang and Zhang1994, Reference Tang and Zhang1998; Liu & Wei, Reference Liu, Wei, Xia and Zhang1995; Su et al., Reference Su, Lin and Ma2002).
During the past 2 decades a dramatic population decline, attributed largely to hunting, has been documented in Guangxi (Hu & Wei, Reference Hu and Wei2002; Hu et al., Reference Hu, Dong, Wei, Zhu and Duan2004) and North Vietnam (Nadler et al., Reference Nadler, Momberg, Nguyan and Lormee2003), with the population estimated at c. 300 each in Guangxi (Li et al., Reference Li, Huang, Ding, Tang and Wood2007) and North Vietnam (Nadler et al., Reference Nadler, Momberg, Nguyan and Lormee2003). In Guizhou, however, hunting pressure is generally slight and the population increased from 1,000 in the 1990s (Li, Reference Li, Xia and Zhang1995) to 1,160–1,200 in 2010, although the species has been extirpated from several sites and is now restricted to just five isolated fragments because of extensive human disturbance (Hu et al., Reference Hu, Dong, Luo, Su, Li and Zhou2011).
There is little information available on the population size and distribution of François' langur in Chongqing. There are two reports based on surveys conducted in the 1990s that documented c. 80 individuals in seven groups in Jinfoshan (Tang & Zhang, Reference Tang and Zhang1994, Reference Tang and Zhang1998) and 40–45 individuals in four groups in Furongjiang (Su et al., Reference Su, Lin and Ma2002). The aim of the survey reported here was to assess the current distribution and population of, and the threats to, François’ langurs in Chongqing, to provide the basis for a management plan.
Methods
We reviewed relevant literature, including published articles and unpublished reports, to obtain historical information on the distribution and population of François' langur. We then conducted interviews with staff of local forestry agencies and reserves, hunters and village elders at sites where François' langur was known to occur historically. In the interviews we gathered information about the latest date of direct observations and data on locations and group sizes. We randomly interviewed other villagers and staff to reaffirm the findings. Based on this interview information we chose six sites where langurs were recorded in the 1990s to conduct a 1-month preliminary survey (5 days at each site). During the preliminary survey we found no langurs or traces of fresh faeces at three sites (Simianshan, Wanzu and Gongtan) and so excluded these from more detailed surveys.
The three further sites (Jinfoshan, Furongjiang and Heishangu) covered four counties in south and south-east Chongqing (Fig. 1). These areas have a subtropical monsoon climate (Han & Hu, Reference Han and Hu2010; Table 1). The landscape is dominated by limestone mountains with steep valleys and gentle limestone plateaus at 200–2,250 m altitude. Vegetation composition varies depending on the degree of human impact but the main vegetation types are subtropical evergreen broad-leaf forest at low altitudes and mixed deciduous–coniferous forest at higher altitudes (Tang & Zhang, Reference Tang and Zhang1998; Su et al., Reference Su, Lin and Ma2002).
Distribution of François' langur in southern Chongqing and north-eastern Guizhou. The shading on the inset indicates the location of the main map in China.
Mean annual temperature and total rainfall of the three survey sites in Chongqing (Fig. 1), with area surveyed and area of occurrence of François' langur Trachypithecus francoisi, and the size of populations and groups and number of survey days, during the three surveys from September 2009 to September 2010.
Systematic surveys using transect-based techniques are not suited to the limestone mountain topography of the François' langur's habitat. We therefore used the methods employed by Hu et al. (Reference Hu, Wei and Li1998, Reference Hu, Dong, Wei, Zhu and Duan2004) and Li et al. (Reference Li, Huang, Ding, Tang and Wood2007) for surveys of langurs in Guangxi. We divided the areas to be surveyed into small units of 1–2 km2. Each was surveyed continuously for 2–3 days by teams of 3–4 people. If no langurs were found the surveyors searched the cliffs for fresh faecal traces (with mucilage or traces with lustre) left by langurs.
When langurs were found we recorded the time of detection and either noted the location with a global positioning system or marked the locality on 1 : 10,000 maps. We followed the groups for as long as possible to record group size, sex and age composition. We identified age and sex classes based on fur colour, body size and the colour of the sexual skin at the perineum (the female has a white perineum and the male a black one), which can easily be seen when a langur is sitting facing the observer or if it is running (Hu, Reference Hu2007). Whilst following the groups we recorded the duration of our observations and data on habitat type, feeding patches and sleeping sites.
The reliability of individual counts in forest depends on the frequency of repeat surveys. From September 2009 to September 2010 we surveyed the three areas during three periods totalling 219 days (Table 1). In each period we surveyed each unit at least twice, and three times for units with langur sightings or records of fresh faeces. Surveys normally started at 06.00 and finished at 19.00.
An observation of one langur group on a given day was considered a record. We discarded unclear records (i.e. if the number of individuals was not precisely recorded) and combined duplicated records (records of one langur group by two adjacent survey teams). During each period records of the same langur groups at different times were used for comparison or for revising the data.
Information on hunting, and habitat degradation (cultivated land size, firewood use, tourism facilities and the number of tourists) were collected from records of local governmental agencies and our interviews with 146 local families. When we gathered information on hunting we noted the hunted species, date and location. For information on habitat degradation we recorded the area of cultivated land and amount of firewood used and we collected information on the human population and number of tourists from local governmental agencies.
Results
Distribution and population
In the 219 days of surveys we compiled a total of 145 records, of which 13 were unclear and 52 were duplicates, giving 80 unique records of groups and individuals. The detailed results for each of the three survey periods are presented in Table 1. In Jinfoshan we surveyed c. 150 km2 but found langurs only in the core area in the north-east, where most of the vegetation was evergreen broad-leaf forest with some mixed deciduous/coniferous forest. In Furongjiang langurs were found only in the valley bottom, where fragmented broad-leaf forest patches remain along the banks of the river. We found only four langur groups, one in each of four forest patches, isolated from each other by inaccessible cliffs, waterfalls and cultivated land. In Heishangu we found only two groups, in two remnant broad-leaf forest patches in the valley of the Liyu River. Densities were 3.0–12.0 individuals per km2.
Threats
Hunting of François’ langur in these three sites is low, although hunting for other animals (e.g. wild boar, Chinese muntjac Muntiacus muntiak, clouded leopard Neofelis bebulosa and leopard cat Felis bengalensis) does take place occasionally. The main threat to the langurs is habitat loss and degradation in the form of firewood collection, agricultural encroachment and the construction of tourism facilities and hydroelectric power projects, whose effects varied from site to site.
Jinfoshan is a national reserve with many scenic sites; it attracts >1 million tourists per year. Over the past decade in this reserve there has been a rapid increase in construction of tourism infrastructure and facilities such as highways, hotels, telephone lines and hydroelectric power installations. Most of this infrastructure was built in an area of natural forest, thus misappropriating or destroying actual or potential habitat. In addition, a new tourism zone is planned in the north-east (L.J. Ma, pers. comm.) in the remains of the primary broad-leaf forest (Plate 1) that is the habitat of most of the langur groups. There are c. 8,600 local residents living in villages within or adjacent to the current area of occurrence of the langurs, and the villagers depend largely on the natural forest for wood for cooking and heating and for livestock grazing.
Primary broad-leaf forest in the north-eastern core of Jinfoshan national reserve, Chongqing (Fig. 1).
In Furongjiang firewood collection is a continued source of habitat loss. The majority (97%) of local families collect firewood for cooking and heating; an approximate estimate of consumption is 1,000 m3 per year for all families combined. Almost all suitable flat land and plateaus are cultivated, which restricts the langurs to the forest patches remaining along the two banks of the river. In 2000, when the Jiangkou Hydroelectric Dam was commissioned, the reservoir behind the dam submerged about half of the riverine broad-leaf forests (Plate 2).
The riverine broad-leaf forests remaining along the Wujiang river in Furongjiang, Chongqing (Fig. 1).
In Heishangu primary broad-leaf forest remains only in the north-eastern valley of the Liyu River. Natural forests in other areas of Heishangu have been replaced by forest plantations or farmland (Plate 3) and tourism facilities. The c. 1,100 local families depend mainly on the forest for fuelwood for cooking and heating.
Cultivated land in Heishangu, Chongqing (Fig. 1).
Discussion
Information from local chronicles indicates that François’ langur was formerly widespread in the south and south-east of Chongqing, in seven counties (Jiangjing, Nanchuan, Wansheng, Wulong, Pengshui, Youyang and Xiushan), and our interview data indicated that the species was commonly found in a few sites along the Wujiang River even in the 1980s. Our survey indicates, however, that the species is now restricted to three isolated sites, with a total population of c. 200 individuals. However, underestimation is a problem that plagues estimates of karst-dwelling primates. Considering the broad historical distribution range, the dissected, steep terrain, the shortness of the preliminary survey and the sites chosen for intensive surveys, it is possible that we could have overlooked some langur groups. The data reported here is therefore a minimum estimate of the population and range of the species in Chongqing, and further surveys in other potential sites are needed.
Drastic declines in primate populations in the 20th century were attributable to the effects of the expanding human population and economic development (Strier, Reference Strier2000), and the two main threats to primates are hunting and habitat degradation (Marsh et al., Reference Marsh, Johns, Ayres, Marsh and Mittermeier1987). Although hunting has been considered the main cause of the decline of François' langur in Guangxi (Hu et al., Reference Hu, Dong, Wei, Zhu and Duan2004; Li et al., Reference Li, Huang, Ding, Tang and Wood2007), where langurs have been used to make a medicinal wine (Huang et al., Reference Huang, Wei, Li, Quan and Li2002; Li et al., Reference Li, Huang, Ding, Tang and Wood2007), and North Vietnam (Nadler et al., Reference Nadler, Momberg, Nguyan and Lormee2003), it is less of a problem in Chongqing because there is no tradition of hunting langurs for commercial purposes. From 1989 to 2009 there were only four cases of poaching in Jinfoshan Reserve (Jinfoshan Reserve Management Bureau, 2010, unpubl. data) and two cases in Furongjiang (Pengshui Police Bureau, 2009, unpubl. data) but none of these were poaching of langurs.
Struhsaker (Reference Struhsaker2005) suggested that, besides hunting, the major threats to red colobus monkeys Piliocolobus tephrosceles are habitat loss, degradation and fragmentation. Habitat degradation is severe and common throughout the range of François' langur (Hu et al., Reference Hu, Dong, Wei, Zhu and Duan2004, Reference Hu, Dong, Luo, Su, Li and Zhou2011; Nadler & Streicher, Reference Nadler, Streicher, Nadler, Streicher and Long2004; Hu, Reference Hu2007). In Chongqing the main causes of habitat loss and degradation were formerly fuelwood collection and encroachment for agriculture but increasingly hydroelectric projects and tourism are having significant effects, and recent increases in the latter have aggravated the effects of habitat degradation. Since 2005 the total number of tourists visiting the three sites surveyed has been c. 2 million per year, of which c. 1.2 million were to Jinfoshan, and 400,000 each to Furongjiang and Heishangu (Chongqing Tourism Bureau, 2010, unpubl. data). There has been an increase in construction of tourism infrastructure and facilities, and consequent loss of forest. Habitat loss and degradation may concentrate colobines into smaller patches and aggravate intra- and inter-specific food competition, which may cause a further decline in population viability (Kirkpatrick, Reference Kirkpatrick1996), leading to a population crash (Davies, Reference Davies, Davies and Oates1994). Although François’ langur displays considerable flexibility in feeding and ranging, which helps the species survive and flourish in moderately disturbed habitats (Hu, Reference Hu2007, Reference Hu2009), habitat loss and degradation has extirpated the langur from some of its historical range in Guizhou (Hu et al., Reference Hu, Dong, Luo, Su, Li and Zhou2011).
Although the population of François' langur in Chongqing is only c. 200 in three isolated sites, these areas are adjacent to north-east Guizhou, where there are four sites with c. 60% of the total wild population of François' langur in China (Hu et al., Reference Hu, Dong, Luo, Su, Li and Zhou2011). The geographical proximity of the sites in Guizhou and Chongqing raises the possibility of setting up ecological corridors between some of them (such as between Jinfoshan and Dashahe; Fig. 1). These seven sites together contain almost 70% of the total remaining population of François’ langur and constitute their last remaining stronghold.
Acknowledgements
We thank W. Wang and J.L. Ma for their administrative help and for providing data on poaching, Y.Y. Zhang and the Fauna & Flora International China Programme for their support for this study, the local forestry authorities and all interviewees for their valuable information on the langurs and the number of visiting tourists, C. Chen, H.Q. Li and student volunteers for their help with the survey, Colin Groves, Anthony Di Fiore, Noel Rowe and two anonymous reviewers for their invaluable comments and suggestions, and to L. Liang for drawing the figure. This survey could not have been done without financial support from the China–Europe Biodiversity Conservation Project (No. 00060222), the Scientific Research Fund of Chongqing Education Department (No. KJ061304) and the Chongqing Specific Fund for Outstanding Young Scholar.
Biographical sketches
Zongxian Han has studied the ecology and conservation of François’ langur and other threatened mammals for over 10 years and has participated in several projects in south-west China. Gang Hu has studied the behavioural ecology, socioecology and conservation of langurs and other colobines in south-west China for over 15 years. He currently focuses on the behavioural flexibility of langurs in response to environmental variation and human disturbance, and management planning to minimize conflicts between langurs and local human communities. Shaobin Wu studies the ecology and conservation of rare mammal species in Chongqing, including langurs. Changlei Cao studies the ecology and conservation of rare wildlife and is involved in population surveys and monitoring projects in Chongqing. Xin Dong has research interests in wildlife conservation management, and particularly in habitat quality assessment, monitoring and management.
