Tropical forest region in Southeast Asia
Biogeographic Regions: oriental (Native )
Grubb (2005) also included China, but the occurrence there of T. kanchil is not confirmed, reflecting the difficulty of separation from T. williamsoni. Records from Yunnan Province seem on present evidence likely at least to include T. williamsoni (E. Meijaard pers. comm. 2008) but the additional occurrence of T. kanchil cannot be ruled out. Similarly, records in Viet Nam north to ca 22ºN (Dang Huy Hyunh 1994) and in the west part of Lao PDR’s northern highlands (e.g. Johnson et al. 2003; 21°N), could refer to T. kanchil or to T. williamsoni, or to a mix. It is unclear if there are chevrotains at all in the central and eastern parts of Lao PDR’s northern highlands. The record of chevrotain signs in Evans et al. (2000) for Phou Dendin NPA, far north-eastern Lao PDR and outside the interview derived range as shown in Duckworth et al. (1999: 269), was by a relatively inexperienced observer and upon recent re-examination of his original notes, he has retracted the record, since confusion with muntjac fawn was possible (W.G. Robichaud pers. comm. 2008). More telling, during interviews by the same observer in the same area in 2004–2005, villagers commonly reported that while they knew what a chevrotain ("kai", or "fan kai") is, most said they are essentially absent from the area (including the southern sector), and always have been. But whether they are truly absent or only rare differed somewhat with informants (W.G. Robichaud pers. comm. 2008). Likewise, the northern extent of T. kanchil in Thailand remains unclear. No species of chevrotain appears to penetrate Myanmar away from Tennaserim (Tun Yin 1967; Lynam 2003; in the latter, identifications to species should be disregarded). Well outside the confirmed range of the genus, Khan (1985) speculated that Lesser Oriental Chevrotain might occur in Bangladesh: "over a hundred villagers, whom I have interviewed have said they have either seen, killed or eaten such an animal". But he could procure no physical evidence for this, and none has been traced subsequently (Md Anwarul Islam pers. comm. 2008).
A recent taxonomic revision of chevrotains (Meijaard and Groves 2004), followed here, attaches a slight doubt to assume that any records, other than those validated through examination of specimens, from non-Sundaic South-east Asia, certainly refer to this species; as well as T. williamsoni (of at least northern Thailand, but perhaps with a much wider range) there is also T. versicolor of South-east Viet Nam, also perhaps with a much wider range. This Red List account assumes that the chevrotains of the Mekong basin and Thailand south from 18°N to the Thai–Malay peninsula refer to T. kanchil, on the basis of a fair number of specimens from many localities (E. Meijaard pers. comm. 2008: a listing of the mainland specimens examined for Meijaard and Groves 2004). Further south, T. kanchil is widely sympatric with T. napu, making chevrotains often challenging to identify (particularly for people with limited field experience of the genus) on field views, camera-trap photographs and even in hunting studies (Duckworth 1997; Matsubayashi and Sukor 2005). Many observers have therefore pooled their Sundaic chevrotain records as ‘chevrotain spp.’ (e.g. Bennett et al. 2000; Laidlaw 2000; Linkie et al. 2003; O'Brien 2003; Kawanishi and Sunquist 2004; Azlan 2006; Azlan and Engkamat Lading 2006; Lynam et al. 2007). Thus, the information base for determining this species' status is much thinner than would be expected from the general perception of the genus as being widespread and common, at least in the Sundaic part of its range.
Hoogerwerf (1970) wrote that "in Java T. javanicus was encountered in all provinces, fairly intensively distributed from sea-level to high in the mountains". Present-day occurrence in East Java is questionable: S. Hedges, M. Tyson and E. Meijaard (pers. comm. 2008) know of no certain records (as distinct from listings in collation documents which do not cite information to primary source) from areas like Baluran or Alas Purwo national parks, despite high survey effort during 1991–2000.
Meijaard and Groves (2004) gave only one specific locality for the species: Cheribon (= Cirebon) on the north coast of West Java Province, which is the type locality of T. j. pelandoc Dobroruka, 1967. Dobroruka (1967) also mentioned the western part of Java, to the southern coast, for what he called T. j. focalinus (which is T. javanicus s .s.). Meijaard and Groves (2004) did not list the localities for the many specimens they examined. Dobroruka (1967) and Van Dort (1986) both discuss variation within Javan chevrotains; neither has yet been checked for specific localities. More recent localities, although not of specimen records, include: Gunung Halimun (reportedly camera-trapped some time before 2003 [Suyanto 2003], but the photograph is not reproduced in the appendix and a painting is used for the species instead); Ujung Kulon (1991–1993; van Schaik and Griffiths 1996; C.P. Groves pers. comm. 2008), and the Dieng Plateau (during 1999–2000; V. Nijman pers. comm. 2008). There has been no collation of records from the various surveys over the last 20 years or so, and some observers no doubt are aware of other localities.
Tragulus javanicus, or lesser Malay mouse deer, do not have antlers or horns. Instead, adult males have elongated, tusk-like upper canines (Nowak and Paradiso, 1983). These canines protrude from the side of the mouth. Females lack these canines. Females are also smaller than the males (Nowak and Paradiso, 1983). The cheekteeth of lesser mouse deer have a crescent pattern formed by the enamel ridges. Mouse deer have no upper incisors. The pelage of mouse deer is brown with an orange tint. The underside is white. There is also a series of white vertical markings on the neck (Grzimck, 1994). Malay mouse deer have a triangular head and a round body with elevated rear quarters. The thin legs are about the diameter of a pencil (Nowak and Parasido, 1983). T. javanicus is the smallest artiodactyl, 18-22 inches long with a tail length of 2 inches (Grzimck, 1994). The young look like miniature adults when born; however, the tusk-like incisors in the infant males are not well developed.
Range mass: 1 to 2 kg.
Other Physical Features: endothermic ; homoiothermic; bilateral symmetry
Sexual Dimorphism: male larger
Average basal metabolic rate: 4.883 W.
Catalog Number: USNM 120574
Collection: Smithsonian Institution, National Museum of Natural History, Department of Vertebrate Zoology, Division of Mammals
Sex/Stage: Female; Adult
Preparation: Skin; Skull
Collector(s): Collector Unknown
Year Collected: 1902
Locality: Buitenzorg [= Bogor], near, Java, Jawa Barat, Indonesia, Asia
- Type: Miller, G. S. 19 Mar 1903 . Proceedings of the Biological Society of Washington. 16: 35.
Tragulus javanicus are found in overgrown primary and secondary forests in southeast Asia. They often reside around rocks, hollow trees, and dense vegetation near water.
Habitat Regions: tropical ; terrestrial
Terrestrial Biomes: rainforest
- Nowak, R., J. Paradiso. 1983. Walker's Mammals of the World. Chicago: John Hopkins University Press.
Habitat and Ecology
During the day it may prefer crown-gap areas with dense undergrowth of creeping bamboo, and move to higher and drier ridge areas at night. Home range All Tragulus species are associated with forests, but there is strong evidence that T. kanchil does not require old-growth forest or even particularly mature secondary forest. Its heavy use of disturbed habitats was confirmed by Matsubayashi et al. (2003), who found that in Kabali-Sepilok (Sabah), T. kanchil fed primarily in crown-gap areas dominated by bamboo stands, probably because it prefers fruits and soft leaves of pioneer plants with lower concentrations of secondary metabolites. Later, at another site in Sabah, Matsubayashi and Sukor (2005) again found a strong association of T. kanchil with immature forest. Johns (1997) found Tragulus to be more common in logged forests than in mature forests in Peninsular Malaysia, and densities tended to decrease again as logged forests matured. The abundance of Tragulus spp. (T. napu and T. kanchil combined, with no information on the proportions or even confirmation that both species were present) was higher (strongly statistically significantly so) in areas within 1 km inside the boundary of Bukit Barisan Selatan National Park, Sumatra, than in the interior of the park, suggesting higher numbers in somewhat encroached habitat (O'Brien et al. 2003); note that it is not confirmed that this information refers to T. kanchil. The genus (presumably including this species) was ubiquitous in a study in West Malaysia which paired each of seven of virgin jungle reserves with a nearby unprotected area with seriously encroached habitat (Laidlaw 2000). By contrast, at the Danum Valley, Sabah (East Malaysia), densities of T. kanchil were 21–39 animals per km² in primary forest but only 10–15 animals per km² in selectively logged areas (and there was an even bigger decrease of 55–66% for T. napu); in this area, where hunting is negligible and so is not confounding patterns, chevrotain densities were positively correlated with fruit abundance, and negatively correlated with pioneer trees and grass and herbs (Heydon 1994; Heydon and Bulloh 1997; Davies et al. 2001). Heydon and Bulloh (1997) suggested the obligate frugivory of chevrotains limited their ability to compensate for logging-led loss of fruit trees by browsing the abundant regrowth of logging areas. This may be significant in rationalising the differences between areas concerning the apparent effects of logging on the species: in some logging areas, fruiting trees are common along streams and tend to be left, and pioneer colonists produce abundant edible fruit. Around Buon Luoi, southern Viet Nam, most encounters with chevrotains (assumed to be largely or entirely this species rather than T. versicolor, based on ratios of specimens) were in places where tall forest alternated with scrub thickets, patches of grassland and moist riverine habitats. Riverbank slopes appeared particularly favourable to chevrotains (Kuznetzov and Borissenko 2004). E. Pollard (pers. comm. 2008) sees Lesser Chevrotains commonly in highly degraded forest in Mondulkiri province, Cambodia. In Thailand, R. Steinmetz (pers. comm. 2008) finds the species commonly in edge habitats, also widely, but probably less frequently, within big blocks of closed-canopy forest. By contrast in Lao PDR, the genus seems to be scarce in the interior of old-growth forest, based on surveys of many of the country’s remaining forests. Chevrotains have been camera-trapped in at least one protected area (Nakai–Nam Theun NPA), but only in one part of that (Johnson and Johnston 2007); the report does not discuss habitat. Tracks are readily identified by those familiar with them, but are rarely found within heavy forest in Lao PDR (R.J. Timmins pers. comm. 2008). Chevrotains are easily spot-lit at night where they occur, yet extensive spotlighting in south and central Lao PDR during 1992–1996, with most of such survey in areas below 550 m asl, found the species only once, and there were not many more records by day (Duckworth et al. 1994; Duckworth 1996, 1998; Evans et al. 2000; and a lack of records in many other sources). All or nearly all the handful of sightings (day or night) were in edge or heavily degraded areas (J.W. Duckworth pers. comm. 2008), but some lengthy surveys of lowland encroached forest failed to find the genus (notably in the ‘corridor’ sector of Phou Xang He NPA). This indicates a general overall rarity, or at least a localised distribution, either natural or human-induced. The apparently large numbers on the Vientiane plain (see Threats; Srikosamatara et al. 1992) (at the time of surveys perhaps among the most intensively hunted areas of Lao PDR for mammals of this size-class), including persistence at dreadful sites like Houay Nhang (whence almost all mammals larger than rats had been hunted out; see Threats), argues against human agency being responsible for the absence or low densities of chevrotains elsewhere in lowland Lao PDR in places like the corridor of Phou Xang He NPA and interior forests. Extensive surveys in the 1990s of Lao forest areas suggests almost certainly that in extensive tracts of semi-evergreen and evergreen forest below 1,000 m asl chevrotains are naturally scarce at least away from ‘edge areas’, because hunting pressure from snares, dogs, and day and night hunting with guns is relatively low, and other quarry species captured in similar ways to chevrotains are often common (R.J. Timmins pers. comm. 2008). In contrast, Kuznetzov and Borissenko (2004), who concurred that, in Viet Nam, closed-canopy forests often support few if any chevrotains, proposed that this is an artificial pattern produced through hunting (see Threats). In Lao forest ‘edge’ areas the very low numbers of chevrotains cannot necessarily be concluded to be a natural pattern, as these ‘edge areas’ tend to be in the vicinity of human settlements and in at least the 1990s experienced high levels of snaring and gun hunting (primarily by day) for primarily local consumption or local markets and ubiquitous hunting dog presence (R.J. Timmins pers. comm. 2008). Experiences in 2005, however, suggest that chevrotain presence is perhaps highly patchy and correlated to specific habitat features, as in the Nam Kading National Protected Area there was a clear hunting focus on chevrotains with an evidently significant capture rate; yet general wildlife field surveys in 2005 and 1995 observed the species in the field only once, despite significant spotlighting effort (Evans et al. 2000; Timmins and Robichaud 2005; W.G. Robichaud pers. comm.. to R.J. Timmins 2005; R.J. Timmins pers. comm. 2008). It suggests that the species is tied to specific microhabitats; perhaps this might be related to water requirements as there is a surprising regularity with which authors note commonness in riverine areas. Dumas’s (1944) description of the distribution in Cambodia strongly echoes modern findings that it is localised in Lao PDR and associated with gaps in Sabah: he stated that it was absent from heavy forest of tall trees without understory, and in general it was localised in occurrence but found somewhere in each of most of the forested provinces of Cambodia, wherever there are low, dense, thickets, and within clumps of bamboo, of which it is fond of the young shoots. Such a pattern 70 years ago seems unlikely to have been produced by hunting. In sum, the habitat use of this species with respect to edge–interior areas shows startling heterogeneity, at least in non-Sundaic areas, but this cannot yet be explained; in particular, disentangling the contributions of intrinsic habitat suitability and the effects of hunting is difficult.
The Lesser Chevrotain may even be able to survive in large numbers in plantation landscapes: in the Sarawak Planted Forests, Bintulu, Sarawak, Belden Giman (pers. comm. 2008) has many records from Acacia mangium plantations both mature and immature. However, this landscape contains many patches of secondary and old-growth forest (Belden et al. 2007), so further study is needed to assess use, if any, of landscape-level monocultures.
Most of this chevrotain’s geographic range is dominated by evergreen or semi-evergreen forests, but large areas of Cambodia, Thailand and to a lesser extent Lao PDR and Viet Nam are deciduous. One survey in Mondulkiri province, Cambodia, in 2008, found chevrotains to be widespread and common in mixed deciduous forest, dominated by e.g. Lagerstroemia spp., but they were not found in nearby deciduous dipterocarp forest (C. Starr pers. comm. 2008). In this same general area, J. Walston (pers. comm. 2008) also found them commonly in mixed deciduous forest, during 2002–2006. Field surveys of extensive landscapes of deciduous dipterocarp forest and mixed deciduous forest (in which ground fires are almost annual) in eastern and northeastern Cambodia (including Mondulkiri) have not detected the species (R.J. Timmins pers. comm. 2008) and camera-trapping from the same landscapes also suggest chevrotains are absent from deciduous regularly burnt forest tracts, but occur in semi-evergreen forest patches within the same landscapes (WWF unpublished data). Dumas (1944) stated that chevrotains were unknown in the deciduous dipterocarp forests of Cambodia, No other information on their use of deciduous landscapes has been traced; it is probable that even within deciduous dipterocarp they might use the semi-evergreen riverine strips within.
The Lesser Oriental Chevrotain may be absent from highlands throughout much of its range, and seems to be an extreme lowland specialist in some areas. The clearest analysis, from Borneo, found that it lived mostly below 100 m asl, with no record above 600 m asl (Payne et al. 1985). The maximum altitude in Indonesia (i.e. Borneo and Sumatra) is also said to be 600 m asl (G. Semiadi and Boeadi pers. comm. 2006); specific information from Sumatra has not been traced. In Lao PDR most known sites (see Duckworth et al. 1994; Duckworth 1996; Evans et al. 2000) are also below 600 m asl; however, chevrotains are abundantly hunted in the Nam Ha NPA which has a minimum altitude of 560 m asl (and rises to nearly 2,100 m asl) and is almost entirely rugged (Johnson et al. 2003), although these are perhaps more likely to be T. williamsoni than T. kanchil. There is a chevrotain specimen from Thateng on the Bolaven plateau (FMNH 38012) confirmed as T. kanchil (E. Meijaard pers. comm. 2008), suggesting occurrence at 800–1,200 m asl, although a short-distance trade origin may be possible. A recent finding of the genus being common in one part of Nakai–Nam Theun NPA (Johnson and Johnston 2007) gives no information on altitude of the records, and there are too few relevant specimens to allow a firm prediction as to which species they relate. Good populations certainly occur well above 600 m asl in Thailand: chevrotains are common in the degraded and edge habitats around Khao Yai National Park headquarters at 760–800 m asl (J. W. Duckworth pers. comm. 2008) and a number of specimens from nearby Pakchong suggest that these are likely to be T. kanchil, rather than unsuspected southerly occurrence of T. williamsoni. Also in Thailand, at Phu Khieo Wildlife Sanctuary chevrotains were camera-trapped commonly at 700–900 m asl, although the species involved is/are unclear (Lynam et al. 2001; A.J. Lynam pers. comm. 2008).
Lesser Chevrotains visit saltlicks regularly (Matsubayashi et al. 2007).
Lesser Chevrotain is often stated to be nocturnal, but in fact most activity is diurnal (in the first few and last few hours of daylight), with animals spending most of the night resting (Matsubayashi et al. 2003; Matsubayashi and Sukor 2005). It is highly solitary (Matsubayashi et al. 2006). Apparent territorial behaviour has been observed (Davison 1980); home ranges and core areas overlap considerably between males and females but core areas are completely separate between individuals of the same sex (Matsubayashi et al. 2006). Males are highly philopatric but females establish new home ranges when giving birth (Ahmad 1994 in Meijaard et al. 2005). Matsubayashi et al. (2003) estimated home-range size for females at 4.3 ha and for males 5.9 ha using the minimum convex polygon method, but the difference between males and females was not significant. The mean daily distance travelled for males was 519 ± 89 m, that for females 574 ± 220 m. The Lesser Chevrotain is partly frugivorous, but also eats substantial quantities of shoots and young leaves, mostly of fast-growing gap species rather than closed forest understory species; the latter tend to be richer in secondary protective compounds (Dang Huy Huynh 1968; Medway 1983; Kuznetzov and Puzachenko 1992; Matsubayashi et al. 2003). Among fruits, those of mass 1–5 g and seeds of mass 0.01–0.5 g are particularly taken (Heydon and Bulloh 1997). It seems to be facultively monogamous–polygymous (given the congruence between core areas of individual males and females with a conspecific of the opposite sex; Matsubayashi et al. 2006), with post-partum oestrus, females being almost continuously pregnant (Cadigan 1972 in Meijaard et al. 2005). The gestation period is 140–177 days (Lekagul and McNeely 1977), with 2–3 young per year, the fawns kept hidden (Meijaard et al. 2005).
Habitat and Ecology
T. javanicus is a ruminant and has a three-chambered stomach (Lawlor, 1979). As ruminants, mouse deer use microorganisms that produce enzymes within the stomach to digest their food. In the wild, lesser Malayan mouse deer are commonly herbivores and folivores, eating leaves, buds, shrubs, and fruits that have fallen from trees. In zoos, mouse deer tend to eat insects as well as leaves and fruits (Nowak and Paradiso, 1983).
Life History and Behavior
Perception Channels: tactile ; chemical
Status: captivity: 12.0 years.
Lifespan, longevity, and ageing
Mouse deer can breed at any time of the year. The gestation period is usually 4 1/2 months. Breeding females produce one fawn (Jinaka, 1995). The young are precocial when born and can stand within 30 minutes of birth (Grzimck, 1994). Mouse deer are shy and their fawns tend to be "hiders". The fawn is weaned for 10-13 weeks. It reaches sexual maturity at about 5-6 months. Lesser Malay mouse deer can live for 12 years.
Key Reproductive Features: gonochoric/gonochoristic/dioecious (sexes separate); sexual
Average birth mass: 370 g.
Average gestation period: 144 days.
Average number of offspring: 1.
Average age at sexual or reproductive maturity (male)
Sex: male: 167 days.
Average age at sexual or reproductive maturity (female)
Sex: female: 167 days.
Molecular Biology and Genetics
Barcode data: Tragulus kanchil
Below is the sequence of the barcode region Cytochrome oxidase subunit 1 (COI or COX1) from a member of the species.
See the BOLD taxonomy browser for more complete information about this specimen.
Other sequences that do not yet meet barcode criteria may also be available.
-- end --
Download FASTA File
Statistics of barcoding coverage: Tragulus kanchil
Public Records: 1
Specimens with Barcodes: 2
Species With Barcodes: 1
The mouse deer population is threatened by hunting and habitat destruction.
One method to prevent the extinction of mouse deer is captive breeding. This is done primarily in zoos (Jinaka, 1995).
US Federal List: no special status
CITES: no special status
IUCN Red List of Threatened Species: least concern
- Jinaka, H. 1995. Endangered Animal of February 1999 -Mouse Deer. A Guide to the Threatened Animals of Singapore, 2: 38-39.
IUCN Red List Assessment
Red List Category
Red List Criteria
There are several major uncertainties, although these are not adjudged sufficient for listing as Data Deficient: (1) a rather low proportion of modern records of Tragulus from both the Sundaic and non-Sundaic portions of the range have been identified conclusively to species; (2) the conflicting nature of the information available concerning the effects of hunting (harvest levels are locally very high) and habitat destruction; (3) strong indications that in its non-Sundaic range it is localised in occurrence, a pattern for which the reasons remain opaque, but which has been proposed to result from hunting; (4) the apparent restriction to the extreme level lowlands in at least Borneo, placing it within the altitudinal zone of rapid forest loss.
IUCN Red List Assessment
Red List Category
Red List Criteria
Non-Sundiac populations are poorly documented. Chevrotains are common in Thailand’s Khao Yai National Park, at least in the degraded and edge habitats around the park headquarters (J. W. Duckworth pers. comm. 2008) and Phu Khieo Wildlife Sanctuary (Lynam et al. 2001; identifications to species should be disregarded). The genus seems to be of localised occurrence in Lao PDR (Duckworth et al. 1994, 1999; Duckworth 1996, 1998; Evans et al. 2000; Johnson et al. 2003; Johnson and Johnston 2007; see Habitat and ecology). Chevrotains seem also to be localized in Viet Nam. They are common in Cat Tien National Park, southern Viet Nam (Borissenko and Ivanova 2003; Polet and Ling 2004). Somewhat further north, intensive sign and spotlighting searches at several sites in lowland Dak Lak in 1997 failed to find the genus (Le Xuan Canh et al. 1997; an earlier listing in a report for that area should not be taken seriously, see Duckworth and Hedges 1998: 66–68), although these were mostly in rather or entirely deciduous areas and perhaps outside the species' habitats; however, landscape in parts of the areas surveyed is similar to that in eastern Mondulkiri, Cambodia (see below). Further north again, they were common around the Buon Luoi area, although seem to have declined recently (Kuznetzov and Borissenko 2004). As in Lao PDR, there are many other areas within the genus’s range (as defined by Dang Huy Hyunh 1994) from which reasonably intensive mammal surveys have failed to find the species (e.g. Phong Nha Ke Bang National Park [Timmins et al. 1999] and the Huong Son District of Ha Tinh Province [Timmins and Trinh Viet Cuong 2001]). In Cambodia, chevrotains were recorded commonly by recent surveys in at least one area of Mondulkiri province dominated by semi-evergreen forest (T.D. Evans pers. comm. 2008; E. Pollard pers. comm. 2008; J.L. Walston pers. comm. 2008; C. Starr pers. comm. 2008); but in the primarily deciduous lowlands presence appears to be patchy (see Habitat and Ecology), but no good modern composite picture for the country is available (R.J. Timmins pers. comm. 2008). Dumas (1944) considered that it occurred in most of the forested provinces of Cambodia, wherever its favoured microhabitats were found.
North of the sympatry with T. napu, T kanchil is heavily hunted and is often considered one of the most preferred wild meats (Srikosmatara et al. 1992; Duckworth et al. 1999; Kuznetzov and Borissenko 2004; R. Steinmetz pers. comm. 2006, R.J. Timmins pers. comm. 2008). Nonetheless, it is clear that T. kanchil is adept at surviving in areas (at least in the lowland plains) in the face of very heavy hunting, major habitat degradation and isolation from other forest areas. Exemplifying this, it was one of the few mammals larger than a rat confirmed to persist at Houay Nhang, an isolate of a few km² of highly degraded forest just outside the capital of Vientiane facing, in the early 1990s, exceedingly concentrated hunting with guns and snares (Duckworth et al. 1994). A little to the west it persists in Sangthong district, which was heavily logged just before a 1996 survey and also supported heavy hunting (Duckworth 1996). These two areas are within 10 and 70 km, respectively, of, and ecologically contiguous with, Thangon, from where Osgood (1932) reported specimens of T. kanchil (confirmed by E. Meijaard pers. comm. 2008 as this species), and thus the modern records probably do relate to T. kanchil. Hansel (2004) provided a further record from the adjacent low hills. Numbers of chevrotains in Vientiane markets in the early 1990s (Srikosamatara et al. 1992) were so high that they could not be at the brink of extinction in this degraded and fragmented landscape. Chevrotains are still common in trade in the city, although mostly in transit rather than on open sale in the city’s markets, and their origin is not clear (T. Hansel pers. comm. 2008). The genus was recently camera-trapped quite commonly in a part of Nakai–Nam Theun NPA where ground mammals were much depleted (Johnson and Johnston 2007), perhaps indicating higher resilience than many other mammals, although there is no information to compare with pre-exploitation levels.
While non-Sundaic chevrotains can, therefore, clearly persist under extreme pressures in at least some circumstances, this does not rule out that densities might be quickly reduced through the twin pressures of hunting and habitat degradation. Decreasing population densities between the 1970s and 1990s, in at least parts of southern Viet Nam, were suggested by both spot-lighting surveys and qualitative assessments of footprints. Near Buon Luoi in December 1978, chevrotains were spotlit at about three animals per 5 km of transect route, but by 1993, when forest had undergone heavy degradation and hunting was very heavy, chevrotains were not even found every night with at least 10 km of transect (Kuznetzov and Borissenko 2004). Kuznetzov and Borissenko (2004) proposed that night-time hunting may explain chevrotain’s near-absence from habitats with open-understory (this feature allowing chevrotains to be readily hunted using spotlights and guns) in the Buon Luoi area, because by contrast in Cat Tien National Park, where the use of firearms has been banned for decades and enforcement is fairly effective, chevrotains are encountered with comparable frequency both in thickets and on exposed forest floor (Borissenko and Ivanova 2003). This sighting rates contrast is consistent with a difference between the headquarters area of Khao Yai National Park, Thailand (where chevrotains are readily found by spotlight) and which is relatively well-protected from hunting, and Lao forests (see above) where hunting was effectively unrestrained (J.W. Duckworth pers. comm. 2008). However, a proposal that, where hunting is unrestrained, as it is in most of Viet Nam and Lao PDR, chevrotains would be common in heavy forest but have been effectively eradicated from such habitat, is flawed for at least Lao PDR, where numbers of forest mammals killed through spotlight hunting are probably insignificant compared with those trapped (including with snares) and perhaps even killed using dogs (J.W. Duckworth and R.J. Timmins pers. comm. 2008, based on numbers of active hunters encountered at nigh versus traps found, in many areas of Lao PDR). Kuznetzov and Borissenko (2004) based their reasoning on only two sites and it is possible that other factors explain the differences. In sum, the effects of hunting and habitat modification, and the interaction between them, on non-Sundaic chevrotains remain most unclear.
The effects of potential threats operating in the Sundaic range are not understood well, either. Lesser Oriental Chevrotain is now rather rare on Singapore (K. Lim pers. comm. 2008), presumably nowadays constrained by habitat quality rather than by ongoing hunting, although the role of free-ranging dogs, if any, in restraining population numbers is unclear. However, the ‘island within an island’ situation of remaining forest on Singapore and the extreme habitat loss over the last 150 years prevent useful extrapolation of this species' situation there to other Sundaic areas. Poaching pressure in Kerinci-Seblat National Park, Sumatra, among snared ungulates was greatest for muntjacs and lowest for chevrotains, as assessed by diameter of the cable (Linkie et al. 2003). Chevrotains were more then nine times as abundant in areas of Bukit Barisan Selatan National Park, Sumatra, with low than with high human population density within 10 km of the park boundary, suggesting low resilience to human presence, presumably the effects of hunting (O'Brien et al. 2003); note that it is not confirmed that this information refers to T. kanchil. Bennett et al. (2000) profiled the effects of hunting in both Bornean states of Malaysia (Sarawak and Sabah), each with eight study areas, upon chevrotains (not distinguished to species) through a lengthy hunting study in February 1993 to June 1995. They constituted 8–9% of animals killed. No direct or index measure of chevrotain densities was possible at these sites to determine what effects, if any, such heavy hunting was having on chevrotains. Chevrotains are among the most favoured wild meat by inhabitants of the Sarawak Planted Forests area, Bintulu, Sarawak, after Sambar Rusa unicolor and wild pigs Sus (Belden Giman pers. comm. 2008). Logging can apparently drive substantial reductions in density, at least locally, although other studies have found it allows an increase in numbers (see Habitat and ecology). No information on the use of plantation landscapes has been traced. In the Greater Sundas, lowland forest is being converted at unprecedented high rates. At least in Borneo, where this species is concentrated in areas below 100 m asl, most of its habitat either has been substantially modified during the last two decades or will be in the next two (e.g. Holmes 2000; BirdLife International 2001; Jepson et al. 2001; McMorrow and Talip 2001; Lambert and Collar 2002; Fuller et al. 2003; Kinnaird et al. 2003; Curran et al. 2004; Fuller 2004; Eames et al. 2005, Aratrakorn et al. 2006). However, inferring proportionately large population losses in Lesser Chevrotain populations is not possible: depending on how serious hunting is and what exactly is being done to the habitat post-logging (forest regeneration, tree plantation or non-woody habitat) in any given area, chevrotain numbers might remain broadly unaffected, decrease or increase in logged-over areas.
Chevrotains occur regularly in markets in towns such as Jakarta, Surabaya, Yogyakarta, and Malang, but often they are cramped in small cages, and can therefore be overlooked, and may even be more common than observations suggest; numbers in trade are surprisingly high given the small numbers to be seen in the field nowadays (V. Nijman pers. comm. 2008). They have been traded at high levels for many decades: Hoogerwerf (1970) wrote of “numerous reports of mouse deer being regularly trapped and offered for sale alive” in Java. They are hunted and traded both for pets and as wild meat (S. Hedges pers. comm. 2008). Numbers passing through markets in Jakarta, Bogor, and Sukabumi have recently declined sharply, perhaps because of tightened control by the forest police in those markets; but there is the possibility that falling trade might indicate, at least in part, increasing difficulty to catch the animal and thus a decline in populations. In the Malang area, it is still “relatively easy” to procure one, although it is “getting time consuming” (G. Semiadi pers. comm. 2008). Hunting is probably largely with snares; dogs are also likely to be a serious threat (M. Tyson pers. comm. 2008). They are vulnerable to active hunting at night through a propensity to freeze when spotlit. However, the effects of these comparably high hunting levels on the genus in Java have not been studied empirically.
The continued presence of many animals in markets suggests that significant populations remain somewhere on Java (assuming that the animals are not now imports from elsewhere in Indonesia), and thus that a major decline may not have occurred, despite current indications. A comparable situation was found with inornate squirrel Callosciurus inornatus in Lao PDR, when extensive 1990s surveys found few animals in the field, in contrast with historic statements of abundance and ongoing substantial numbers being traded in fresh meat markets. This led to conservation concern for the species (e.g. Duckworth et al. 1999); but later field survey of degraded and edge areas found out that the species was indeed common and evidently a species associated with degraded areas, and had hence been severely under-recorded by the 1990s surveys (Timmins and Duckworth in press).
The taxonomic revision of Meijaard and Groves (2004) means that T. javanicus is endemic to Java, and, if it occurs, Bali. There are fair indications of a decline, perhaps a major one, and thus this newly-revealed endemic species should be swiftly removed from the anonymity in which it has lain for decades. An urgent first step is a collation of existing information, because it is possible that many more records exist than were traced during the preparation of this assessment. This should include trawling likely observers and examination of collections not covered by Meijaard and Groves (2004). If insufficient number of specimens are found to clarify the number of taxa on Java, more should be obtained. Some may come from markets but, because of the pre-eminence of locality in determining systematics among very similar taxa (e.g. Groves in press: discussion under Wapiti group), specimens of known locality origin must form the basis of analysis. Whatever the number of species on Java, the difference in sighting rate between Hoogerwerf (1970) and observers from the 1990s onwards is suggestive of a major decline, although other explanations are possible (see Population and Threats). Current status needs to be clarified through specific surveys for the genus (camera-trapping, but undertaken in a way more suitable for smaller species than is usual, spotlighting, and hunting surveys may all play a role). Surveys must take care to investigate secondary and edge areas which are often eschewed by general wildlife surveys in favour of the less encroached areas.