|Scientific Name:||Tragulus javanicus|
|Species Authority:||(Osbeck, 1765)|
Cervus javanicus Osbeck, 1765
|Taxonomic Notes:||Grubb (2005) stated that “Meijaard and Groves (2004) were not convinced that the original description of Cervus javanicus Osbeck, 1765 is a mouse-deer and preferred to date the name from Tragulus javanicus (Gmelin, 1788) [=Moschus javanicus Gmelin, 1788]”. However, Meijaard and Groves (2004) actually dated Cervus javanicus Osbeck to 1757, not to 1765. The situation is complicated and Meijaard and Groves (2004) indicated that they would be making a submission to the International Commission for Zoological Nomenclature. Pending the result of this, the name for this species follows usage by Grubb (2005).
Whoever the name is credited to, the lack of any biological type material has led to its application within the genus Tragulus having a contorted history, elaborated in Meijaard and Groves (2004). In sum, usage in past publications of T. javanicus can rarely if ever be taken at face value to represent the species as circumscribed here, unless explicit mention is made that the statements refer to the animals of Java. Up to and including Chasen (1940), the name T. javanicus was used for the larger chevrotains, with Moschus kanchil Raffles, 1821 for the smaller ones (dated as 1822 by Meijaard and Groves 2004). Van Bemmel (1949), building on Chasen’s (1940: 194) comment that “I rather doubt if javanicus [i.e. the larger chevrotains] occurs there [ = anywhere on Java]”, proposed that T. javanicus should be used for the smaller chevrotains, with Moschus napu F. Cuvier, 1822 being the first available name for a larger chevrotain. This system almost universally was followed up to Meijaard and Groves (2004) who proposed that the smaller chevrotains contained several species, and that the form on Java was specifically distinct and endemic to the island. This form being T. javanicus, nearly all the off-Java forms fell under T. kanchil, as they previously had done under Chasen (1940). Tragulus kanchil williamsoni Kloss, 1916, placed within T. javanicus from Van Bemmel (1949) onwards, was segregated as a monotypic species by Meijaard and Groves (2004).
Further change within the taxonomy and nomenclature of the genus is likely, given the ominous statement in Meijaard and Groves (2004) that "because many type specimens, especially of small island taxa are located in the Smithsonian Museum, which we did not visit, we were unable to assess the validity of all taxa. We intend to address the variation between these island taxa...in future publications". Whatever the findings of these examinations, they are unlikely to affect usage of the name T. javanicus for a chevrotain on Java. A separate issue is whether there is truly only one species on the island. Meijaard and Groves (2004) reviewed past claims of two subspecies of smaller chevrotain and concluded that "our craniometrical data indicate that there may indeed be two distinct mouse-deer taxa on Java. The data are, however, inconclusive...further research...is required". These two taxa would both belong in the smaller chevrotains (the T. javanicus group). A yet further unresolved issue concerning the genus on Java is that three chevrotain specimens objectively and clearly identifiable as large chevrotains (T. napu group) were reported by the collector, E. Dubois, to be from Java; two have the precise and traceable locality of Mount Willis (Meijaard and Groves 2004). The latter suggested that "it may be that Dubois collected these specimens during his earlier fieldwork in Sumatra, after which they were mislabelled", a speculation for which there seems to be no foundation other than the lack of other specimens of T. napu from Java.
In this preliminary assessment, the provisional course of assuming that all records of chevrotains from Java relate to T. javanicus is taken with reservations. Field identification of Tragulus is difficult, even under the pre-2004 conventional arrangement of two species widely sympatric. If there is only one species on Java, all records of the genus can safely be assumed to relate to that species; but if there are two or even three, assignment to T. javanicus without critical evaluation is not possible.
|Red List Category & Criteria:||Data Deficient ver 3.1|
|Assessor(s):||Duckworth, J.W., Timmins, R. & Semiadi, G.|
|Reviewer(s):||Brook, S.M. & McShea, W.J.|
Tragulus javanicus is listed as Data Deficient partly reflecting the lack of clarity over how many species of chevrotain occur on Java and therefore how much of the available information about the genus refers to T. javanicus. However, even taking an assumption that there is only T. javanicus on the island, information applicable to red listing is limited. Given the considerable uncertainties that exist for all members of the genus (see Red List accounts for other species), largely as a result of general oversight of chevrotains in faunal investigations, the comparison of some historical accounts of T. javanicus with recent fieldwork results is difficult to interpret, not least the species tolerance of hunting is poorly known, as is it's likely response to habitat disturbance. Additionally the presumed short generation length of the species, considered to be likely under five years, also influences assessment, in that, for decline criteria to be invoked in Red Listing one would have to assume relatively high rates of decline in a relatively short window of time, ca 10-15 years, since the present. There are fair indications of a decline, perhaps a major one, and a category such as Vulnerable is quite likely to be applicable. Dedicated field investigations (throughout Java) of status are urgently warranted, and the species Red List status should be reviewed regularly in light of current uncertainty and concerns.
Tragulus javanicus as here defined is endemic to the island of Java, Indonesia, according to Meijaard and Groves (2004). The latter authors did not mention the island of Bali, but a sighting was reported from Bali Barat National Park, Bali, in a bird watching trip report (Birdquest 2006). The genus was not listed for Bali in the exhaustive reviews of Chasen (1940; of the genus) or Meijaard (2003; mammals of Indonesian islands), nor by Grubb (2005). Given the live-animal trade in the genus on Java, further records are needed to confirm whether or not there is a native population of the genus on Bali. If there is, biogeographic considerations suggest it would be most closely related to the Javan population and quite probably conspecific.
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.
|Range Map:||Click here to open the map viewer and explore range.|
|Population:||There appears to have been no field study specifically of chevrotains on Java but current wildlife surveyors agree that they are rarely seen compared with results from similar styles of observation in Kalimantan (S. van Balen, V. Nijman, E. Meijaard, M. Tyson and S. Hedges, all pers. comm. 2008). Even Hoogerwerf (1970), who clearly found the species with some regularity (explicitly not daily, but enough for him to learn its calls and thereby establish dates of a probable mating season), complained that “it is particularly difficult to obtain any insight into the situation of this species in Java”, but concluded that “it is improbable that the species is in danger of extinction”. The basis for an unattributed statement in Whitten et al. (1996), that “it is still numerous and can be seen easily in many areas such as the tourist park part of Pangarandaran Nature Reserve” is unclear. The genus was recorded at five of ten sites on the Dieng Plateau surveyed in 1999–2000, but mostly only through single observations (V. Nijman pers. comm. 2008). S. van Balen (pers. comm. 2008) points out that in Java the genus seems very shy (compared with animals in Malaysia and Kalimantan), so most records come as footprints. This contrasts with the assignment by Hoogerwerf (1970) that the genus was common and widespread. It is therefore quite plausible that a major decline has taken place, although other explanations remain to be explored which could have bolstered Hoogerwerf’s sighting rates, such as his having a dog with him which flushed the chevrotains, or his spending a lot of time in the species' favoured microhabitats. Specifically, neither M. Tyson nor S. Hedges (pers. comm. 2008) saw chevrotains during a rhinoceros survey in Ujung Kulon in 1992, whereas this was Hoogerwerf’s (1970) main site in assessing the species as relatively common and readily found. It was camera-trapped there five times during 1991–1993 (van Schaik and Griffiths 1996), a rate comparable with that of many other species in the study and certainly not suggesting out-and-out rarity. Very recently, numbers in trade in Java have dropped sharply in most cities, and it is plausible that this reflects increasing difficulty in procuring the animal (G. Semiadi pers. comm. 2008).|
|Habitat and Ecology:||Hoogerwerf (1970) wrote of chevrotains on Java occurring "from sea-level to high in the mountains". In the Dieng plateau area, V. Nijman (pers. comm. 2008) found them only a few times in the lowlands (400–700 m asl), where most survey took place, and had no records from above about 1,500 m asl. They have been found on Gunung Gede–Pangangro up to about 1,600 m asl (V. Nijman pers. comm. 2008). Hoogerwerf’s (1970) description of favoured habitats on Java suggests that chevrotains there might be an 'edge' species, certainly seeming to prefer areas with thick understorey vegetation, such as that along riverbanks. This would not be unusual within the genus (see other Tragulus accounts).|
|Use and Trade:||For Use and Trade information see under Threats.|
Java has highly fragmented natural habitats and has done for centuries, reflecting longstanding high human population densities. Many protected areas were established during the Dutch colonial period but from independence up until the 1970s they were largely under-funded and neglected. After hosting the World Parks Conference in 1982, the Indonesian government gazetted a swathe of national parks and more structured conservation planning began, funded by the World Bank and other donors. The focus was largely on the 'multi-function' national parks and much money was spent on infrastructure, some staff training and increased personnel. The 'lesser' protected areas such as "game reserves" and "nature reserves" still had few staff and resources, and that has continued to the present. During the 1980s to the mid 1990s, guns were tightly controlled and the military and police were feared and respected. However, the strong culture of caged bird keeping meant that hunting, including that within protected areas, was primarily for birds and some small game, through various forms of trapping, including snaring; this latter could well have included chevrotains. There was some habitat loss from protected areas through illegal logging, agricultural encroachment and other offtake, but the national parks of Java remained remarkably intact for much of this period. Socio-political changes from 1997 led to a reduction in the respect for the police and military and the rise of a viewpoint that protected areas were the peoples' resources and would therefore benefit from decentralised management (S. Hedges and M. Tyson pers. comms. 2008). This policy change, which risked a ‘tragedy of the commons’, has indeed led to increased destruction and poaching in the past decade. There is some evidence that species readily uses edge and secondary vegetation, meaning that effects of forest encroachment will be less serious than for old-growth forest obligates. Moreover, chevrotains already seemed rare for surveyors operating in the early–mid 1990s, when habitats had been relatively stable since 1970s or early 1980s. For any decline which may have occurred in the 1980s and early–mid 1990s, therefore, habitat factors are an unlikely driver.
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 2008).
Chevrotains on Java occur in some protected areas, e.g. Ujung Kulon (Hoogerwerf 1970; van Schaik and Griffiths 1996) and were earlier said to “occur in all game sanctuaries in Java and in most of the nature reserves established on that island...." (Hoogerwerf 1970). The species has been officially protected since 1931, yet it is still hunted.
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, 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.
|Citation:||Duckworth, J.W., Timmins, R. & Semiadi, G. 2015. Tragulus javanicus. The IUCN Red List of Threatened Species. Version 2015.2. <www.iucnredlist.org>. Downloaded on 05 September 2015.|