Ungulate taxonomy revisited: the evidence for the splits of G&G

TeaLovingDave · 25 Nov 2018

lintworm said:
Differences in karyotype have been reported between populations, as has ancient dna introgression with Gaur.

And, more notably, prehistoric hybridization between burmanicus and Kouprey - this being the reason for the unclear taxonomic status of the latter taxon for a time.

lintworm · 27 Nov 2018

TheGerenuk · 27 Nov 2018

Will there be more ungulate groups done?

Jurek7 · 27 Nov 2018

Thank Lintworm, very interesting.

So most of the proposed splits don't stand.

I notice that you are already applying very easy criteria to split the species anyway. Size difference in Wisent is used as an argument, but bigger size difference in African Buffalo signifies nothing but ecomorphs. Size of white patches is an argument to split Common and Pyrenean chamois, although color patches on wild sheep signify only subspecies.

It looks like an argument often used is that two forms show some difference consistently (even not so big one). This will however easily lead to error with small sample size or a local inbred sub-population.

lintworm · 28 Nov 2018

TheWalrus said:
Will there be more ungulate groups done?

All the other Ungulate groups will be covered in due time, but it will take some time before this list will be completed.

Jurek7 said:
Thank Lintworm, very interesting.

So most of the proposed splits don't stand.

I notice that you are already applying very easy criteria to split the species anyway. Size difference in Wisent is used as an argument, but bigger size difference in African Buffalo signifies nothing but ecomorphs. Size of white patches is an argument to split Common and Pyrenean chamois, although color patches on wild sheep signify only subspecies.

As you will have seen, all the splits you mention here are not only based on morphological data, but also on genetic data (see the references at the end of each profile). I wouldn't split Wisent, for a lack of genetic evidence, even though morphologically the animals are quite distinct. We don't know anything about their evolutionary history though. In the case of African buffalo there are size differences between the forest and the savanna buffaloes. The genetic data do however show that the forest ecotype has developed multiple times independently out of a savanna ancestor and all are nested within savanna types. So it is wrong to recognize the forest ones as a different species, as then you would end up with 10+ different buffalo species, which hardly differ genetically.

In the case of Common and Pyrenean chamois, the morphological difference is backed up with very clear genetic data showing a long time of divergence between two species, this is now partly visible in different color patters, mating behavior as well as in some skull characteristics. Pelage differences between different Urial or Argali subspecies are not backed up with genetic data, as in they have diverged only very very recently and many of the proposed subspecies aren't even monophyletic based on genetic data.

I have always tried to use a combination of data sources, where possible multiple morphological traits as well as genetic data, most often only Mtdna is available, but where possible also nuclear dna.

Jurek7 said:
It looks like an argument often used is that two forms show some difference consistently (even not so big one). This will however easily lead to error with small sample size or a local inbred sub-population.

That is indeed the whole problem of the strict PSC approach that G&G have used in their book. Combined with statistical abuse this has resulted in a large number of splits that don't stand, especially when they are integrated with other data sources. G&G are very much old school taxonomists and their work has been really valuable, but with their strict PSC approach they missed the whole story that genetic data can tell.

lintworm · 7 Jan 2019

MOSCHIDAE

The Musk deer family consists of only one genera: Moschus in which 7 species are currently recognized. Though they are called deer, they are actually the closest relatives of the Bovids. There has been considerable taxonomic debate about the number of valid species of Musk deer, ranging from one to seven depending on the author. Currently 7 species are generally recognized, though there still is some debate on the validity of some species and genetic data are mostly lacking.

Siberian musk deer
The Siberian musk deer (Moschus moschiferus) is the most widespread musk deer species and occurs from extreme NE Kazakhstan throughout much of E Russia, N Mongolia, N China and the Korean Peninsula. Five subspecies are currently recognized, but their validity is questionable:

M.m. moschiferus Kazakhstan, Siberia & Mongolia
M.m turovi Far Eastern Russia
M.m. arcticus Veryokhansk ridge
M.m. sachalinensis Sakhalin Island
M.m. parvipes Korean peninsula

Moschus moschiferus moschiferus

@Dormitator , Edinburgh Zoo, UK

@Tomek , Zoo Berlin, Germany

Anhui musk deer
The Anhui musk deer (Moschus anhuiensis) is a monotypic species that occurs only in the Dabie Mountains in the west of the Anhui Province, China. There has been debate as to whether this species is actually a subspecies of Forest musk deer (Moschus berezovskii)

No pictures of this species have been uploaded to the gallery.

Forest musk deer
The Forest musk deer or Dwarf musk deer (Moschus berezovskii) occurs throughout China and also in N Vietnam. Currently four subspecies are recognized:

M.b. berezovskii Sichuan, Qinghai, Tibet
M.b. bjiangensis northwestern Yunnan
M.b. caobangis Yunnan, Guangxi, Guangdong, northern Vietnam
M.b. yunguiensis Yunnan-Guizhou highland, Hunan, Jiangxi

Only one picture of this species has been uploaded to the gallery:

@Deer Forest , Museum of Biological Specimens, Beijing, China

Black musk deer
The Black musk deer (Moschus fuscus) is a monotypic species that occurs in the Himalayas in China, Nepal, Bhutan, India and extreme NW Myanmar.

No pictures of this species have been uploaded to the gallery.

Alpine musk deer
The Alpine musk deer (Moschus chrysogaster) is a Musk deer from Central and Southern China, and extreme N Nepal, Bhutan and India. Two subspecies are generally recognized:

M.c. chrysogaster SE and S Tibet, Nepal, Bhutan, India
M.c. sifanicus Qinghai, Gansu, Ningxia, western Sichuan and northwestern Yunnan

No pictures of this species have been uploaded to the gallery.

Himalayan musk deer
The Himalayan musk deer (Moschus leucogaster) was formerly considered a subspecies of the Alpine musk deer, but is now generally recognized as a distinct monotypic species. It occurs in the foothills of the Himalayas from Bhutan, through N India, Nepal and extreme S China.

No pictures of this species have been uploaded to the gallery.

Kashmir musk deer
The Kashmir musk deer (Moschus cupreus) was described as a subspecies of Alpine musk deer, but is now generally regarded as a separate monotypic species. This species occurs in the Himalayas in extreme NW India and N Pakistan, the Kashmir Region and N Afghanistan.

No pictures of this species have been uploaded to the gallery.

Chlidonias · 12 Jan 2019

lintworm said:
Wild sheep

Argali

No pictures of darwini, nigrimontana, severtzovi, karelini, arabica, isphaganica or laristanica have been uploaded to the gallery yet

I have put a couple of photos of Ovis ammon darwini in the Mongolia wildlife gallery.

lintworm · 11 Feb 2019

CERVIDAE

Cervidae, the Deer family, is the second largest ungulate family and traditionally +- 55 species of Deer have been recognized, this number was increased somewhat by G&G, but not as crazily as for Bovids. Deer taxonomy on both the species and the genus level is still a hot topic and future changes are to be expected.

Cervinae

Muntiacini

Tufted deer

The Tufted deer (Elaphodus cephalophus) is a small deer species from China and possibly NE Myanmar. Traditionally four species have been recognized, though G&G do not name any in their book because of lacking data, though the few samples available seem to show large variability.

E.c. cephalophus SW China, NW Myanmar
E.c. michianus SE China
E.c. ichangensis C & S China
E.c. fociensis distribution unclear, validity doubtful

E.c. cephalophus

@Maguari, Lowry Park Zoo, Tampa

E.c. michianus

@Tim May , Wroclaw Zoo, Poland

lintworm · 11 Feb 2019

Red muntjac species complex

The Red or Indian muntjac (Muntiacus muntjak) was previously considered as a single species ranging from India to Indonesia. Mainly based on a different chromosome number this species has been split in 2: Northern red muntjac (Muntiacus vaginalis) that ranges from the Indian peninsula and Indochina to N Malaysia and the Southern red muntjac (Muntiacus muntjak) in Malaysia and Indonesia. This split is accepted by e.g. the IUCN but is not free of discussion, mainly because chromosome numbers have only been sampled from very few individuals. In the book Ungulate Taxonomy G&G further split Red muntjak additionally recognizing Muntiacus malabaricus from Sri Lanka and W Ghats, India; Muntiacus aureus NW & C India and parts of Myanmar; Muntiacus nigripes from Hainan, Yunnan and N Vietnam. Muntiacus montanus from the Sumatran highlands.

Apart from Muntiacus montanus these splits are generally not accepted and the IUCN red list authors can tell exactly why:

Groves (2003) and other authors (e.g., P.M. Giao et al. 1998) have suggested that northern forms here placed within M. vaginalis may constitute more than one species, and Groves and Grubb (2011) separated the taxon into four species, M. vaginalis s.s.(central range, the most widespread of the four), M. malabaricus (extreme southwest of range), M. aureus (western range with an outlier in the central part of the range) and M. nigripes (eastern parts of the range). The purported differences between these taxa include body size, antler size and general and or midback, nape and crown pelage colour and or more specifically its darkness. The purported descriptions presumably meant to indicate diagnostic differences and the grounds for considering the species valid are often internally inconsistent, for example “shorter antlers” is clearly implied as a characteristic of M. vaginalis, M. malabaricus is stated to have “shorter antlers [than M. vaginalis]”, in M. aureus the “antlers short” and in M. nigripes once again “antlers short”. Specimens used and sample sizes are not given, nor is there in general any discussion of variation, nor reference to alternative explanations for the geographical distribution of traits. With the obvious propensity for muntjacs to undergo chromosome rearrangements, and the clear indications from certain areas of multiple sympatric and or closely parapatric muntjac species, it is clearly quite possible (perhaps even probable) that M. vaginalis as used here may involve multiple phylogenetic species. However the documentation of such species, especially for the sake of taxonomic stability, requires the burden of proof to be placed on those authors attempting taxonomic splits. For the reasons already outlined, small sample sizes (morphological or genetic) from disparate range locations are not adequate proof, for such a widespread, ecologically tolerant species, that effective reproductive isolation has occurred between sister lineages.

In the context of the above it has been pointed out for instance by Groves (2003) that dark-legged Chinese and Indochinese animals (‘M. nigripes’ of Groves and Grubb 2011) differ from paler, more uniform, animals to the south and west (‘M. vaginalis’ of Groves and Grubb 2011, for further details, see also C.P. Groves in litt. 1999 to Duckworth et al.1999). However, recent data (especially the results of camera-trapping studies) from Indochina suggest that pelage colour and pattern variation is little more than a geographically based polymorphism within a widespread species, and might not warrant even use in designation of subspecific taxa (R.J. Timmins pers. comm. 2008, based on unpublished data from various camera-trapping projects by WWF/WCS/FFI/SFNC/IUCN). There are certainly no grounds from field studies to suggest that more than one species-level taxon is present in Indochina (R.J. Timmins pers. comm. 2008).

Three species are for now recognized:

Northern red muntjac
The Northern red muntjac (Muntiacus vaginalis) ranges from in the Indian Peninsula through Indochina S to N Malaysia at the Isthmus of Kra. Chromosome count 2n=8 in females, 2n=9 in males. The following 8 subspecies have traditionally been recognized, though their validity is sometimes doubtful.

M.v. vaginalis Nepal. Bhutan, NE India, Bangladesh
M.v. annamensis S Laos, S Vietnam & Cambodia
M.v. aureus Pakistan, N & C India
M.v. curvostylis Myanmar, Thailand
M.v. malabaricus S India & Sri Lanka
M.v. menglalis S China, possibly N Indochina
M.v. nigripes Hainan Peninsula
M.v. yunnanensis S China

G&G merge annamensis, menglalis into curvostylis and yunnanensis in nigripes.

No subspecies:

@Giant Eland , Shadow Nursery, USA

Muntiacus vaginalis vaginalis

@Chlidonias , Assam State Zoo, India

Muntiacus vaginalis curvostylis

@Giant Eland, Khao Keow Forest & Wildlife Reserve Park, Thailand

Southern Red muntjac

The Southern red muntjac (Muntiacus muntjak) is a monotypic species from Malaysia, S of the Isthmus of Kra and Sumatra, Java, Borneo, Bali and associated islands. Chromosome count 2n=6 in females, 2n=7 in males.

@Giant Eland , Taman Safari Bogor, Indonesia

Sumatran mountain muntjac

The Sumatran mountain muntjac (Muntiacus montanus) is a species of questionable validity of which hardly anything is known. This species is limited to mountanous areas of W Sumatra, Indonesia.

No pictures of this species have been uploaded to the gallery.

Chlidonias · 11 Feb 2019

lintworm said:
The Red or Indian muntjac (Muntiacus muntjak) was previously considered as a single species ranging from India to Indonesia. Mainly based on a different chromosome number this species has been split in 2: Northern red muntjac (Muntiacus vaginalis) that ranges from the Indian peninsula and Indochina to N Malaysia and the Southern red muntjac (Muntiacus muntjak) in Malaysia and Indonesia. This split is accepted by e.g. the IUCN but is not free of discussion, mainly because chromosome numbers have only been sampled from very few individuals.

I didn't know they had been split. I'll have to have a look at this and see what I think of it - I might be able to add an armchair tick!

Chlidonias · 11 Feb 2019

Chlidonias said:
I didn't know they had been split. I'll have to have a look at this and see what I think of it - I might be able to add an armchair tick!

Turns out I don't think much of that split. It is likely I actually already knew about it and forgot again because (to quote the IUCN) it "rests on little evidence". Phrases used throughout the IUCN taxonomy account are in the vein of "postulated", "assumption", "assumed", "purported", "minor", and it ends with "There appears to have been no significant further investigation of this taxonomic split since 2008, and thus the 2014 reassessment largely for the sake of status quo maintains this taxonomic treatment".

Groves (2003) elected to raise non-Sundaic forms of M. muntjak (s.l.) from subspecific taxa to the species M. vaginalis, leaving the Sundaic forms to constitute M. muntjak (s.s.). In 2008 the IUCN Red List chose to follow this split, with the following caveats. This position, already postulated by previous authors (e.g. Groves and Grubb 1990), rests on little evidence, primarily the assumption that all taxa within M. muntjak in the Sunda region carry a unique karyotype different from all M. vaginalis in northern regions. However, the Sundaic karyotype seems to have been documented in only a few individuals from the Malay Peninsula south of the Isthmus of Kra (Wurster and Aitkin 1972, Groves and Grubb 1990, Groves 2003, Tanomtong et al. 2005), assumed to be M. muntjak (the type locality of the later is Java). Other purported differences (e.g. dorsal darkening and short nasals, Groves 2003, Groves and Grubb 1990), if they can be considered diagnostic characters rather than traits (see below), appear minor rather than ones likely to separate species-level taxa. Although differences in chromosome number between muntjacs appears likely to constitute a good species boundary, preventing extensive interbreeding (see Groves and Grubb 1990), a much wider sampling of karyotype (especially within the presumed range of M. muntjak (s.s.)) is needed to place on solid ground the systematic position which assumes separate and widespread ranges for the two so far identified karyotypes. There appears to have been no significant further investigation of this taxonomic split since 2008, and thus the 2014 reassessment largely for the sake of status quo maintains this taxonomic treatment.

TheGerenuk · 11 Feb 2019

Glad to see that this thread is still active! Keep up the good work!

ThylacineAlive · 11 Feb 2019

The US population has traditionally listed as Muntiacus muntjak vaginalis. Would this not translate into M. v. vaginalis once split, or were the subspecies not described until after the split?

~Thylo

Chlidonias · 12 Feb 2019

ThylacineAlive said:
The US population has traditionally listed as Muntiacus muntjak vaginalis. Would this not translate into M. v. vaginalis once split, or were the subspecies not described until after the split?

The subspecies were already in place - they have just split the species in two (and apparently dropped any subspecies which were formerly recognised in the Sunda region, to make muntjak monotypic).

UngulateNerd92 · 5 May 2019

I am sorry to bump this thread, but I can't help but to wonder what Colin Groves' and Peter Grubb's rebuttals would be to the findings/analysis done by @lintworm here, if they were still alive.

Great work by the way! I can't thank you enough for putting this thread together!

lintworm · 11 May 2019

Hog deer

The Hog deer (Axis porcinus) is a small species of deer from Southeast Asia. This species has recently been placed in the genus Hyelaphus and one genetic study placed it as a sister to Rusa instead of Axis deer, but later genetic studies have confirmed that this was a mistake and Hog deer are indeed sister species to Axis deer and do not warrant recognition of a separate genus as the estimate date of divergence is only 2.6 MYA. Traditionally two subspecies have been recognized:

A.p. porcinus Indian hog deer; Pakistan, N and NE India, Bangladesh and Nepal
A.p. annamiticus Indochina hog deer; Myanmar, Cambodia & Manipur, India

The exact historic distribution is unclear and only recently was a population in E India reliably assigned to annamiticus. Both subspecies are elevated to species status by G&G

Sample sizes

Skins
No sample sizes are given

Skulls
porcinus 5-6
annamiticus 2-4

Skins

annamiticus is described as being much brighter, more ochery and less grey and buffy compared to porcinus. annamiticus is also not speckled.

Skulls

porcinus and annamiticus cannot be distinguished based on skull characteristics.

Additional data

There has been one recent study that compared porcinus and annamiticus hog deer genetically (Gupta et al. 2018). This study found a sequence divergence of +- 4% and has an estimated time of divergence of 220.000 years ago (95% HPD 110.000-338.000 years ago).
From personal observations in zoos it seems that there is high variation in coat colouration between porcinus Hog deer between individuals and throughout the year, with colour ranging from ochery to dark grey, which is used here to differentiate between annamiticus and porcinus, the more ochery coloured porcinus do however normally have spots.

Summarizing

The genetic data indicate the existence of two clearly differentiated subspecies that might be approaching species status, though on PSC criteria they represent clearly separate species. Morphological differences are however very slight, which could be due to the extremely low sample sizes used by G&G. More information on the exact distribution and improved morphological analyses, including antlers, would be necessary to assess whether annamiticus indeeds should have species status. For now treating them as a separate management unit is certainly necessary.

Axis porcinus porcinus

Male

@Goura Burgers' Zoo, Arnhem, Netherlands

Female

@Chlidonias Kaziranga NP, India

Axis porcinus annamiticus

@Giant Eland , Khao Kheow Forest & Wildlife Reserve Park, Thailand

Reference

Gupta et al. (2018): Genetic analysis of endangered hog deer (<i>Axis porcinus</i>) reveals two distinct lineages from the Indian subcontinent

lintworm · 13 May 2019

Apart from Hog deer, three other species are recognized in the genus Axis

Axis deer

The Axis deer (Axis axis) is a medium-sized deer from India, Sri Lanka, Nepal, Bhutan and Bangladesh. It is considered monotypic

Male

@Chlidonias , Wilpattu NP, Sri Lanka

Female

@devilfish , Tatu Carreta, Argentina

Bawean deer

The Bawean deer (Axis kuhlii) is a monotypic critically endangered deer species that is restricted to the island of Bawean between Java and Borneo.

Male

@Maguari , Edinburgh Zoo, UK

Female

@Giant Eland , Taman Safari Bogor, Indonesia

Calamian deer

The Calamian deer (Axis calamianensis) is a monotypic critically endangered deer that is restricted to the Calamian Islands in the Phillipines.

Male

@Blackduiker , Los Angeles Zoo, USA

Male & female

@Ituri , Phoenix Zoo, USA

lintworm · 13 May 2019

The sister genus to Axis is Rucervus, until recently Brow-antlered deer were also assigned to this genus, but multiple studies that worked with Mtdna have clearly shown that Brow-antlered deer are sister to Pere-david deer and is sister to Cervus. This means that either Cervus will need to be enlarged or Brow-antlered deer should be placed in the monotypic genus Panolia.

The genus Rucervus now consists of only one extant and one extinct species.

Barasingha
The Barasingha or Swamp deer (Rucervus duvaucelii) is the only extant member of the genus Rucervus. This is a large deer from India and S Nepal and three subspecies are generally recognized:

R.d. duvaucelii N India, S Nepal
R.d. branderi C India
R.d. ranjitsinhi NE India

G&G elevate all subspecies to species status.

Sample sizes
No new data seem to have been used and all shown results seem to be based on Groves (1983).

Skins
No sample sizes are given

Skulls/antlers
duvaucelii 9-13
ranjitsinhi 2-3
branderi 2-5

Sample sizes are thus mostly extremely limited.

Skins
No details on skins are given for branderi except that this species has "well-knit" feet with hairy pasterns, possibly an adaptation to living in less marshy areas than the other Barasingha. ranjitsinhi is described as having less prominent white spots in moult, short tail and small pointed ears with very little white hair on the inside when compared to duvaucelii.

Skulls/antlers
Several differences in antlers and skulls are described by G&G. Though these might be true when looking at the averages, there is however wide overlap in these measurements when looking at the standard deviation. Taking into account the extremely small sample sizes for especially the skull measurements, this means no conclusion can be drawn from these data. The only possible difference might be that branderi is relatively small with relatively large antlers, but given the paucity of data this cannot be reliably concluded.

Additional data
A recent genetic study (Kumar et al. 2017) using both nuclear and Mtdna found that each subspecies is genetically different. They did however find evidence for relatively recent colonization, though the date of divergence was not estimated. The genetic distance between the taxa was relatively high and would approach species status. Given the fact that genetic distance increases faster in small isolated populations, which was likely the case for Barasingha, this might however not equal species status.

Summarizing
Overall there seems to be no strong case for splitting Barasingha, except if you strongly favour a PSC approach. Given the current morphological data, differences can hardly be identified reliably. The genetic results show that the three subspecies are clearly identifiable, but differences are not huge and estimated times of divergence are not known. Future research on Barasingha should however be a priority to assess how long the different populations have been isolated and whether (and which) morphological differences do exist.

It remains unclear to which subspecies captive Barasingha in Europe and America belong, though they are often labelled as duvaucelii. The only Barasingha of which pictures have been uploaded from the wild are from Kaziranga, India.

Rucervus duvaucelii ranjitsinhi

@Chlidonias , Kaziranga NP, India

Zoo Barasingha; male

@ro6ca66 , ZSL Whipsnade Zoo, UK

Female

@ro6ca66 , ZSL Whipsnade Zoo, UK

Schomburgk's deer

The Schomburgk's deer (Rucervus schomburgki) is an extinct deer species that originally occurred in Thailand, though rumours of continued survival exist, this species is generally considered to be extinct since 1938.

Uploaded by @Joker1706 , Zoo Berlin, Germany

References

Groves (1983): https://www.biodiversitylibrary.org/page/48745060#page/686/mode/1up

Kumar et al. (2017): Mitochondrial and Nuclear DNA Based Genetic Assessment Indicated Distinct Variation and Low Genetic Exchange Among the Three Subspecies of Swamp Deer ( Rucervus duvaucelii)

ThylacineAlive · 13 May 2019

lintworm said:
It remains unclear to which subspecies captive Barasingha in Europe and America belong, though they are often labelled as duvaucelii. The only Barasingha of which pictures have been uploaded from the wild are from Kaziranga, India.

What about the small number of branderi listed for some European zoos? Most are listed as nominate, but some of the French collections and a single animal at Port Lympne are listed as branderi.

~Thylo

lintworm · 13 May 2019

ThylacineAlive said:
What about the small number of branderi listed for some European zoos? Most are listed as nominate, but some of the French collections and a single animal at Port Lympne are listed as branderi.

~Thylo

Zootierliste states that based on the first ISB it is unclear where all the founder animals were sourced and it is thus impossible to assess to which subspecies they belong... I have not been able to find any other information on their origin, except it being stated somewhere that most if not all American & European are likely duvaucelii... But if someone else has new information that would shed light on the status of these branderi specimens, that would be great...

Ungulate taxonomy revisited: the evidence for the splits of G&G

Moderator

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Moderator

Well-Known Member

Well-Known Member

Moderator

Moderator

Well-Known Member

Well-Known Member

Moderator

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member