Callerebia dibangensis ( Lepidoptera : Nymphalidae : Satyrinae ) , a new butterfly species from the eastern Himalaya , India

A new species of butterfly in the genus Callerebia (Butler, 1867) is described from the Upper Dibang Valley District, Arunachal Pradesh, India. A combination of very distinctive characters: large size; highly rounded wings; striking under hindwing white scales; distinctive under hindwing tornal ocelli; large round forewing orange apical spot and a dark brown under ground colour distinguishes this butterfly from any other Callerebia species.


INTRODUCTION
Callerebia (Butler, 1867) is a genus of approximately 11 species and 20 subspecies (Appendix 1), almost entirely confined to the Sino-Himalayan region.The genus is composed of medium to large butterflies with small bodies in relation to their broad, rounded wings.The hindwing in many species is produced at the tornus to form a slight lobe.The antennae are thin, only slightly thickening towards the tip.The markings, though very variable follow a fairly constant arrangement: the upper ground colour is dark brown to blackish; the forewing with a bi-pupilled apical ocellus; the upper hindwing with or without one tornal spot; the under hindwings are often covered by whitish scales of different densities, tones and patterns, with or without one or two tornal ocelli; in addition, some species have a set of up to four post-discal white dots on the under hindwing.All species may also have additional ocelli as part of their variation.Older literature on Indian butterflies (Talbot  Bruna et al. (2000) treat these species as being under Callerebia, Hemadara and Paralasa.Huang (2003) described a new species of Callerebia -C.ulfi from northwestern Yunnan and did a revision of C. polyphemus and its subspecies and had a note on the status of C. suroia.Bruna et al. (2000) comments on the taxonomic uncertainty of this genus, the remarkable variation in wing pattern and genitalia morphology.It is certainly the case that the variation in pattern that exists between different populations has made the use of key-based identification very difficult and has led to confusion over the status and identity of certain species and forms.Several of the eastern Himalayan species are characterised by their large orange-ringed forewing apical ocellus contrasting prominently against a dark brown, blackish upper ground colour.The under hindwings are covered with prominent white striations.Though many of the species appear confusingly similar, this new species is visually very distinct and is a beautiful and striking addition to the eastern Himalayan fauna.
The Upper Dibang Valley District (Fig. 1) lies between 95 0 15'-96 0 35'E & 28 0 22'-29 0 27'N.It is further north than Bhutan, Sikkim and eastern Nepal.It is situated in a very complex and active geological zone where the main Himalayan axis turns abruptly north-west to south-east forming a ridge of moderately high mountains to the north and east, between 4000-5400 m, which separate the Dibang Valley from the Pemako region of Tibet to the north and Chayu county of Tibet to the east.A western ridge largely above 3600m separates the Dihang and Dibang valleys.The district capital Anini (1,968m) is situated on a plateau between the confluence of the Dri and Mithun rivers and its annual rainfall (2003) was 3,281.33mm (Upper Dibang Valley official website 2012).The predominant natural vegetation around 1800m is sub-tropical hill forest and temperate forest.However, there are quite large areas of grasslands in the Dri and Mithun valleys probably formed principally by human activity.The district is sometimes regarded as India's remotest, on account that the road connection is frequently closed by landslides from April to October.72.32% of the district is recorded as forested (Department of Environment and Forests, Government of Arunachal 2012) and has the lowest population density of any district in India (<1km -2 ) (Upper Dibang Valley official website 2012) (Fig. 1 -Satellite Image of Upper Dibang Valley District).

METHOD AND MATERIALS
The majority of the sites visited were in the Dri and Mithun valleys within a day's return walk from Anini (c.2,000m) from the period of 17 July to 13 August 1987.14 to 16 August 1987 were spent travelling to the Ithun Valley on the southern edge of the Upper Dibang Valley District.The Dibang Valley government provided a guide and assistance, which was invaluable.Only species considered of interest and which could not be determined in the field were collected.These are deposited in the Natural History Museum, London, U.K.
Previous to the visit to the Upper Dibang Valley, the author had undertaken numerous visits to Nagaland, Manipur and Arunachal Pradesh and so was familiar with many of the butterfly species that commonly occur in northeastern India.

Diagnosis
There does not appear to be any other species of Callerebia like this species in the Natural History Museum collection.
The following is a list of characteristics that are unique to this specimen and/or rarely expressed by other Callerebia species.Reference can also be made to images polyphemus, is similar to C. suroia, but is usually not so well marked on the under hindwing.Comparison has only been made against male specimens.Females are usually larger, upper ground colour paler, underside markings more pronounced and the apical ocellus rounder.
(i) Size: One of the largest male Callerebia specimens (Table 3); only large forms of C. polyphemus and C. scanda opima are of comparable size.Within the NHM collection, C. dibangensis sp.nov. is noticeably larger than other male specimens.
(ii) Wing shape: The forewings are very large and broad, the costa convex and forming a smooth curve all around the apex to the termen.The hindwing tornus is barely produced and rather round.Only specimens of C. scanda opima from eastern Bhutan express a similar forewing shape, but their hindwing tornus tends to be noticeably produced.
(iii) Underside hindwing pattern: The under hindwing patterns of Callerebia (Image 4) are complex and difficult to describe.In several species the scales on the hindwing are arranged as short little white lines (iv) Underside hindwing tornal ocelli: The ocellus in space 2 is larger than is typical for Callerebia and the ocellus in 1c tiny and blind.It is the only species where the ring colour is orange rather then yellowfulvous.When ocelli are expressed in other species this combination of a large ocellus in space 2, with a tiny blind ocellus in 1c, does not appear typical.
(v) Forewing apical ocellus: The ocellus is very large, rather round and has a wide and fairly regular orange ring of a uniform colour.C. suroia and C. polyphemus also have a large ocellus, but it is very different from the new species: when large the ring is constrained distally; far more irregular; elongated posteriorly and with a greater amount of red suffusion.The ocellus of C. orixa is the most similar in terms of appearance and possible species of Callerebia.There was no comparable species in the allied genera of Hemadara, Loxerebia, Argestina and Paralasa which are comprehensively covered by Bruna et al. (2000).The plates of d 'Abrera (1985'Abrera ( , 1990'Abrera ( , 1992) ) covering Satyrinae in the Holarctic and Oriental regions were also referred to and there appears to be no other genera this species may belong to.
On the basis of the unique combination of these characteristics mentioned above I propose this as a new species.
Upperside: Ground colour dark chocolate brown, slightly blackish, paler towards margins.Very large round orange-ringed apical ocellus (11mm), black inner with two white pupils.Edges of ring distinct and regular.Hindwing single tornal ocellus in space 2. Narrow reddish ring, black inner and white pupil.
Underside: Ground colour dark chocolate.Forewing apical ocellus large as on upperside, but with an additional thin reddish outer ring.Forewing termen margin covered with a thin scattering of white scales tapering towards tornus.Hindwing white scales forming prominent small lines (striae) with slight violet tones covering the whole wing apart from a narrow area around the tornal ocelli.Striae become darker towards the costa and apex giving a fading effect.In the basal half they are more separate and alternate with the dark brown ground colour producing a distinctive snow-drift like appearance.They are densest towards the discal region in spaces 1, 2 and 3 where they merge with one another.Hindwing two tornal ocelli, ocellus in space 2 much larger then 1c which is blind.Both ocelli orangeringed with a black centre.Ocellus in space 2 with a white pupil.

Etymology
Named after the Dibang Valley District in Arunachal Pradesh, India.

Location
Only one specimen was observed and collected on the 27 July 1987 and is the holotype.The species was collected along the track that leads from Anini to Mipi between the altitude of 1600-1800 m approximately The closest species appears to be C. scanda opima (Images 2-4, Fig. 2).Fourteen males of C. scanda opima in the NHM collection were compared with C. dibangensis sp.nov.and the key external features are summarised in Table 1.
Table 2 is a list of all Callerebia species with their key characteristics.It has been highlighted where they differ from C. dibangensis sp.nov.either as a group or individually.
Species names and genera were referenced in LepIndex (Beccaloni et al. 2003) and Savela (2012)   Eastern Himalayan (roughly east of Eastern Nepal) species with a large to very large ocellus that substantially occupies the apical region of the upper f/w.Ring colour usually orange fulvous but may be also be yellowish-orange to reddish.Ocellus outer rings often irregular especially when ocellus ring is wide.Under h/w with prominent white striations overlaid with strong to weak bands and/or variegation.Often large butterflies.These species differ significantly in wing shape from C. dibangensis sp.nov.: f/w less rounded and more produced; h/w produced and with lobe; under h/w overlaid with bands and variegation.
(i) There appear to have been very few butterfly surveys of the Dibang Valley.Evans (1914)  (ii) The specimen was collected at the end of July in Upper Dibang Valley.Many places such as Eaglenest (West Kameng District) have access problems in the monsoon and are consequently poorly surveyed in the months of July and August which also appears to be the main flight period for C. scanda opima according to the NHM collection labels.
(iii) The most surveyed areas in Arunachal Pradesh appear to be Eaglenest (West Kameng District) and Namdapha (Changlang District).Both these areas are on  Images 2-4).In addition an allied species Hemadara narasingha has been recorded from Sikkim, Arunachal Pradesh (Abor Hills) and the Lower Tsang Po.This species is smaller, with narrow forewings and the hindwing angled on the termen and produced at the tornus.A description may be found in Talbot (1947).I cannot find much literature regarding recent records of Callerebia in Arunachal Pradesh.Betts (1950) records C. orixa from West Kameng 'Rupa, 5000ft.Sept. Common in grassland in sheltered valleys in the temperate Momba country.' Older literature records two more possible Indian species.South (1914) records a specimen of C. polyphemus collected by F.M. Bailey at Tawang, Arunachal Pradesh.It should be noted that C. polyphemus has not been recorded so far west and C. suroia was only described in 1914 so the identification of this record needs to be confirmed.South (1913) also records that F.M. Bailey collected a specimen of C. baileyi from Dichu (6000ft, 1829m).The Dichu River, Anjaw District, Arunachal Pradesh is just before the border with Tibet.The majority of the specimens were from Drowa Gompa, Tibet (3,048m) to the north of the Mishmi Hills.I have not seen any contemporary records for this species in India.
Rather speculatively I think the new species is closely allied to C. scanda opima on the basis of wing shape and the under hindwing pattern.There also appear to be minute white dots in space 6 of the under hindwing.These white dots do not appear to occur in C. polyphemus, C. suroia, C. annada and C. orixa. Bruna et al. (2000) comment that most Callerebia species are allopatric.However, the distributions of C. suroia and C. orixa overlap in Manipur and Nagaland, though it is not known if they fly together in the same habitat.It will be interesting to have contemporary field data and phylogenetic studies to clarify the relationships of the different forms and species.

Conclusions
Callerebia dibangensis sp.nov is a new species of Satyrine butterfly from the Dibang Valley in the eastern Himalayan region of India possibly allied to Callerebia scanda opima.I propose that this common name is determined by the people of the Dibang Valley, since this butterfly appears to be endemic to that valley.

Table 2 . Key identification features of male Callerebia species.
lists the butterflies collected by F.M. Bailey during his tour of southeastern Tibet.He started his trip from the Mithun Valley in mid May 1913, but does not appear to have recorded any butterflies from the region.Borang et al. (2008) conducted a survey of the Dihang-Dibang Biosphere Reserve during November 2006 and October 2007 and recorded 134 species/subspecies, but this covered many places outside of the Dibang Valley and the survey dates were probably rather late for higher elevations and they did not record any Callerebia species.Endemicity is not a feature of Himalayan butterflies, but as outlined in the introduction the Dibang Valley is in a very unique geographical zone.Two Callerebia species from neighbouring regions of Tibet -C.tsirava and C. baileyi appear to have very restricted distributions.