On Hypolycaena from Maluku, Indonesia, including the first description of male Hypolycaena asahi (Lepidoptera, Lycaenidae)

Abstract The taxonomy and distribution of the five species of Hypolycaena in Maluku are discussed and new locality records given. Corrections are made to the published taxonomy and distribution of Hypolycaena phorbas (Fabricius, 1793). This clarification enables a better understanding of the biogeography of the genus. Hypolycaena asahi Okubo, 2007, was originally described from a single female from Ambon and is here recorded from Seram. The male is described for the first time.


Introduction
The Indonesian provinces of North Maluku and Maluku consist of numerous islands, yet their butterfly fauna remains less well described than those of the principal surrounding areas of the Philippines, Sulawesi and New Guinea. Vane-Wright and Peggie (1994) comment that, geologically, the northern islands of Halmahera, Ternate, Morotai and Bacan form a complex of land areas variously related to New Guinea, while the Buru, Ambon, Seram arc is related to North-West Australia. The Sula islands of Taliabu, Mangole and Sanana, in the west of Maluku were included faunistically in the "Sulawesi region" by Vane-Wright and de Jong (2003), while Burrett et al. (1991) link Sula geologically with Banggai and Obi. The islands of the Aru group in the south east of Maluku share the continental shelf of, and are faunistically close to, the New Guinea mainland.
Thus Maluku sensu lato remains an area of immense biogeographical interest, with the largest of its islands forming the northeasterly part of Vane-Wright's "Wallacea": the land between the Sunda and Sahul shelves. To facilitate testing of biogeographical hypotheses, it is important that the taxonomy and distributional data of all butterfly families represented in Maluku is accurate and as comprehensive as possible. The extensive lycaenid fauna is perhaps the least understood.
The genus Hypolycaena C. & R. Felder, 1862, (Lycaenidae, Theclinae, Hypolycaenini) consists of about 25 species in the Indo-Australian region as well as about 20 species in Africa. Fiedler (1992) included Chliaria Moore, 1884, and Zeltus de Nicéville, 1890, within Hypolycaena whilst Eliot (1992, retained these as separate genera in his subtribe Hypolycaeniti. H. asahi was described by Okubo, in 2007, from a single female specimen. The male is described here for the first time and enables its relationship with other species of the genus to be more closely determined. This paper is primarily concerned with Hypolycaena species in Maluku. However, it is necessary to discuss in some detail the taxonomy and wider distribution of H. phorbas (Fabricius, 1793) and its allies. These taxa form a species group in which the males exhibit a large circular dark brand of apparently normal scales on the upperside of the forewing and in which the early stages are polyphagous and strongly myrmecophilic (Fiedler, 1992). This study will confirm the identity of the taxon found on Aru Islands and also clarify D'Abrera's record of H.erasmus Grose-Smith, 1900, in Halmahera. A more accurate understanding of the taxonomy and distribution of the phorbas species group will in turn lead to a better understanding of the biogeography of the Papua mainland and the islands to the East and West of it.

Geopolitical terminology
The Indonesian western half of the Island of New Guinea and its associated offshore islands, which has previously been known as Irian Jaya, now consists of two provinces: Papua and West Papua. However "Papua" has also been used to denote this whole area. For simplicity we will use the term "Papua mainland" to describe the whole area excluding offshore islands.

equipment and methods
The preserved material forming the basis of this study is primarily that of the collections of the Natural History Museum London (BMNH) and of the second author. Where their reliability is assured, other distributional data have been accepted in correspondence from curators of other private collections.
Male genitalia were prepared by soaking in 0.1N potassium hydroxide solution for 24 hours at room temperature prior to dissection. Micro-photography of the genitalia, while suspended in 80% Iso-Propanol, was with an AIGO GE-5 digital microscope and the images were subsequently processed using Helicon-Focus 5.0 software (Helicon Soft Ltd. 2010) to enhance depth of field.
All photographs of preserved adult specimens, except those kindly provided by Mr. Yusuke Takanami, were taken using a Nikon D80 digital SLR camera fitted with a Micro-Nikkor 60mm macro lens. The photographic images presented were postprocessed for exposure compensation, cropping, resizing and sharpening using Adobe Photoshop Elements 6.0. The scale on photographs represents multiples of 5mm.   The external morphology of all of these new females is inseparable from that of the H. asahi holotype, and their identification as examples of H. asahi is assured. We propose the hypothesis that the male specimen is also H. asahi because of its underside markings and its sympatry with the aforementioned female from Seram. ♂ Upperside. Forewing length 13mm. Both fore and hindwings metallic blue with dark borders. The forewing black border about 1mm wide at the tornus but rapidly widening along the termen to meet the costa at its mid-point, then running down to the base above vein 12, but not quite entering the cell. The forewing also with basal swelling of veins 2, 3 and 4 with a faint brand of seemingly normal (not androconial) scales surrounding these swollen veins. The hindwing black between veins 7 and 8, and with a dark grey dorsal border in spaces 1 and 1a. In space 1a a small black tornal lobe with a white marginal streak. Filamentous white-tipped black tails at veins 1b and 2, 2mm and 3mm long respectively.

Hypolycaena asahi
Underside. No significant differences exist between the undersides of the females from Ambon and those of both sexes from Seram. ♂ Genitalia. Saccus short, bluntly pointed. Brachia long and tapering to a fine point, with a broad elbow and a pronounced lobe at the proximal junction with the tegumen. Valvae short, broad and conjoined basally, tapering distally with the apex rounded and the inner margins finely serrate. Aedeagus medium length, the sub-zonal portion shorter that the supra-zonal portion.
Remarks. The early stages are unknown. The females of H. asahi (Figs 3-8) from both locations show varying amounts of basal blue scaling not evident in the holotype. Otherwise, they conform closely to Okubo's description.
Okubo notes the similarity between this species and two allied species from the Philippines: H. shirozui (Hayashi, 1981) and H. toshikoae Hayashi, 1984 (Fig. 10). In all three species the underside hindwing tornal orange area extends into space 3 and the sub-marginal black spot in space 3 is much larger and darker than in space 4. On the underside in both sexes, asahi shows a much more marked dislocation of the post-discal band than in shirozui, while this dislocation is absent in toshikoae. On the male upperside, the apical black border in asahi is much broader than in either of the other two species, most notably in spaces 2, 3 and 4. On the female upperside, neither H. shirozui nor H. toshikoae exhibits the white forewing discal patch of H. asahi. Both Philippine species have orange inwardly surrounding the hindwing sub-marginal lunules, which is absent in H. asahi from both Seram and Ambon.
In the male genitalia, H. asahi is distinguished from these other species by the more elongate valvae, which have more rounded apices, and by the shorter suprazonal portion of the aedeagus.
In the same paper Fruhstorfer described H. erylus pigres from Obi, based on a series of eight males. Having examined the pigres and thyrius holotypes, as well as a long series of Obi and Halmahera specimens, we can not see any clear differences between the two taxa and therefore consider pigres to be a synonym of thyrius, which appears earlier in Fruhstorfer's work.

Hypolycaena erylus incertae sedis. Figs 23-28.
Although H. erylus is widespread in the South-East Asian islands and into New Guinea, material from South and South-East Maluku is scarce. We have seen a single male from Banda and BMNH has three males of H. erylus from Tanimbar (20 miles north of Saumlaki, Yamdena -1917-1918. These are all difficult to assign to a particular named subspecies (the males of the different subspecies tend to be fairly similar whilst the females vary more). See Figs 23-26.
In addition BMNH holds one female from Manawoka Island (label reads: Manovolka. 13.xi.(18)99. H. Kühn) in the Gorong Islands, which is unlike any other subspecies, having extensive pale areas on the upperside -especially the forewing -and may represent a new subspecies. See Figs 27, 28.
We await further confirmatory material before naming any further subspecies based on these few specimens.
D'Abrera and Seitz, 1926, considered Felder's taxon dictaea to be a subspecies of phorbas found on Waigeo only, whereas Parsons "provisionally treated" dictaea as a separate species and stated the range to include Aru, Waigeo, mainland New Guinea and its varying outlying islands as far south east as Australia. He went on to specify a number of island localities. Therefore according to Parsons, two "species", H. phorbas and H. dictaea, occur on Waigeo as well as mainland New Guinea. However, we consider that phorbas and dictaea are conspecific and that only one subspecies, H. p. silo, occurs in political Maluku, on Aru, its type locality and on mainland Papua. As Aru is separated from mainland New Guinea only by shallow water, and may well have been directly connected at the surface during the last glaciation, it can be regarded biogeographically as part of Papua. Parsons did not locate the holotype female of dictaea although it is deposited in BMNH. We have examined this specimen along with a series of female specimens in BMNH from Waigeo and it is clear that they all have undersides that are significantly paler in ground colour and weaker in the post discal striae than the underside of the holotype female of silo (also at BMNH).
We have also examined females, whose undersides are dark and therefore match that of the silo holotype, from Papua mainland, Aru, Roon, Biak and Yapen in BMNH and the collection of the 2 nd author.

Hypolycaena phorbas silo
Within Maluku this subspecies is only found on Aru Islands. We have examined five males and five females from Aru. All display the typical silo phenotype with the exception of one female from Wokam (Figs 35,36), which has a slightly lighter underside than the other four.

Other phorbas material examined, from outside Maluku.
Undersides of series of males from Batanta, Papua mainland, Aru, Roon, Biak and Yapen are all of the darker form matching the holotype female silo underside. The three males and three females from Misool in BMNH all have the paler undersides matching the dictaea type specimen. There is one male from Salawati in BMNH whose underside is of this form.
We therefore consider that H. dictaea is not a separate species but is a subspecies (or possibly just a form) of H. phorbas occurring on the islands of Waigeo, Misool and possibly Salawati, which all lie to the west of mainland New Guinea. We consider that the taxon present in Papua mainland, Aru, Roon, Biak and Yapen is H. phorbas silo.
Batanta is a new distribution record for H. phorbas. We have examined four males collected in October 2009 on the South Coast of the island. These all have the darker underside pattern. In the absence of females we prefer not to assign subspecific status.
We also make the following comments on Parsons' suggested wider eastern distribution of dictaea, in the sense that he uses that name. The males from Waigeo, Misool, Batanta, Salawati, Papua mainland, Aru, Roon, Biak and Yapen, all localities within the western part of the species' range, share the same shade of dark blue upperside. These contrast with the more purple colour of nominate phorbas and a number of unnamed specimens in BMNH from the eastern islands of Papua New Guinea including Yule, Woodlark and Kiriwina (= Trobriand Islands).
In addition these more purple males have much darker undersides than the dictaea type. The origins of these un-named specimens match many of the localities given by Parsons included in his distribution of dictaea.
We believe he mistakenly included these together with Waigeo specimens in his provisional assessment of dictaea. As this latter, more purple, group is beyond the geographical scope of this article, we do not describe any of these specimens further, but await a more comprehensive revision of the genus. Nevertheless, the clarification herein of the status of the Maluku fauna should aid in such a revision.

Hypolycaena sipylus (Felder, 1860) (TL: Ambon)
H. sipylus is widespread in Indonesia as well as occurring in the Philippines and New Guinea region (Rawlins 2007). It is the Type Species of Hypolycaena, although little is known of its early stages. In Maluku there are three recorded subspecific taxa. Fruhstorfer, 1912 (TL: Sulawesi). Figs 43-46.

Hypolycaena sipylus numa
H. sipylus numa occurs on Wetar Island (Rawlins 2007) within S. W. Maluku as well as along the Lesser Sunda chain.

Hypolycaena danis danis
BMNH also has one male and three females from Obi which match this taxon. However there is a second male labelled Obi which is typical of the nominate subspecies from Halmahera. The specimen bears two labels: 1. "Obi, ex J. Waterstradt, 1904". 2. "Ex Oberth Coll, Brit. Mus. 1927 Without further males to examine it is hard to draw a conclusion from this, but based on the other four Obi specimens we include these within danisoides. Therefore we extend the range of H. danis danisoides to include Obi, Seram and Ambon as well as Kei. (Hewitson, [1878] D'Abrera records distribution as: "Aru (?) Papua and islands of Louisade Archipelago". We assume he intends the "(?)" to refer to Aru, although Hewitson states the holotype to have been collected in Aru by Wallace. Although we could find no specimens from Aru in BMNH, M. Nagai (pers. comm.) says his son, K. Nagai, has collected three males and five females in Aru, confirming the type locality. This subspecies also occurs widely on the island of New Guinea including both Papua New Guinea (Parsons) and Papua mainland (Timika -1♂ vi/2002, Nabire -1♀ ii/2003, 1♀ iii/2003, Fak Fak -iv/2003).

Discussion
Vane-Wright and Peggie (1994) conclude that the fauna of Central Maluku (Buru, Ambon, Seram, Seram Laut) is most strongly related to New Guinea and to Sulawesi plus the Philippines. The distribution of H. asahi in Ambon and Seram, with two similar species in the Philippines, conforms to this pattern and supporting evidence for the theory. It also suggests that a closely-related species might occur in Sulawesi. H. umbrata Seki and Takanami (1988) is a strong but not quite conclusive candidate. It shares with asahi, shirozui and toshikoae the larger hindwing tornal orange spot and deeply conjoined valvae, although the outer edges of the valvae are noticeably excavate with a sub-apical point. These four taxa might be shown in future to constitute the shirozui species group, but their monophyly is as yet uncertain.
Our extensive study of the phorbas species group taxa from Maluku, Papua Mainland and the islands of West Papua has clarified the status of the taxa dictaea and erasmus. This new information, when combined with further study of the related specimens from the islands to the East of Papua Mainland, should provide valuable evidence about the biogeography of the island arc from North Maluku to the East of Papua New Guinea and confirm the apparent monophyletic status of the species group.

Conclusions
Examination of the male confirms the specific status of H. asahi which is now recorded from Seram as well as the type locality Ambon. The species of Hypolycaena most closely resembling asahi occur in Sulawesi, Mindanao and Mindoro.
There is no confirmed record of H. periphorbas (= H. erasmus) occurring in Maluku. The distribution of this species remains extralimital, to the East of the region studied.
We synonymise H. erylus pigres Fruhstorfer, [1912], with H. erylus thyrius Fruhstorfer, [1912], the latter having page priority. The low number of specimens of H. erylus available at this time from South and South East Maluku, especially of females, makes determination at subspecific rank for those islands speculative. The