DID A MEMBER OF THE VANESSA INDICA COMPLEX (NYMPHALIDAE) FORMERLY OCCUR IN NORTH AMERICA?

The North American Oligocene fossil Vanessa amerindica is thought to be most like V. indica. Based on an 18th century painting made by the English naturalist Henry Seymer Jr., the possible existence of a member of the indica complex that occurred in North America as late as circa 1770 is demonstrated. New information on the classification of the nine extant species currently included in Vanessa sensu stricto strongly suggests that this apparently undescribed species is most closely related to the Atlantic Islands endemic, V. vulcania. Three competing scenarios that attempt to explain the highly disjunct distribution of the species that make up the V. indica complex are discussed, and it is concluded that the genus Vanessa most probably originated in North America, and that V. vulcania represents a separate, Atlantic colonisation event, separate from the Pacific colonisation event that gave rise to the Asiatic V. indica-group. This implies that, contrary to earlier hypotheses that sought to explain the distributional gap between the Canaries and India, the indica complex may never have been established on the western Palaearctic mainland, or in the Eremic Zone (Morocco to Somalia and Tien Shan). An African species formerly placed in Antanartia is formally transferred to the genus Vanessa (Vanessa abyssinica, comb. nov.). Additional key words: Tertiary, relict, introduction, biogeography, Newfoundland, Macaronesia, Henry Seymer. According to Art Shapiro (1992a), “The most peculiar Palearctic butterfly distribution is that of the Indian Red Admiral, Vanessa indica (Herbst).” Although the taxonomy of this member of the Nymphalidae has since changed — V. indica (Herbst, 1794), in Shapiro’s sense, is now divided into three allopatric species — the enigma to which he referred remains: how can the 7000 km gap in distribution between the Indian red admirals found on Madeira and the Canary Islands, and those occupying the rest of the range, in the Oriental region and far eastern Palaearctic, be explained? No extant member of the species complex to which these butterflies belong is known from the rest of the western Palaearctic, North Africa, or North America. Fresh interest in this problem has come from the recent discovery of an 18th century painting of a member of the Vanessa indica complex, supposedly based on a specimen collected in Newfoundland circa 1770. This illustration is one of some 300 surviving images of exotic butterflies and moths made by the little-known British naturalist Henry Seymer (1714–1785), together with his son, Henry Seymer Jr. The Seymers obtained their natural history specimens, notably of molluscs and insects, through dealers, travellers, military personnel and other contacts. Based on an extensive analysis of all the known Seymer Lepidoptera paintings, and their notes and records, it has been demonstrated that the vast bulk of their exotic material came from China, Java, India, West Africa, South Africa, South America, Jamaica, and the early British colonies in North America (Vane-Wright & Hughes, 2005: table 1, p. 254). Their Lepidoptera collection apparently totalled some 20,000 specimens, but it was dispersed immediately after Henry Sr.’s death, and nothing is yet known to have survived (Vane-Wright & Hughes, 2005; Barker & Vane-Wright, 2007). The Seymer paintings, made during the period 1755–1783, appear to have been intended as a virtual record of the collection. If so, it is fortunate they had such foresight. The level of accuracy achieved ranges from good to outstanding. Minute detail is often finely rendered, and the coloring remains authentic in all but a few instances (Vane-Wright & Hughes, 2005). The lack of degradation of tint that might be expected in such old watercolors is no doubt a consequence of the fact that, through the intervening years, the pictures were rarely on show, and were evidently preserved in library conditions. The idea that a member of the V. indica complex recently occurred in North America is so surprising that, without a specimen and independent verification, considerable doubt must be accepted—although we 200200 JOURNAL OF THE LEPIDOPTERISTS’ SOCIETY believe the case for authenticity (presented below) is good. However, whatever the final conclusion regarding the painting, it has stimulated us to review the “Vanessa indica problem” and, in turn, challenge two previous hypotheses regarding the biogeography of these butterflies, and support instead an alternative hypothesis in which North America plays a key role. Phylogenetic Relationships of the Red Admiral Butterflies The type species of Vanessa Fabricius, 1807, Papilio atalanta Linnaeus, 1758, is the familiar Red Admiral butterfly. In his major revision, Field (1971) placed five species of red admirals in the genus: V. atalanta (L., 1758), V. tameamea Eschscholtz, 1821, V. indica (Herbst, 1794), V. dejeanii Godart, 1824, and V. samani (Hagen, 1895), but more species are now recognized (see below). The two other main species groups usually included within the genus are the painted ladies (placed by Field in the genus Cynthia; type species Papilio cardui) and the antipodal admirals (placed by Field in Bassaris; type species Papilio itea). DNA sequence data (Wahlberg et al., 2005) suggest that within Vanessa and contrary to earlier phylogenetic work based solely on morphology (e.g. Craw, 1989, Holloway & Nielsen, 1999), the red admirals (Vanessa sensu stricto) have a sister-group relationship with Vanessa abyssinica (Felder & Felder, 1867), a montane butterfly from East Africa. Previously V. abyssinica was treated as a member of the endemic Afrotropical genus Antanartia Rothschild & Jordan, 1903 (e.g. Howarth, 1966; Ackery et al., 1995). Superficial comparison indicates that this new arrangement is credible: all five remaining Antanartia, including the type species, Papilio delius Drury, 1782, have distinct hindwing tails at vein M3, whereas V. abyssinica does not, looking instead rather like a small and drab red admiral (Fig. 1a). The work of Nakanishi (1989) on the early stages of abyssinica is consistent with this placement, as it has a peculiar setal arrangement in the first larval instar otherwise known only from Vanessa, and it shares the habit, in later instars, of making a nest by tying both edges of a leaf together with silk. Nakinishi recorded that neither Antanartia schaenia nor A. hippomene exhibit these Vanessa characters. However, as also pointed out to us by Thomas Dimock, adult abyssinica are highly distinctive compared with all other Vanessa s.s. Notably, at least two of the hindwing ocelli always have blue pupils (invariably black in other Vanessa s.s.); the hindwing marginal band widens at cell M1, lacks any black submarginal spots (as seen in an aberration of V. atalanta: Frohawk, 1938: p. 86), and continues anteriorly into cell R5 (unlike other Vanessa s.s.); and hindwing vein Sc+R1 is relatively elongate, giving the wing a unique, almost square aspect. All of these differences can be seen as autapomorphies, except the first, which may be a symplesiomorphy (e.g. this condition is frequent in subgenus Cynthia, in V. (C.) cardui, for example, being referred to as form Fig. 1. (on facing page) The species of the genus Vanessa sensu stricto. Left halves show upperside, right halves corresponding underside. All figures (with exception of b) have been brought to the same forewing length to facilitate comparison; information on actual size is included with each separate legend. With the exception of b and h, all images are based on specimens in the Natural History Museum, London (BMNH); fw-l. = forewing length. a, V. (Vanessa) abyssinica abyssinica (Felder & Felder, 1867), male [Ethiopia: Mt Zuquála, over 9000 ft, 25–27.x.1926, H. Scott; BM1927-127; fw-l. 21 mm.] [Howarth, 1966: 31, indicates a range of 17–22 mm for male V. abyssinica, and 20–24 mm for female]; b, V. (Vanessa) sp. nov. (V. vulcania-group), female? [Newfoundland, ca 1770; from ‘profile’ image made by Henry Seymer Jr., ca 1773; VaneWright & Hughes, 2005: 164/5; fw-l. estimated at 35 mm — see text]; c, V. (Vanessa) vulcania Godart, 1819, female [Spain: Canary [Islands], iv.1885; Leech Collection, BM1901-173; fw-l. 34 mm] [Field, 1971: 24, gives male 26–32 mm, female 29–33 mm, but there are larger and smaller examples in the BMNH collection — see text]; d, V. (Vanessa) indica indica (Herbst, 1794), male [China: Siao-Lou, 1900; Oberthür Collection, BM1927-3; fw-l. 33 mm] [Field, 1971: 21, gives male 25–34 mm, female 27–37 mm]; e, V. (Vanessa) indica pholoe (Fruhstorfer, 1912), male [SW India: Anamully Hills, 3000–4000 ft., Davison; Godman-Salvin Collection, BM1903-4; fw-l. 30 mm] [Field, 1971: 22, gives male 27–29 mm, female 28–30 mm]; f, V. (Vanessa) indica nubicola (Fruhstorfer, 1898), male [Sri Lanka: Newara Eliya, vi.1921, W. Ormiston; BM1922-315; fw-l. 27 mm] [Field, 1971: 22, gives male 26–31 mm, female 30–33 mm]; g, V. (Vanessa) buana Fruhstorfer, 1898, male [Indonesia: S Sulawesi, Bonthain, 5–7000 ft., x.1895, A. Everett; Rothschild Bequest, BM1939-1; fw-l. 24 mm] [Field, 1971: 23, gives male 27.5 mm; assuming figures in Tsukada, 1985: 82, are life-size, females are ca 25–27 mm]; h, V. (Vanessa) dilecta Hanafusa, 1992, male [Indonesia: W Timor, Mt Mutis, v.1992; Hanafusa Collection] [Hanafusa, 1992, gives male fw-l. 27.5–30.5 mm, female 30–31 mm.]; i, V. (Vanessa) samani (Hagen, 1895), male [Indonesia: SW Sumatra, Danan Bento Morass, Ft. of Korintji Peak, 5000 ft., viii.1921, C.F. & J. Pratt, 7.22; Joicey Bequest, BM1934-120; fw-l. 23 mm] [Field, 1971: 27, gives male 25 mm, female 23 mm]; j, V. (Vanessa) dejeanii dejeanii Godart, 1824, male [Indonesia: E Java, H. Fruhstorfer; Fruhstorfer Collection, BM1937-285; fw-l. 24 mm] [Field, 1971: 26, gives male 23–26 mm, female 23–27 mm]; k, V. (Vanessa) dejeanii sambaluna (Frushtorfer, 1898), male [Indonesia: Lombok, Sambalun, 4000 ft., 1896, H. Fruhstorfer; Oberthür Collection,