Eublemma versicolor Walker

Autoba versicolor Walker, [1863] 1864, J. Proc. Linn. Soc. (Zool.), 7: 57.

Thalpochares rubricosa Snellen, 1880, Tijdschr. Ent. 23: 64.

Diagnosis

See brachygonia and indigofera. There has been much confusion over the identity of versicolor in the past, and many records probably are of perversicolor sp. n. described below. True versicolor is a pinkish violet-grey in ground colour with reddish brown fasciation and margins. The holotype (OUMNH: 1352) is strongly darkened basal to the straight medial fasciae, therefore it and a more typical specimen from Sulawesi are illustrated. It is relatively large compared to the previous few species and perversicolor. It is one of a major complex of species of this facies type as discussed in the note following, and this complex can be identified in the male genitalia by the elongate, narrow valves with a small, tongue-like, setose saccular process close to the ventral margin, but usually accompanied by a slight angle to the costal margin opposite. The valves expand to this position from the base, then taper away again to a rounded apex. The female genitalia have the ostium flanked by a pair of finely rugose pads rather than a distinct pocket; the ostium extends posteriorly along a narrow ventral declivity through the eighth sternite. The corpus bursae is an elongate pyriform on a relatively short, coiled neck and contains two small unequal spines*.*They vary in size slightly through the range of the complex but show no marked differences.

Taxonomic note

The species typifies Autoba, the genus in which it was placed by Poole (1989) and Edwards in Nielsen et al. (1996). The versicolor subgroup of species ranges from the Indian Subregion to the Solomons and Australia. In the Indian Subregion the group is represented by angulifera Moore, where there appears to be some sexual dimorphism with males having a fawn ground colour and females a pinkish grey ground. The taxa grisescens Warren and ochreirufa Warren are probably synonymous with angulifera, based on female and male material respectively. The original description of grisescens included subspecies lilacina Warren from Australia and vinosa Warren from New Guinea and eastwards. Roepke (1916) separated lilacina as a good species from grisescens (angulifera) as discussed in the section on biology below. Edwards in Nielsen et al. (1996) established its synonymy with loxotoma Turner, and also included vinosa with this, though vinosa is generally smaller and more darkly marked. The genitalia of loxotoma (slide 20922) are somewhat more robust than in grisescens (slides 20774, 20803), but otherwise very similar with the valves slightly broader. Poole (1989) treated pallescens Warren (Woodlark I., New Guinea, Sulawesi) as distinct, but it is retained as a synonym by Edwards in Nielsen et al. (1996). Given the complex situation just outlined, it is probably distinct.

Geographical range

Borneo, Peninsular Malaysia (Barlow colln), Philippines, Sulawesi; the species illustrated for Thailand by Kuroko & Lewvanich (1993) has males concolorous with females so may also be versicolor.

Habitat preference

The holotype of versicolor was taken in Sarawak by A.R. Wallace, probably in the lowlands. The species has not been reliably recorded in recent surveys.

Biology

Robinson et al. (2001) listed numerous host plants, with the larva feeding variously on flowers, pollen, fruits, shoots and leaves:

• Anacardium, Buchanania, Mangifera (Anacardiaceae)
• Ceiba, Durio (Bombacaceae)
• Baccaurea (Euphorbiaceae)
• Bauhinia, Canavalia, Cynometra, Moullava, Senna (Leguminosae)
• Loranthus (Loranthaceae)
• Lagerstroemia (Lythraceae)
• Memecylon (Melastomataceae)
• Eugenia, Psidium (Myrtaceae)
• Nelumbo (Nelumbonaceae)
• Macadamia (Proteaceae)
• Ziziphus (Rhamnaceae)
• Dimocarpus, Lepisanthes, Nephelium (Sapindaceae)
• Theobroma (Sterculiaceae).

However, many of these records probably refer to other members of the versicolor complex and also to perversicolor. A specimen from the Philippines in USNM that is probably versicolor was reared from Ternstroemia speciosa (Theaceae) (S.E. Miller, L. Helgen, pers. comm.). Descriptions of the biology of some other members of the group follow.

Roepke (1916) identified and described the life history of a species in Java as lilacina (previously identified as versicolor), promoting this to full species status. The larva is typical of the genus, reddish or greenish grey, with a light brown, rounded head. The primary setae are small. It feeds on the rind of the fruits of cocoa (Theobroma cacao; Sterculiaceae) and nam-nam (Cynometra cauliflora; Leguminosae), these growing in a similar manner from the trunks and branches of the trees. The larva eats grooves in the rind that become covered in frass. Pupation is on the fruit or on a stem in a robust cocoon that is also adorned with frass. Occurrence of the larva is positively correlated with that of the mealybug Planococcus lilacinus Cockerell (Pseudococcidae, as Pseudococcus crotonis Green; see Williams (2004: 641) for an account of this species), and the larva may also consume these secondarily, perhaps more so in earlier instars.

Aiyar (1943) described eggs and larvae attacking mango (Mangifera) in India that may be referable to angulifera. Bell (MS) also described the larva, but recorded a different suite of hosts. The eggs are pale red, hemispherical, ribbed, laid singly on the flower buds of mango. The larva is smooth, spindle-shaped, its body pale green with the spriacles finely ringed with black, and its head and prothoracic shield are orange or yellowish brown. There are prolegs only on A5 and A6, and these are short and weak. It constructs silk-lined galleries on the main stem of the flower spikes of mango and other host plants. The slightly oval cocoon is formed at the base of the flower spike under a leaf and incorporates frass. Bell recorded Buchanania, Canavalia, Memecylon and Wagatea as hosts, observing that the larva fed inside the pods of Canavalia. When flower-feeding, the larvae appeared to be attended by red Oecophylla ants, but Bell suspected that these were attending coccids that the larvae might also prey on.

Genitalia: