heteropanini

Yen et al. (2005) stated that the only included genus in their system, Heteropan Walker, was not chalcosiine despite general similarity to Arbudas Moore and allies. However, in their analyses, it was consistently placed as sister to the rest of the Chalcosiinae, so could logically be treated as a tribe thereof for the present, e.g. as by Tarmann (2004: 35), but see comments on the generic account with reference to immatures of Heteropan. It is likely that the elucidation of the relationship of Heteropan to the Chalcosiinae would be facilitated by more detailed investigation of the immature stages with reference to the definitive features of clades within the Callizygaeninae-Chalcosiinae complex.

The tribe currently includes (Yen et al., 2005) only Heteropan, excluding the other genera of the concept of Alberti (1954), but Yen (2003) noted diversity in structure of the genitalia and androconial systems, and suggested that the genus may therefore require subdivision.

The species are all rather small and delicate. The features of the thorax shared with the Chalcosiinae have been mentioned on the account of the subfamily, but it is the head that provides the only unique features for the tribe and also further parallels with, or discrimination from, the Chalcosiinae. The upper part of the frontoclypeus is projected as in many Chalcosiinae, a character, together with a crested vertex, suggested by Epstein et al. in Kristensen (1998) to be diagnostic for the Chalcosiinae, but noted by Yen (2003) to be unreliable. The tongue is relatively

long within the family, and the labial palp has the terminal segment longer than the second; it is equal or shorter, and the segments may be reduced to two or one segment in the Chalcosiinae proper. The pilifer is much reduced, rudimentary.

The only unique features for the group noted by Yen et al. (2005) are in the antennae, where the flagellomeres are marginally ridged and strongly keeled. The antennae are not pectinate.

The male genitalia are diverse in structure, though those of the generic type species have a relatively short and broad uncus, with a pair of small spines ventrally near the apex; the other three Bornean species have a slender, tapering distal section to the uncus, and there is a slender, digitate, apically acute process from the base of the uncus, and the valve; this is largely fused to the valve in the type species apart from a free, more sclerotised apical spine (Fig 28).

The cornuti of the aedeagus vesica, when present, can be slender, threadlike spines, a feature also seen in Cyclosia Hübner. In the female, the pseudobursa is absent as in outgroups, whereas it is present in the Chalcosiinae proper. The type species (Fig 27) has a moderate ductus leading into a rather elongate corpus bursae with a large curved spine near its base. In another Bornean female dissected that belongs to the typical group (Fig 30), the ductus bursae has a short but slightly broader antrum and then a long narrow section that expands gently to curve into a more or less spherical corpus bursae set somewhat asymmetrically on it; there is no signum.

Horn-like signa are sometimes present as in some outgroups, but these occur also in “Pidoros” gemina Walker, as a Chinese outlier of an otherwise Australasian clade of the Chalcosiinae. The functional ovipositor is shared with Chalcosiinae (Yen, 2003), but is not strongly extensile.

No early stage data were included for Heteropan in the analysis of Yen et al. (2005). These are partially available for H. scintillans Walker, the type species of the genus (T.R.D. Bell, MS). However, it is not possible to extract from this account the critical characters of the larva and pupa. The larva has a retractile head. The body is flat beneath and convex transversely above. The head is light orange as are T1 and T2, but the rest of the body is shining, light bluish grey, with a large bluish black patch laterally incorporating the dorsolateral and supraspiracular tubercles. The segments are constricted where they join, and each one has dorsolateral and supraspiracular verrucae with many radiating black hairs or bristles. Below each spiracle is a further tubercle with radiating off-white hairs. Prolegs are present on A3-6.

The larvae are found on the underside of the leaves of host plants. The pupa is enclosed in a flat, oval, disc-like, pure white cocoon made along the midrib of the underside of a leaf or elsewhere; it is densely woven, with a few tiny balls of white silk stuck onto the outside, and with a slight flange all round. Bell did not describe the pupa. Comments on adult behaviour may be found under H. leesi sp. n. below.

The host plants recorded by Bell were Phrynium (Marantaceae) and Curcuma (Zingiberaceae), both monocotyledonous genera in the Zingiberales (Mabberley, 1987).

The description of the cocoon is consistent with placement of Heteropan within the Callizygaeninae-Chalcosiinae clade. The presence in the larva of secondary setae on verrucae is more characteristic of other zygaenid subfamilies than of Callizygaeninae and Chalcosiinae, where most taxa have primary setae only (p. 33). However, secondary setae are known from genera in Clades 1 and 5 (Yen et al. 2005: fig 57) within the Chalcosiinae: Aglaope and Arbudas respectively.

The genus Heteropan ranges from the Indian Subregion through the Indo-Australian tropics, though excluding Australia. There are at least five species in New Guinea and an outlier further east in Vanuatu and Fiji, H. dolens Druce (Robinson, 1975).