Chalcosiinae

Endo & Kishida (1999) have published an illustrated guide to most of the Chalcosiinae of south-east Asia, with supplementary material in Kishida & Endo (1999). Kishida noted that it is difficult to collect chalcosiine zygaenids in S.E. Asia because their seasonality is not well understood, and many fly in the forest canopy (Endo & Kishida, 1999). This has also been the field experience of the author and H.S. Barlow (pers. comm.); most of the species covered here are illustrated by old material.

The analyses of Yen et al. (2005) indicate that the monophyly of Chalcosiinae is strongly supported with or without inclusion of the Heteropanini (Heteropan Walker), though few of the characters involved are concerned. However, no data for Heteropan immatures were included, yet the seven characters indicated (Yen et al., 2005: 252, 272) as unique to the Heteropan + Chalcosiinae clade include one larval (retractile cervical gland on T1) and four pupal characters (thoracic spiracles not concealed; abdominal tergites with several rows of spinules; a transverse tuberculate band on A10). Therefore these could also be considered as unique to the Chalcosiinae. The two adult characters shared uniquely between Heteropan and the Chalcosiinae proper are of the thorax: patagia strongly sclerotized and sac-like; and more developed than the parapatagia.

One of the characters uniquely supporting the monophyly of the Chalcosiinae proper is also of the patagia, where the scaling is bent downwards. The other character is at the anterior margin of the second abdominal tergite, where a transverse row of four syndeses (elongate, transverse, translucent windows) is distinctly submarginal rather than marginal.

Unfortunately the character list in Appendix 4 for the critical node defining the Chalcosiinae in Yen et al. (2005) is confusing because the character numbers are not sequential and appear to contain several typographical errors such as a zero character. This prohibits exploration of them in the list of Appendix 2 and the matrix of Appendix 3. For this, one is restricted to Appendix 5 where some additional features may be of value: the second segment of the labial palp is distinctly shorter than the basal one and is oval in shape; the antennae are usually (but not always) bipectinate in both sexes, the rami longer in the male and platelike. Neither of these character states is seen in Heteropan, however.

Tarmann (1992a, 1994) and Naumann et al. (1999) suggested that an unusual male androconial organ involving bristles from the inner base of the hindwing and a pleural pouch or lateral fold on each side sternally of the second abdominal segment provided an autapomorphy for the Chalcosiinae, together with lack of accessory glands at the oopore and an elongated ovipositor. Tarmann (1992a) also listed specialism in the chemical defence system of the larvae. Yen et al. (2005) indicated that occurrence of elongation of the ovipositor was inconsistent within the subfamily, although present in the majority of taxa. Further surveys of larval defence systems through the family as a whole have shown these are complex and do not provide such clear taxonomic indicators as had originally been hoped.

Tarmann (1992a) described the surface of the pleural pouch as covered with a large number of setae which he suggested were associated with pheromone glands. The bundle of bristles on the hindwing is S-shaped and can be inserted into the pleural pouch; it arises from the anal axillary sclerite at the base of the hindwing. Yen et al. (2005) recorded complex variation on this theme, with the occurrence of long hair bristles, rather than modified scales, at the base of the hindwings, together with a doubly folded pleural pouch as being definitive for the Chalcosiinae.

The Chalcosiinae are by far the most diverse zygaenid subfamily in the Oriental tropics and exhibit a similar diversity of form, particularly wing shape, and facies, including many instances of mimicry. The Chalcosiinae contain the largest species of the family, sometimes, but not always, combined with robustness, and those with the largest wing area, the wings being much deeper than in other subfamilies. However, a few genera, such as Phlebohecta Hampson, contain narrow-winged species, and some of the deeper winged genera are also very delicate, such as Heteropan and Arbudas Moore.

The deeper wings may well be associated with the development of branching systems among the radial and medial sectors of the forewing. R1 and R2 arise independently from the cell and run anteriorly to the costa; only very rarely is R2 a part of the radial branching system (Yen et al., 2005: fig 18B). The branching system is most frequently (R3, R4) R5) but occasionally (R3 (R4, R5)), the latter state occurring most frequently in the more basal clades in the analyses of Yen et al. (2005): Heteropanini; Agalopini; Cyclosiini; the majority of the Arbudas-complex of Tarmann (1992); but also Psaphis Walker in Clade 17. M2 and M3 can also be stalked, but this feature is similarly widely scattered, as in: some Agalopini; Corma Walker in the Cyclosiini; Amesia Duncan in Clade 12; Psaphis in Clade 17. The course of these distal veins is often gently sinuous, again throughout the subfamily.

There are also trends in the structure of the male genitalia, with the Chalcosiini tending to have the uncus shorter, broader and often bilobed compared with the Cyclosiini and Heteropanini, where it is narrow, tapering, undivided, and the Agalopini where it is also mostly undivided (Achelura Kirby is an exception). Within the Chalcosiini there is a tendency with progression from the basal clades to the distal clades in the analyses of Yen et al. (2005) for the aedeagus more often to be short, strongly sinuous and to have a medioventral protuberance. These features correlate to some extent with increasingly strongly modification of the eighth abdominal segment, particularly the sternite, which is often enlarged and has prominent lateral arms. Such features are strongly developed in virtually all the clades from 11 to 18.

The larvae are often highly aposematic, but mostly have primary setae only, even if set on prominent verrucae. The setae can be plumose, e.g. in the Agalopini. Secondary setae occur in Aglaope Latreille, the only European (and N. African) chalcosiine genus, Heteropan and Arbudas.

The host plants recorded for the Chalcosiinae are diverse, and have been collated by Tarmann (1992b). There is some frequency of records in Rosaceae for the Agalopini and in Euphorbiaceae for Cyclosiini and some Chalcosiini. The Heteropanini may favour Zingiberaceae. The Chalcosiini have been recorded from a wide range of families, particularly Lauraceae, Melastomataceae, Myrtaceae, Symplocaceae and Theaceae.

The Chalcosiinae, apart from the Mediterranean Aglaope, range through warm temperate and tropical regions of southern and eastern Asia through the Indo-Australian archipelago to the Solomons. However, they are absent from Australia, where only Procridinae occur (Tarmann, 2004). About three-quarters of the genera are restricted to the Oriental Region and east Asia. A small number of genera is restricted to the Philippines and/or Sulawesi, mostly containing very few species. A slightly larger number of genera is found further east, most restricted to New Guinea, but some having outlying species in the Moluccas, Sulawesi or the Philippines. Only Heteropan and Achelura Kirby extend from India to New Guinea.

The more easterly genera occur mostly in four of the numbered clades in the analysis of Yen et al. (2005: fig 57, reproduced in an abridged form here in Fig 1): 6 and 7; 17 and 18. If coloration characters are excluded (and therefore the tendency to mimicry that is developed as much or more in the New Guinea genera in comparison with Oriental ones), then clades 6 and 7 come together (Yen et al., 2005: fig 64). This combined clade may therefore represent an early radiation of nine of the eleven genera endemic to, or centred on New Guinea, though the clade also includes one monobasic genus endemic to the Moluccas (Sciodoclea Jordan), one small Sulawesi genus (Hemichrysoxyptera Roepke) and an extreme outlier, “Pidorust” gemina Walker from southern China. There is no representation of this clade in Sundaland unless gemina genuinely occurs in Borneo. Hemiscia Jordan is split between clade 7 and clade 10, the latter with further Pidorust species.

Clades 17 and 18 are the terminal ones in the analysis, with the sequence of clades 6 and 7 up to Clade 17) all consisting of Oriental genera. Clade 17 contains the genus Eusphalera Jordan with over 20 species in New Guinea, two in the South Moluccas and two, in different species groups, in the Philippines. Docleopsis dohertyi Rothschild from Sulawesi is placed as sister to Eusphalera (Yen et al., 2005: fig 57), and this clade in turn is sister to the genus Psaphis Walker that has several species in Sundaland and extends to the Philippines and Sulawesi.

Clade 18 includes one major subclade that shows a progression from west to east in its branching: (Sundaland (Philippines (Sundaland (Philippines (Sulawesi, New Guinea + Moluccas))))). It brings in several of the small Wallacean genera such as Boradiopsis Hering and Allocaprima Hering (Philippines), Euxanthopyge Hering (Sulawesi) and Aphantocephala Felder (New Guinea, with one Moluccan species), and also species of Docleopsis Jordan, Phlebohecta Hampson and Trypanophora Kollar that are assigned by the analysis to ‘geographically sensible’ positions. These three genera are some of several that are not recovered in their traditional composition by the analyses of Yen et al. The most distal representation of Sundaland in the clade above consists of the “Anarbudas” bipartita Walker group present in Borneo and indicated by Yen et al. to need transference to a new genus, effected here by transfer to Boradiopsis.

Some of the more basal clades are represented in New Guinea, such as the Heteropanini and the Agalopini (Clade 1). In the latter, Achelura Kirby has a small group of New Guinea species that is placed with the Oriental members of the genus in the analysis, and the monobasic New Guinea genus Atelesia Jordan is placed in a clade with mainland Asian and Taiwanese taxa. This interesting distribution was discussed also by Yen (2003).

The two remaining genera endemic to the Philippines, Cryptophysophilus Hering and Docleomorpha Hering, are grouped with Heterusinula Bryk and typical Anarbudas Jordan, then Corma Walker and Eucormopsis Jordan within Clade 2.

As indicated above, several traditionally defined genera have their component species scattered widely through the clades in the analysis of Yen et al. (2005: fig. 57). However, very many other species-rich genera are recovered in more or less their traditional composition. This tends to support the assumption that such ‘dispersed’ genera are genuinely para- or polyphyletic as traditionally constituted, and therefore that they require revision. Some more extreme examples, with the clades (those with Bornean species are in bold) to which their species have been scattered, are: Pidoras Walker (7, 8, 9, 10, 12, 16); Phlebohecta (15, 18 (two subclades)); Trypanophora (15, 18 (two subclades)); Docleopsis (17, 18 (two subclades)); Eterusla Hope (13, 14, 18).

The order of genera and higher taxa in the systematic section following is based on that in Appendix 1 of Yen et al. (2005) which follows to a large extent the branching sequence from Clade 1 to Clade 18 in their fig. 57 (Fig 1). Comments will be made on each clade in the description of the first component genus thereof to be treated.