Notodontidae

The Notodontidae are a family that has proved very popular with amateur entomologists over the past two centuries. The species are mostly moderate to large in size with biologically cryptic but aesthetically pleasing wing patterns. The wings, especially the forewings, are usually long and relatively narrow, as is the abdomen. The wing scaling is often coarse and the patterning never very crisply defined; the body and legs are usually clothed densely with longer scales giving the whole insect a rather shaggy appearance.

In collections specimens are prone to become greasy, suggesting a high fat content in the adult moth. In this they resemble the Cossidae and, to a lesser extent, the Lymantriidae.

The larvae are often highly modified into bizarre shapes, again primarily cryptic, but often with additional aggressive defences such as protrusible lashes on modified anal claspers.

There are 122 species known at present from Borneo, a fauna not as rich as that of Sumatra or the north-eastern reaches of the Himalayan ranges in Sikkim and Assam, but over twice as rich as that of Australia (about 50 species listed by Kiriakoff (1968)) and richer than that of Japan (about 105 species listed by Inoue (1956)). New Guinea has a fauna of about 100 species (Kiriakoff 1968) but a major portion of these are drawn from the predominantly endemic genera Omichlis Hampson and Cascera Walker, as well as from the largely montane genus Quadricalcarifera Strand. The Bornean species are here assigned to 65 genera. There are 16 endemic species.

The majority of species are virtually restricted to the lowlands but 22 range from the lowlands to about 2000 m and 29 are probably exclusively montane.

Twenty five new taxa, mostly species, are described. Past taxonomic work on the Oriental Notodontidae has often been rather superficial or careless so a considerable amount of revisional work involving synonymy and new combinations has had to be undertaken to try and introduce some stability into the nomenclature. A very high proportion of the type specimens involved are found in the British Museum (Natural History) and others have been studied on loan from the various Institutions listed in the Acknowledgements.

The taxonomic work involved the preparation of almost 300 slides of male, and sometimes female, genitalia. The male genitalia of all Bornean species are illustrated either here or by Holloway (1976, 1982).

Information on life histories and host-plants is very sparse for the Indo-Australian tropics and an attempt has been made here to collate what is available. Nevertheless, some interesting, possibly coevolutionary relationships between host plants and genera or groups of genera are already evident.

Adults

The Notodontidae have a metathoracic tympanal organ and are therefore in the superfamily Noctuoidea (Brock 1971) or placed in a separate superfamily, Notodontoidea, associated with the Noctuoidea (Common 1970, 1975). The former arrangement is perhaps preferable as the thoracic tympanal organ is one of very few really reliable apomorphic (derived rather than primitive) characters in the higher classification of the Ditrysian Lepidoptera.

Characters for defining the Notodontidae themselves are less satisfactory but there is general agreement that they retain a higher proportion of primitive, or plesiomorphic, characters than other families in the Noctuoidea (e.g. indications of median veins in the wing cells of Phalera bucephala Linnaeus), and hence may be the sister group to the rest. Both Brock (1971) and Common (1970) cited the possession of a metascutal tympanal bulla as a character defining the family, and Common also regarded the ventral direction of the tympanum itself as diagnostic, though this character is not infallible. In his key to superfamilies Common (1970) distinguished his Notodontoidea from the Noctuoidea using a character of wing venation: the base of vein 5 (M2) is usually nearer vein 6 (at the anterior angle of the end of the cell) than to vein 4 (at the posterior angle of the cell) rather than vice versa. Again this is not infallible.

Several characteristics of the male genitalia and abdomen may prove to be of value either in defining the family or in subdividing it.

Many species have lateral spurs to the fourth abdominal sternite that bear a fringe of flattened, blade-like setae*; the fourth sternite is also often distally bilobed, the lobes bearing more persistent scales or fine hairs marginally. This character is seen, for example, in Ambadra Moore, Caschara Walker (Fig. 90), Archigargetta Kiriakoff, Phalera Hübner, Brykia Gaede, Ginshachia sumatrensis Gaede (but not G. bronacha Schaus), Allata Walker and Saliocleta Walker (e.g. barasamphia Schaus).

The majority of genera have some sort of modification to the male eighth sternite, sometimes also the tergite, but there is great variety in these modifications and they cannot be categorised in terms of a typical ‘notodontid form’ though they may prove of value in subdividing the family or recognising groupings of genera.

The presence of deciduous stellate spicules in the aedeagus vesica is another character widely distributed in the family, e.g. in Phalera, Hyperaeschrella Strand, Brykia, Epistauropus Gaede, Euhampsonia Dyar, Teleclita Turner, Pseudoteleclita Kiriakoff, Disparia Nagano, Oxoia Kiriakoff, Notodonta Ochsenheimer, Suzukiana Sugi, Allodonta Staudinger and Hexafrenum Matsumura. In some instances the stellate spicules become asymmetric with one spine of each spicule being emphasised, the other spines fringing the base (when still attached to the vesica) as in Turnaca punctata sp. n. and, reduced to one massive spine at the end of a long, tube-like vesica, in Caschara punctifera Walker. In Ambadra and Fentonia Butler there are merely many simple, small, needle-like deciduous spicules. These can be of value in assigning males to females in certain species as a female that has been mated will retain them in her bursa copulatrix, and their size and shape may be diagnostic.

Almost all species have the gnathal processes separate, robust, prominent, and often complex. The uncus itself is usually modified to some degree, sometimes with elaborate socii.

A large number of species have the saccular area of the valve of the male genitalia membranous and somewhat corrugate, another character probably unique to the Notodontidae in the Noctuoidea. This character is seen in species of Euhampsonia, Calyptronotum Roepke, Chadisra Walker, Caschara, Allodonta, Hexafrenum, Hyperaeschrella, Disparia, Notodonta, Clostera Samouelle, Micromelalopha Nagano and Plusiogramma Hampson.

The forewing pattern is generally noctuoid with reniform and orbicular stigmata and simple postmedial and antemedial fasciae. The hindwings are usually unmarked except sometimes along the costa (Quadricalcarifera Strand and relatives, Stauropus Germar and relatives) or at the tornus (the Chadisra Walker and Teleclita Turner groups of genera). The presence of such hindwing markings may not be of great taxonomic significance but merely indicative of exposure of that area of the hindwing when the moth is at rest. As mentioned in the Introduction, the forewing patterns are cryptic, suggestive of tree bark or lichen; species with lichenous or bark-like patterns usually rest with the wings obtusely angled over the body. Other species, such as in the genus Phalera Hübner or in the Gargetta Walker group, rest with the wings scrolled round the body, causing them to resemble broken twigs.

In many species the dorsum of the forewing bears a central angular projection. Examples are Norraca lativitta Walker and the genera Suzukiana, Hyperaeschrella, Peridea Stephens, Notodonta, Hexafrenum, Caschara, Allata and allies, and Rodneya Kiriakoff.

The forewings and abdomen are usually long and narrow, the abdomen extending well beyond the hindwing margin. The antennae are usually bipectinate or pectinate in the male, but sometimes filiform with ciliae. The female antenna is filiform in the majority of cases, less elaborate than the corresponding male antenna.

Larvae

Gardner (1943) stated that the larval characteristics justified inclusion of the family in the Noctuoidea and drew attention to the resemblance of those notodontid larvae with normal prolegs to the larvae in the noctuid subfamily Acronictinae.

All four pairs of prolegs are present and nearly equal except in Gargetta Walker and allies, where the first, and sometimes the second, are reduced. The claspers have crochets in a uniordinal homoideous mesoseries (a single uniform band) and always with conspicuous secondary setae (Gardner 1943).

The generic characters are sharply definable and will no doubt lead to much revision of the current taxonomy when more life histories are known; they may even provide a basis for a rational subdivision of the family. Features include a range of form of the anal prolegs (or claspers) from vestigial through normal to slender and stematopodiform.

Stematopodiform claspers are often held out horizontally or erect and may contain protrusible lashes. They are seen in the stout, anteriorly squarish larvae of the Cerura Schrank group of genera and in the slender, cylindrical larvae of the Gargetta group of genera. The Australian species Pheressaces cycnoptera Lower (Turner 1903) and the Indian Fentonia tenebrosa Walker (unpublished notes by T.R.D. Bell) also have a *Gargetta-*like larva and the Indian species must therefore be misplaced in Fentonia Butler.

Another, perhaps natural group of taxa with modified anal prolegs has these filiform, vestigial, set on dorsoventrally flattened terminal abdominal segments, the latter resembling a leaf to some extent. The anterior part of the larva is also modified, usually by a number of dorsal humps. The development is at its most extreme in larvae of Stauropus Germar and Neostauropus Kiriakoff, but is also seen, with the leaf-like appearance most definite, in Teleclita and Pseudoteleclita and their African relatives in the genera Afroplitis Kiriakoff and Amyops Karsch (Pinhey 1975).

Many other genera have dorsal prominences of some sort and most are strikingly, if cryptically, patterned.

The literature on the early stages of Oriental tropical and Australasian species is sparse and scattered but an attempt is made in the main text to collate and precis what there is and assemble all the host-plant data. A major source of information was an unpublished typescript by T.R.D. Bell in the British Museum (Natural History) (BMNH), referred to in the text following as (Bell, MS), accompanied by watercolours of many of the species described, reared in South India. The adult material reared was deposited in the BMNH so it has been possible to check the identity of the species. Other sources are Barlow (1982), Bell (1935), Dodd (1902), Gardner (1943, 1946), Issiki (1969), Kalshoven (1981), Pant & Chatterjee (1951), Sevastopulo (1938-1947, 1940, 1949), Sugi & Nakatomi (1969) and Turner (1903).

* The cteniophores of Jordan (1923) who suggested they were associated with abdominal glands and and scent scales on the hindwings or legs.

Economic Importance

Few species are of economic importance though some are defoliators of fruit trees (Kalshoven 1981). Their role as defoliators of rain forest trees and seedlings has yet to be elucidated. They may have some value as an indicator group for monitoring rain forest type, and are frequently taken in light-traps set on the forest floor amid the understorey, though they are more numerous in samples made with the light overlooking the canopy.

Host-plant relationships

The references listed at the end of the discussion on larvae, together with Browne (1968), Pholboon (1965) and unpublished records in a card index maintained by Lepidoptera specialists of the Commonwealth Institute of Entomology, are the sources used in this survey of host-plant relationships and in the discussions of individual species later.

Several interesting patterns or relationships are becoming evident despite the sparseness of the data.

The majority of species defoliate dicotyledonous trees and shrubs but the Pydna Walker and Ambadra Moore groups of genera, together with Phalera combusta Walker (extending from India to Sumatra but not Borneo), feed on monocotyledons, mainly palms, bamboos and grasses. The streaky fawn, yellow or red forewing patterns of these groups reflect their feeding habit, paralleling the adults of other monocot-feeding species in other families such as the Mythimna and Sesamia groups in the Noctuidae, the crambine Pyralidae, and Psalis pennatula Fabricius in the Lymantriidae.

Several species are polyphagous over a range of plant families, such as Neostauropus alternus Walker and the Chadisra species from India. Species of the genus Phalera tend to be polyphagous but Oriental tropical species are found more frequently or exclusively (P. javana Moore) on Leguminosae.

A preliminary survey of host-plant relationships in the Indo-Australian macrolepidoptera has pinpointed a number of genera in other families that are restricted to Leguminosae such as, in the Ophiderinae (Noctuidae), the genera Lacera Guenee, Ericeia Walker and Rhesala Walker. The Notodontidae probably also include legume specialists, most conclusively the Allata Walker group of genera - Allata, Eguria Matsumura, Rosama Walker (Bell MS; Gardner 1943, 1946). Besida xylinata Walker has also been recorded from the family (Pholboon 1965).

Netria viridescens Walker would appear to be restricted to the Sapotaceae, feeding on several genera of the family such as Mimusops, Madhuca, Sideroxylon (Bell MS) and Manilkara (CIE records) in India.

The genus Dudusa Walker may prove to be restricted to the Sapindaceae, with D. synopla Swinhoe recorded from Schleichera in India (Bell 1935, as nobilis Walker), a record from Thailand on Nephelium also attributed to nobilis (Pholboon 1965) and D. vethi Snellen recorded from Nephelium in Sumatra and Java (Kalshoven 1981, as nobilis).

The genera Teleclita and Pseudoteleclita have only been recorded from Terminalia (Combretaceae), in India (Bell MS), the Philippines (original material of T. cathana Schaus) and Australia (Dodd 1902; Turner 1903). African relatives in Afroplitis, Amyops and Galona Karsch have similar larvae that feed on Terminalia and Combretum in the Combretaceae but also on Parinari in the Chrysobalanaceae and guava (Myrtaceae) (Pinhey 1975). Two chloephorine noctuid genera, Aiteta Walker and Westermannia Hiibner, also appear to be restricted to Terminalia in the Indo-Australian tropics but this plant genus is host to an unusually large number of Lepidoptera species, mostly polyphagous ones.

An interesting parallel in host-plant relationships is evident between the Clostera group of genera and the Cerura group; it is also shown by the nymphalid butterfly Phalanta phalantha Drury (Corbet & Pendlebury 1978). In the Palaearctic Region the host plants are Salicaceae, especially Salix and Populus, but in the tropics the majority of hosts are in the Flacourtiaceae, genera such as Casearia, Flacourtia, Scolopia (in Australia (Common 1963)) and Xylosma being recorded. In India both plant families have been noted as hosts and also in Hong-Kong (M. Bascombe in litt.). Terminalia is also involved in India (Brown 1968). Notocerura spiritalis Distant, an African member of the Cerura group with a characteristic larva, also has Flacourtia as a host (Pinhey 1975).

This relationship is less exclusive in Clostera than in the Cerura group. Records from a few other plant families have been made for the two major species groups in Clostera that extend into the Palaearctic. In the anachoreta-fulgurita group Palaearctic records are from Salicaceae. In India they are from both families (Browne 1968; Gardner 1943; CIE records) and also from Elaeodendron in the Celastraceae (Gardner 1943). There are no records from South-east Asia but the group has been reared from Flacourtiaceae in Hong-Kong (M. Bascombe in litt.). The anastomosis-restitura group is known from Salicaceae in the Palaearctic and Hong-Kong, from both families in India but also from Elaeocarpus (Elaeocarpaceae) there (Sevastopulo 1940). The Melanesian representative of the group, rubida Druce, has been reared from Terminalia in Papua New Guinea (CIE records). Southern African Clostera feed on Proteaceae (Pinhey 1975).

The two plant families are not taxonomically related according to Willis (1973), the Flacourtiaceae being linked with families such as the Passifloraceae, Euphorbiaceae and Tiliaceae rather than the Salicaceae.

There are several host-plant records for the Gargetta group of genera (Bell MS; Gardner 1943; Bascombe in litt.). Porsica ferreopicta Hampson, P.ingens Walker and Phycidopsis albovittata Hampson have all been recorded from Antidesma, Gargetta costigera Walker and G. ?divisa Gaede from Briedelia, Porsica punctifascia Hampson from Aporusa and Porsica curvaria Walker from Bischofia. All four plant genera were previously placed in the Euphorbiaceae but only Briedelia and Aporusa are still included (Willis 1973). Antidesma is the only genus of the Stilaginaceae, suggested by Willis to be intermediate between Icacinaceae and Euphorbiaceae, and Bischofia is likewise unique to its family, the Bischofiaceae, related by Willis to the Staphyleaceae with a comment that relationship with the Euphorbiaceae was probably illusory. The association of the three plant genera in respect of their utilisation by what is probably a natural group of Lepidoptera might indicate that their taxonomic relationships should be reexamined.

Zoogeography

The information on distribution and habitat preference is summarised in the following table:

* Shared with Sumatra only. ** Shared with Peninsular Malaysia only.

The two species in the ‘Other’ category are Ambadra suriga (Borneo and Wallacea) and Quadricalcarifera nigribasalis (Sumatra, Borneo, China, Taiwan).

Only 13 of the species are widespread in the Indo-Australian tropics, and a similar number extend from the Himalaya to Sundaland. About 30 of the species belong to morphologically isolated groups, many monobasic, restricted to Sundaland or extending through Sundaland to Wallacea. The monobasic groups include the genera Besida, Blakeia, Cerasana, Medanella, Pseudohoplitis, Caschara and Rodneya, and the species Turnaca spinifera, Neostauropus major and Ginshachia sumatrensis. Largely allopatric arrays of species with a similar distribution occur in Hunyada, Pantanopsis, Pseudostauropus, Stauroplitis, Sagamora and the Clostera dorsalis group. Other groups with this geography include two or more sympatric species, namely Brykia, Parasinga (though further species from outside the area may await transference from Quadricalcarifera), the Clostera bramah group and the Turnaca punctata group. Greatest complexity is shown by the Erconholda group of species within Phalera where there is as much diversity in the Philippines as in Sundaland.

The single association with New Guinea is shown by Euhampsonia gigantea, though the genus Omichlis has its greatest diversity in New Guinea and the Sundaland species may prove to have a sister relationship to the east; the male genitalia are very different from those of the Himalayan species O. rufotincta Hampson.

There are two Oriental/Melanesian sister pairs of the sort described by Holloway (1982b). One is found in the genus Archigargetta, and Blakeia marmorata may have as its sister species Pseudo gargetta diversa Bethune Baker (New Guinea, Solomons).

The Bornean species of Cerura, Oxoia, Teleclita and Allata, together with those of the Phalera grotei group, belong to arrays of species, largely allopatric, that extend throughout the Indo-Australian tropics, the species falling into localised distribution categories as discussed by Holloway (1982a), e.g. mainland Asian, Sundanian, Wallacean, Lesser Sundan, and Melanesian.

A number of endemic or Sundanian species have sister-relationships with mainland Oriental species as follows:

• Dudusa vethi: D. nobilis Walker (S. China, Taiwan).
• Gargetta costigeroides: G. costigera Walker (India).
• Epistauropus apiculatus: E. vinaceus Moore (India).
• Tensha postobscura: T. striatella Matsumura (Taiwan).
• Saliocleta nannion: Saliocleta sp. (India) + S. barasamphia Schaus (Philippines).
• Liparopsis sundana + L. dympna Schaus (Philippines): L. postalbida Hampson (N.E. Himalaya, Taiwan).
• Fentonia sumatrana + F. helena: F. ocypete Bremer (Himalaya to Amur) + F. excurvata Hampson (N.E. Himalaya).
• Fentonia bipunctus + F. talboti: F. notodontina Rothschild (N.E,. Himalaya).
• Notodontella nieuwenhuisi: N. ferrifusa Dudgeon (N.E. Himalaya).
• Chadisra borneensis: C. bipars Walker or C. albobrunnea Rothschild (Himalaya).
• Pseudoteleclita flavisticta: P. centristricta Hampson (Sri Lanka, India).
• Ginshachia bronacha: G. gemmifera Moore (N.E. Himalaya, China, Taiwan).
• Clostera angularis: C. transecta Dudgeon (N.E. Himalaya).
• Plusiogramma aurosigna: P. argentata Oberthur (China).

Many of these sister-patterns are reflected at a subspecific level by species such as Formofentonia orbifer (cf. Ginshachia), Quadricalcarifera nigribasalis (cf. Tensha), Porsica ingens (cf. Clostera) and Tarsolepis rufobrunnea (cf. Pseudoteleclita).

Several other species are related to taxonomically unexplored mainland Oriental complexes, such as Euhampsonia roepkei, Phalera javana, Phalera acutoides, Antiphalera sumatrana, Peridea albipuncta, Disparia sundana and Ramesa tripunctata. Other species belong to genera with a more even distribution of diversity throughout the Oriental tropics, such as Gargetta, Hyperaeschrella and Hexafrenum. The Chadisra complex extends throughout the Indo-Australian tropics, and Porsica and Ambadra have their centres of diversity in Sundaland. Quadricalcarifera is very diverse throughout the Indo-Australian tropics, particularly in montane habitats. Relationships within Saliocleta, Norraca and Micromelalopha await resolution.*

Clostera fulgurita* and C. restitura are widespread Oriental species that have similarly widespread Palaearctic counterparts, and the latter also has a relative in Melanesia.

There are three instances where a Bornean endemic has as a sister species a more widely distributed yet sympatric species: Benbowia kiriakoffi/virescens; Turnaca altipunctata/punctata; Clostera bramoides/bramah.

Comparisons with Sumatra

Of the 61 Bornean species restricted to the Sunda Shelf, 31 are shared exclusively with Sumatra yet only two are shared exclusively with the Malay Peninsula and none exclusively with Java. No doubt further collecting in the last two localities will change this picture but the difference is sufficiently striking for one to suggest that it will persist.

The list of Sumatran species by Bender & Dierl (1977) totals 154 species. Thus, the recorded fauna would appear to be significantly larger than that of Borneo despite the recording of several species twice through synonymy or errors in past literature (e.g. the record of Antiphalera bilineata probably refers to A. sumatrana, also recorded, though in combination with Grangulina). Species shared with Sumatra comprise roughly 80% of the Bornean total, suggesting close association between the two faunas in the past either directly through exchange or indirectly through gaining the same species when mainland Asian taxa were extending their ranges. It is interesting, therefore, to consider the 25 or so Sumatran species not represented in Borneo that are listed by Bender & Dierl, and also the relationships of the Bornean endemics.

The majority of these Sumatran species are found also in the Himalaya and a few are known also from Java and Bali; a distribution typical of the ‘Sumatran track’ pattern recognised by van Steenis for montane plant taxa (Holloway 1970 and in press a). Representation of this distribution type is very much weaker in Borneo than Sumatra. It is not clear from the list of Bender & Dierl which species are montane but greater penetration by montane mainland taxa might be one explanation for the larger fauna. Species concerned are Baradesa omissa Rothschild, Omichlis rufotincta Hampson (though the genus has its centre of diversity in New Guinea), Hyperaeschra ochropis Hampson, Chloroceramis viridinota Hampson, Metaschalis disrupta Moore and Rosama plusioides Moore.

In addition, Formotensha basalina Gaede has close relatives in the Himalaya and elsewhere and Betashachia angustipennis Matsumura, described from Taiwan, has a distinct subspecies in Sumatra, a remarkably disjunct distribution that may represent the extreme of the pattern shown by Quadricalcarifera nigribasalis and the genus Tensha. Perhaps further collecting in Borneo and the Philippines will ‘fill in the gaps’ in this pattern and reveal it as ‘Luzon track’ montane distribution (Holloway in press a). Mesophalera obliterata Kiriakoff appears to be close to M. lundbladi Kiriakoff from northern Burma.

Another group of Oriental species with similar extension to Sumatra, and sometimes Java, falls within the Pydna Walker group of genera recognised by Kiriakoff (1962), again much richer in Sumatra than Borneo. The species in Sumatra not recorded from Borneo are Antheua servula Drury, a legume feeder in a genus somewhat distinct from the rest, Poncetia albistriga Moore, Norraca longipennis Moore (records of the Philippine Oraura ordgara Schaus may refer to longipennis), Mimopydna essa Swinhoe, Niganda strigifascia Moore, Pydnella rosacea Hampson, Pydnella galbana Swinhoe and Hypambadra delineivena Swinhoe (see Holloway 1982 a: 208; the monotypic genus Hypambadra Kiriakoff may be allied to Tensha Matsumura, indicated by similar modification of the uncus, and the pair a sister-group to Ambadra Moore). Host-plant records for the group are mainly of palms, grasses and bamboos. Therefore the species may be characteristic of open or disturbed habitats where such plants are well represented. Such habitats may have been more extensive from the seasonally arid parts of Burma through Sumatra to Java during dry periods of the Pleistocene glaciations, more so than in Borneo. Phalera combusta Walker, also a grass feeder, falls into this category too.

The majority of species supposedly endemic to Sumatra listed by Bender & Dierl have now been recorded from Borneo, or have Bornean sister species or close relatives. Exceptions are Neopheosia albiplaga Gaede (the relationship with N. fasciata needs to be reexamined), Corinella vittata Gaede (affinities with Allodonta, Hexafrenum and allies), and Closteroides dorsalis Kiriakoff, none of which have close relatives in Borneo.

There are more endemic Quadricalcarifera species in Sumatra than Borneo. The genus contains numerous species, mainly montane, from the Himalaya to New Guinea, and is clearly in need of revision.

Most Bornean endemics are closely related either to species found in Sumatra or to widespread Sundanian species. Exceptions are Phalera styx, with its closest relatives in the Philippines, Turnaca spinifera, with relationships obscure but possibly overlooked in Sumatra, Quadricalcarifera trioculata, possibly related to more easterly species in Sulawesi or New Guinea, Chadisra borneensis, with a Himalaya sister species but itself possibly overlooked in Sumatra, and Micromelalopha cornutijuxta, possibly the sister to the more widespread leucorhetha group.