University of Turku | Department of Biology | Teglab | Nymphalidae Systematics Group | Lepidoptera Links

QUADRIFINE NOCTUIDAE

Hindwing venation of quadrifines apomorphic position

additional links

Erebus macrops Erebinae

Basic definitions

Poole (1995) and Kitching & Rawlins (1998) suggested that a scale-less lower clypeofrons provided a reliable means of distinguishing most quadrifine subfamilies from the trifine complex. It is only weakly evident in Aganainae and Rivulinae s.s. but is present in the Catocalinae complex, Herminiinae, Hypeninae, Stictopterinae + Euteliinae sister-pair and a few smaller subfamilies (Holloway 2008). Therefore the presence of extensive scaling on the clypeofrons immediately above the base of the tongue is taken to indicate trifine affinities in conjunction with other features (Fig. Scale-less and fully scaled of clypeofrons in Catocala and Rivula: Lafontaine & Fibiger 2006).

According to Lafontaine & Fibiger (2006), quadrifine Noctuidae refers to subfamilies in which vein M2 in the hindwing arises in the lower third of the discal cell and is as strong as vein M3, so the cubital vein appears to have four branches; in basal groups of quadrifine Noctuidae, vein M2 is in the same position as it is in trifine Noctuidae (Fig. plesiomorphic quadrifines hindwing in Chytolita, Herminiinae: Lafontaine & Fibiger 2006), whereas in more distal groups the vein is adjacent to vein M3 (Fig. apomorphic quadrifines hindwing in Melipotes, Catocalinae: Lafontaine & Fibiger 2006). Lafontaine & Fibiger (2006) refered to the latter as the true quadrifine hindwing venation. Furthermore, the families Nolidae, Strepsimanidae, Arctiidae and Lymantriidae also have a true quadrifine hindwing.

Classification of Kitching 1984

The re-classification of the major groupings of quadrifines was initiated by a major and comprehensive review of noctuid classification by Kitching (1984), who presented a cladogram representing relationships between the various noctuid subgroups. He united Catocalinae (presence of spines on the mid-tibia) and Ophiderinae (absence of spines on the mid-tibia) in one group, Catocalinae, that he considered was better characterised [than Hypenodinae and Hypeninae] (fused pleural sclerite in the male genitalia and pupa with a whitish bloom) but the majority of genera and species of this vast subfamily have yet to be examined, especially with regard to the pupal character (Holloway 2005). He illustrated three branches for Catocalinae in his cladogram (Catocalinae, Anomis/Alabama, Catocala/Othreis), though he indicated these were merely examples of many genus-groups that could be recognised. Nevertheless, one of these included Catocala Schrank and Othreis Hübner (with Ophideres Boisduval a synonym), the type genera of the two original subfamilies (Catocalinae & Ophiderinae), united by possession of a chitinous projection from the inner margin of the tympanal frame (Holloway 2005).

The Nolidae + Euteliinae clade included four subfamilies (Euteliinae, Nolinae, Chloephorinae and Sarrothripinae), tentatively grouped on the absence of a cremaster. However, as Kitching mentioned, this has not been confirmed for the Nolinae. He also could not find any feature that supported Rivulinae monophyly, but placed it tentatively in a basal position among the quadrifine groups. Hypenodinae was characterized by lack of ocelli, although he suggested that loss of ocelli was highly homoplastic across Lepidoptera and little phylogenetic significance. A similar situation pertained to Hypeninae since its putative apomorphies of long, deltoid palps and lashed eyes were not particularly convincing.

Noctuidae sensu Kitching (1984) was restricted to those groups possessing a post-spiracular hood (although this was absent in Pantheinae and greatly reduced in many other genera scattered among the other subfamilies). As a result, this interpretation omitted a paraphyletic/polyphyletic assemblages of Arctiidae, Aganaidae, Herminiidae, Rivulinae and Acontiinae. The position of the acontiine groups were vague (acknowledged by him) since they were misplaced between Catocalinae and Nolidae + Euteliinae lineages. He showed two branches for acontiines included Acontiinae and Acontiini and stated that only Acontiini was well-characterized, based mainly on the reduced or absent tympanal hood, enlarged alula and sclerotized, and paired anal hair-masses in the male. Arctiidae could be characterised by three apomorphies (none of which were present in all genera), of which two were perhaps quite reliable: the presence of a tymbal organ and of two subventral SV setae on the meso and metathorax in the larvae.

Kitching (1984) stated that no apomorphies could be discovered for the aganaid genera (uncertain position) but it is possible they were more closely related to the arctiids and some to the noctuids, and may even be that some represent the sister-group of the herminiids.

His reason for excluding of the so-called fan-foots and their relatives from Noctuidae (branch 4) was primarily on the basis of the plesiomorphic pre-spiracular hood. Furthermore, this group might be characterised by two apomorphies: modified foretibia in the males of most genera and a swollen metepimeron ventral to pocket IV.

The Euteliinae lineage was perhaps the most well-defined subfamily of noctuids, characterized by the reduced frenulum in the females, larvae feeding mainly on Anacardiaceae, a pair of anal hair pencils in the male, and the attitude of the adults at rest. He mentioned there could be little doubt they represent a monophyletic taxon.

Phylogeny of Noctuoid moths and the utility of combining independent nuclear and mitochondrial genes by Weller et al. (1994)

The molecular analyses of Weller et al. (1994) showed rather surprisingly that the quadrifine examples, including a core catocaline (Catocala), a core calpine (Plusiodonta Guenée) and a hypenine (Plathypena Grote), were grouped with Arctiidae and Lymantriidae rather than with the group of trifine exemplars in analyses of combined DNA data (see the samples). These results raised serious questions about the status of the Noctuidae, indicating that the family is not monophyletic (phylogeny trees of Weller et al. (1994)).

Weller et al. (1994) used the mitochondrial gene ND1 and a nuclear ribosomal RNA (rRNA) data set in their phylogeny estimations. They analyzed the mtDNA data first, prior to combining the two data sets, under different assumptions of character treatment, such as unweighted sites, transversion parsimony, and third codon site omitted. The trees produced from the combined data did not support the monophyly of Noctuidae. Instead, they implied that the cutworms were a grade rather than a clade.

As Weller et al (1994) observed, and as discussed in Yela (1988), this interpretation should be regarded as preliminary because the analyses were based on a very small number of selected species (n = 26, of which only ten were noctuids).

Classification of quadrifines (mostly Catocalinae s.l.) by Speidel & Naumann (1995)

The classification and phylogeny of quadrifines (mostly Catocalinae) was reviewed further by Speidel & Naumann (1995), who suggested that the monophyly of the subfamily was highly doubtful and recommended as a first step the definition of monophyletic groups within the complex and only then attempt to establish the phylogenetic relations among these monophyla (will be discussed in next paragraph). This philosophy has been generally followed subsequently, e.g. by Speidel et al. (1996), Kitching & Rawlins (1998) and Holloway et al. (2001; 2005; 2008). It was also endorsed at a workshop on Noctuidae classification at the Societas Europaea Lepidopterologica (SEL) Congress at Korsør, Denmark, in June, 2002, and also by Kühne & Speidel 2004. At the meeting it was suggested that a review of family-group names would provide a useful foundation for the nomenclature of the group, and this has now been published (Kühne & Speidel 2004; Speidel & Naumann 2005). However, prior to this SEL meeting, Speidel et al. (1996) suggested a division of the Catocalinae/Ophiderinae complex into two groups, reflecting proboscis structure, Catocalini and Calpini and placed the Erebini within their Catocalinae. They did not examine closely many of the genera in the “framed corematous” sequence (Holloway 2005).

Speidel & Naumann restated the chaotic situation in Noctuidae, particularly Catocalinae, the least defined and largest subfamily of Quadrifinae. They logically argued about the centre of quadrifines concentration by expressing out only a few species occur in Europe since the situation is entirely different in other biogeographic regions. For instance, in the tropics Catocalinae is the most diverse of all subfamilies and has the highest number of species. So to compare the species richness of the different subfamilies of Noctuidae, two areas of approximately equal size (Borneo and central Europe) were selected (in both the fauna was comparatively well known). The figure amply illustrates the marked difference in the number of species within the traditionally defined subfamilies between the palaeotropics and Europe. The number of Catocalinae species represented in Borneo may be even higher than illustrated, because at that time, the Bornean catocaline fauna had not been well studied by Holloway. New information on Bornean Catocalinae suggests there are 591 species distributed in 222 genera in Borneo (Holloway 2005). Regardless, Catocalinae contains far more species (about 200 more) than any other noctuoid subfamily on this island. About 2,500 species of Catocalinae occur in the Neotropical region (Heppner 1991), representing more than half the species of the entire subfamily.

Speidel & Naumann stated that the subfamily is defined basically as the quadrifid noctuids according to the classification of Hampson remaining after removal of the following groups (with the exception of the Acontiinae, they are well supported monophyla):

Herminiinae (M2 of hindwing parallel to M3);
Euteliinae + Stictopterinae (female frenulum consisting of one bristle only);
Plusiinae (lashed eyes);
Acontiinae (M2 of hindwing weakly sclerotized and originating more remotely from the lower angle of the discal cell);
Camptolominae + Chloephorinae + Sarrothripinae + Nolinae (bar-shaped male retinaculum, if not reduced).

Thus, the Catocalinae remained an artificial taxon, defined by the exclusion of other taxa from the so-called Quadrifinae. Speidel & Naumann (1995) also surveyed the traditional (Hampson 1902) concepts of Catocalinae (characterized by the midtibiae with spines) and Ophiderinae (lacking such spines). This division was rejected by Berio (1959) who recognized that closely related groups existed in Catocalinae and Ophiderinae, and this indicated the spines had been convergently lost (or gained) several times. As a result, Berio proposed that the two subfamilies (Catocalinae and Ophiderinae) should be combined. Speidel & Naumann concluded that the monophyly of this concept of subfamily Catocalinae was highly doubtful. The main reasons, in addition to its complexity, were the heterogeneity of the taxon and the large number of species. The main discoveries of Speidel & Naumann (1995) were the monophyly of tribe Euclidiini (based on the shape of the internal female genitalia (ductus receptaculi)), and the separation of the two tribes Melipotini and Eulepidotini (they found no synapomorphies to unite the two tribes).

The phylogeny of Noctuidae by Speidel et al. 1996

Speidel et al. (1996) presented a phylogeny of both quadrifine and trifine Noctuidae on which ten putative synapomorphies were mapped, with emphasis on the male genitalia and the tympanal region. The tree and relationships within the Noctuidae, was supported by the following characters (Lafontaine & Fibiger 2006):

Family Noctuidae was characterized by the presence of an adenosma ( a tube-like ventral prothoracic gland) in the larva, this structure being absent from the two families basal to the Noctuidae in their phylogeny, namely Arctiidae and Aganaidae; absence of the gland in Rivulinae and one species of Hypeninae were treated as secondary losses;
Aganaidae (here Aganainae) were treated as the sister group of Noctuidae on the basis of a shared elongated uncus;
The remainder of the phylogeny was similar to that of Todd (1983) and Fibiger and Hacker (1991), with Herminiinae treated as basal due to the prespiracular hood prespiracular hood (similar to that of Arctiidae and Aganaidae);
The noline-chloephorine-sarrothripinae clade was included in Noctuidae next to Eustrotiinae;
Amphipyrinae were retained as a mixture of Cuculliinae s.l. genera, resembling previous lists;
On the basis of the nodular form of the tympanal sclerite, the genera Aedia Hübner and Tyta Billberg were moved from the quadrifine noctuids to form two subfamilies near Acontiinae, in which the tympanal sclerite is a flat plate or is bar-like, as in the quadrifine Noctuidae and the more basal subfamilies of trifine Noctuidae;
On the basis of the presence of basal abdominal brushes and associated structures, the genera Pseudeustrotia Warren, Sinocharis Püngeler, and Lophonycta Sugi were removed from Acontiinae s.l. and placed in monobasic subfamilies next to the higher trifine subfamilies (the cuculliie s.l. and ncoctuine s.l. groups of subfamilies).

As can be seen from the phylogenetic tree of Speidel et al. (1996) no apomorphic characters were proposed to support the monophyly of the quadrifine groups of Noctuoidea.

Classification of Noctuidae by Kitching & Rawlins (1998)

Kitching & Rawlins in their classification made extensive use of character systems of both adults and immature stages (see the phylogeny tree by Kitching and Rawlins 1998: Lafontaine & Fibiger 2006). The main departures from previous classifications were the following (Lafontaine & Fibiger 2006) (the non quadrifine´s alterations have been omitted):

Nolidae, Sarrothripinae, Chloephorinae and Eariadinae were treated as subfamilies of family Nolidae (quadrifine noctuoid);
Nolidae were associated with Arctiidae on the basis of the elongate, bar-shaped retinaculum of the male and other characters;
Pantheinae were removed from Noctuidae and treated as family Pantheidae, which was closely associated with Lymantriidae;
Calpinae were segregated as a subfamily separate from Catocalinae;
A group of quadrifine subfamilies (Herminiinae, Catocalinae, Calpinae, Hypeninae, Aganainae, Cocytiinae, Strepsimaninae, Euteliinae, and Stictopterinae) were associated on the basis of the bare lower frons.

It should be noted that Kitching & Rawlins divided the Catocalinae/Ophiderinae complex into Catocalinae, (including Erebini) and a restricted concept of Calpinae (including Gonopterini and Anomiini), noting the unsatisfactory nature of the classification, including paraphyly of their Catocalinae.

It is necessary to mention that the manuscript was finished in 1992 and represented the state of knowledge and the authors´s concepts at that time.

Revision der Systematik der Noctuidae (=Revision of systematics of the Noctuidae) by Beck (1960, 1989, 1999-2000)

Beck devoted his research career to the study of noctuid larvae. Several years after Kitching’s publication, Beck produced a paper containing a phylogenetic tree of noctuid moths based mainly on larval characters (Beck 1992). His phylogeny differed from that of Kitching, reflecting some discordance between adult and larval morphology but chiefly the author’s particular viewpoint (Yela & Kitching 1999). In fact, his interpretation was not cladistic in a strict sense (Yela & Kitching 1999). Beck’s scheme reinforces the necessity of considering larval morphology and the need for further research. It represented an advance because Beck’s scheme was an alternative against which that of Kitching could be compared and because the paper contained a great deal of previously unpublished information on larvae.

The classification of Beck (1999-2000) split the Noctuidae into a large number of tribes, subtribes, and genera, which were grouped into a relatively small number of subfamilies. The classification included a very large number of family-group names, of which 68 were proposed by Beck either in this work or in a previous checklist. These family-group names were diagnosed but not defined in a hierarchical classification on the basis of derived character states (Lafontaine & Fibiger 2006).

Phylogeny of Noctuidae by Yela (1998)

Yela (1998) recognised seven main phyletic lines within noctuids (six of them quadrifinoids), that were identified as: (1) aganainoid (Aganainae); (2) pantheinoid (Pantheinae); (3) chloephorinoid (Eariadinae + Chloephorinae + Sarrothripinae + Nolinae); (4) eutelinoid (Euteliinae + Stictopterinae); (5) deltoid-catocalinoid ((Herminiinae + Strepsimaninae + Hypeninae) + (Gonopterinae + Calpinae + Catocalinae)); (6) eustrotinoid (Eustrotiinae + Bagisarinae); and (7) trifinoid (Acontiinae + Plusiinae + Acronictinae + Raphiinae + Bryophilinae + Heliothinae + Stiriinae + Agaristinae + Dilobinae + Cuculliinae + Hadeninae + Noctuinae, arranged in four large dichotomous clades).

Noctuid phylogeny revisited by Yela & Kitching (1999)

Yela & Kitching (1999) recognized three main lineages of Noctuidae, although acknowledged that the family might not be monophyletic (Mitchell et al. 2006). The first lineage contained Nolinae s.l., Pantheinae s.l. and Aganainae s.l.; the second contained the remaining quadrifine subfamilies, divided into a deltoid-catocalinoid grade and the Euteliinae + Stictopterinae clade; the third contained the trifines sensu Poole (1995), including (in contrast to Kitching & Rawlins 1998) the acronictinoid subfamilies (Acronictinae, Raphiinae and Bryophilinae), and the Bagisarinae, which Poole regarded as quadrifines.

They recognised Catocalinae, Calpinae and Scoliopteryginae (as Gonopterinae) as subfamilies, and suggested that the second and third were more closely related to each other than to Catocalinae. They excluded Bagisarinae, Aediinae and Tytinae, indicating these are probably best included in the trifine sequence of subfamilies (Holloway 2005).

With the exception of Beck (1992), who conceived noctuid evolution in four clear-cut phases, and leaving aside the work of Kitching (1984) (since his viewpoints up to 1992 are set forth in Kitching & Rawlins, 1998), in general there is some agreement for the recognition of a heterogeneous group of basal paraphyletic or polyphyletic quadrifinoid lineages and an advanced, probably monophyletic, trifinoid group. Most other noctuid researchers shared this viewpoint. It is worth mentioning that by this time the quadrifinoid and trifinoid nomenclature within noctuids had become artificial. Historically, it had been used for designate groups with quadrifid and trifid hindwing features, but currently it makes no sense. A trend of quadrifinoid-trifinoid polarity has been recognised in the gradual reduction of the M2 hindwing vein (lymantriids and arctiids have a well-developed vein with a quadrifid origin which was considered the plesiomorphic state in noctuids). But reduction has probably taken place several times during noctuid evolution (Lafontaine & Poole, 1991; Kitching & Rawlins, 1998). Hence, there are species with vein M2 fairly well developed that are not easily included in any group, while the trifinoid noctuids include several subfamilies displaying a more or less clear quadrifid hindwing venation (Eustrotiinae, Bagisarinae, Eucocytiinae, Acontiinae, Plusiinae, Aediinae, Condicinae, Stiriinae and Dilobinae, and also some acronictinoids). For the moment the terms are being used for descriptive, but should perhaps eventually be dropped.

With regard to the composition of and phylogenetic relationships among the aforementioned heterogeneous group of quadrifinoids, viewpoints differ radically among authors. The nolinoid clade, recognised as such by all authors, is another mosaic of (numerous) plesiomorphic and (scarce) apomorphic character states relative to noctuids, and its assignment and placement is the second of the noticeable disagreements. Beck (1992) placed it among basal trifinoids, with which it has no connection at all. Speidel et al. (1996) postulated an advanced position within quadrifinoids; Yela (1998) suggested a basal position among quadrifinoids; and Kitching & Rawlins (1998) put them outside noctuids, as family Nolidae. The (unsolved) problem regarding the relations and position of Pantheinae is the third big puzzle. Authors vary in their opinions, putting the group within the more advanced trifinoids (Poole, 1995), among the basal trifinoids (Beck, 1992), in a median position within quadrifinoids (Speidel et al. 1996), in a basalmost position (Yela, 1998), and even outside noctuids (Pantheidae; Kitching & Rawlins 1998). With regard to Aganainae, they were placed by Kitching & Rawlins (1998) in a median position within quadrifinoids (between the catocalinoids and Cocytiinae), by Yela (1998) as a basal clade and by Speidel et al. (1996) outside noctuids. The putative deltoid-catocalinoid clade (Herminiinae, Strepsimaninae, Hypeninae, Rivulinae and the groups around Catocalinae) was recognised by Yela (1998) and, tentatively, Kitching & Rawlins (1998), who cited the information of Arn et al. (1986) and Renou et al. (1988) of a common basic chemical structure of the female sex pheromones. Beck (1992) split deltoids from catocalinoids, and Speidel et al. (1966b) put Herminiinae outside the rest of deltoids due to its prespiracular counter-tympanal hood and placed the eutelinoid line between postspiracular deltoids and catocalinoids (which is unusual). For the Euteliinae + Stictopterinae clade, consensus among authors is probably highest. It is placed in various positions among the quadrifinoids, both subfamilies always in a sister group relationship.

Eustrotiinae and Bagisarinae were considered to be trifinoids by Beck (1992), Speidel et al. (1996) and Kitching & Rawlins (1998) and quadrifinoids by Yela (1998). Poole (1995) listed only Eustrotiinae among the trifinoids. If those subfamilies displaying separate counter-tympanal cavities and a fused sclerite between vinculum and tegumen are termed quadrifinoids, then Eustrotiinae and Bagisarinae are undoubtedly trifinoids. The close relationship between both subfamilies is widely recognised (Speidel et al. 1996; Kitching & Rawlins, 1998; Yela, 1998). As for the rest of the trifinoids, which are better known, the schemes in Speidel et al. (1996), Kitching & Rawlins (1998) and Yela (1998) are relatively concordant, with minor differences, although the previously noted radical departure regarding acronictinoids should again be highlighted. The cladogram in Poole (1995) diverges in several details from those of Speidel et al. (1996) and Yela (1998) and is rather different from the scheme discussed by Kitching & Rawlins (1998). It has to be noted that Poole himself recognises that he could not characterise some groups suitably, and that his proposition is a hypothetical one.

Phylogeny of Catocalinae by Fibiger (2003)

Fibiger (2003), whilst attempting to define monophyletic units within the European fauna, nevertheless felt it necessary to use and define various systematic levels or ranks: tribes, subtribes, genus-groups, subgenera and species groups. He also attempted to divide a much wider selection of genera into the two subfamilies recognised by Kitching & Rawlins (1998): Catocalinae and Calpinae (Holloway 2005).

As can be seen from the phylogenetic tree proposed by Fibiger (2003), he recognized three tribes of Catocalinae and nine subtribes of Catocalini in Europe. He stated that a few other tribes of Catocalinae were found outside Europe. His Calpinae (as subfamily) comprised only two genera, each in its own tribe. Indeed, Fibiger’s philosophy was based on an attempt to provide clearer definition of the Catocalinae by circumscribing monophyletic units based on the Hennigian principles, and excluding those supposed para- or polyphyletic groups that were not confirmed by apomorphies (Fibiger 2003 in Goater et al. 2003). He acknowledged that, in the near future, , it was more likely that Catocalinae would be formally elevated to family level, and separated, along with other the quadrifines, from the trifine Noctuidae, and that many of the subsidiary levels would thus have to be elevated accordingly.

Catocalinae was defined by Fibiger based on three autapomorphic character states (Goater et al. 2003):

Hindwing with a black terminal patch and fringes between the M3 and Cu1 veins. In genera or species-groups where a different hindwing pattern appears to have survival value this character is sometimes lost. However, in several of such groups, one or more members are often found to have retained a more or less recognisable dark or blackish patch.
Posterior apophyses rod-shaped throughout, or slightly wider basally, but without the usual flat, diamond-shaped basal structure, close to the ovipositor.
Further character states occur in combinations of two or more within the subfamily (combinations of characters cannot constitute an autapomorphy as proposed by Fibiger (2003), but they can be apomorphic):

Wings held flat in resting position;

a A quadrifid hindwing venation, presence of a fully developed M2;

A more or less sclerotized pocket in the scaphium, built to receive the often sharply pointed tip of the uncus. It must be noted that Lödl (2000) demonstrated this character is also a common feature in the Hypeninae;

Most genera, with the exception of those in a single tribe, have no corona (the ancestral condition = plesiomorphic);

Almost all genera of Catocalinae have a more or less heavily sclerotized scaphium, many genera also have a less strongly sclerotized subscaphium;

All species have more or less asymmetrical male genital armature, often including the valvae;

A further character, common among genera of Catocalinae, is a scale-less area directly above the proboscis;

A sclerotized, often striated, elongated patch, which can be short or distinctly long, near the apex of the vesica and extending into the basal part of the vesica ejaculatorius;

Finally it should be noticed that most Catocalinae species have coloured or patterned hindwings a surprisingly rare condition among Noctuidae in general.

Fibiger & Lafontaine (2005) discussed the character states of Fibiger (2003), and concluded that the first character is also present in some species of Hypeninae and Acontiinae; the second was considered a synapomorphic character for Erebidae (by Fibiger & Lafontaine 2005); and the third character, which was a unique combination of several character states rather than a true synapomorphy, was still useful for defining groups of Catocalinae, but they were not uniquely derived character states (= apomorphic) that would define strictly monophyletic groups. Fibiger & Lafontaine (2005) proposed that, with the removal of Tytini, Armadini and Aediini, Catocalinae (described as Catocalini by Fibiger 2003) became monophyletic.

Finally, it should be noticed that Fibiger noted several genera from the list of Calpinae that might, after a more intensive examination, be transferred into Catocalinae; those where the shape of juxta is similar to those of the genus Aedia. These genera, including Aedia, might prove to be a separate tribe within Catocalinae.

The Moths of Borneo Family Noctuidae, subfamily Catocalinae: by Holloway (2005)

Holloway (2005) in volume 15-16 of “The Moths of Borneo” series made a comprehensive contribution towards the classification of Catocalinae s.l. He argued that, for various reasons, it was not possible to follow the system of Fibiger (2003) in its entirety. This system was presented as a highly structured phylogeny, at least for the Catocalinae (Holloway 2005). In fact, Holloway reverted to the spirit of the discussion at Korsør, endorsed by Kühne & Speidel (2004), and merely identified tribes, interspersed by the approach of Forbes (1954), who recognised sequences of miscellaneous genera. The dangers of adopting too precise and formally defined an approach using only a limited suite of characters is illustrated by the fact that those used by Fibiger (2003) have caused him in a few cases to divide between his Catocalinae and Calpinae groups of genera clearly linked by synapomorphies.

The Catocalinae of Holloway was treated in a broad sense, and included the old concept of Ophiderinae, but as he stated, this assemblage is probably paraphyletic. As a contribution to the development of a better classification of the group, the species were divided onto 18 tribes, four further groups of genera that may be monophyletic, and six miscellaneous sequences of genera. Two new tribes were also proposed, Sypnini and Anobini.

Holloway´s contribution covered 591 species in 222 genera of Oriental Catocalinae. Two tribes, four genera, 99 species and one subspecies were described.

He added that many species in Catocalinae are highly mobile and have very wide geographical ranges; some migrate over long distances, and species of Mocis Hübner, the Anomis Hübner group and Anticarsia Hübner have proved particularly successful in reaching remote oceanic islands (Holloway & Nielsen 1999). Of 30 noctuid species recorded as migrant to Norfolk Islands, half are Catocalinae in the broad sense (Holloway 1996).

A review of the higher classification of the Noctuoidea (Lepidoptera) with special reference to the Holarctic fauna by Fibiger & Lafontaine (2005)

The study of Fibiger & Lafontaine is indeed the most recent and most comprehensive morphological classification, included both adult and immature characters, was concordant with the most recent and most comprehensive molecular classification, the main differences being the rankings given to various subfamilies and tribes.

The classification of Fibiger & Lafontaine (2005) was built mainly on the framework of the Kitching and Rawlins (1998) classification, but the following changes were proposed:

Placing the arctiid group of families (Nolidae, Arctiidae and Lymantriidae) in front of the upgraded family Erebidae so that their close relationship with the quadrifids (=quadrifines) was better reflected;
Moving the Lymantriidae from a position in front of Nolidae to a position after Arctiidae to reflect the close association of the arctiids and lymantriids following Mitchell et al. (1997, 2000);
Treating the quadrifine subfamilies (Herminiinae, Hypenodinae, Hypeninae, Catocalinae, Calpinae, Hypeninae, Stictopterinae, and Euteliinae) as subfamilies of the re-established family Erebidae and as the sister group to the arctiid families, and reinstating the subfamily Rivulinae following Fibiger and Hacker (1991), Speidel et al. (1996), and Beck (1999-2000), the subfamilies Boletobiinae and Phytometrinae following Beck (1999-2000), and the subfamilies Scolecocampinae and Erebinae on the basis of data presented in their own argumentations;
Hypenodinae is reinstated for Hypenodes Doubleday and its relatives, and the taxon Strepsimaninae is treated as family Strepsimanidae;
Revising the classification of the subfamilies Catocalinae, Calpinae and Erebinae on the basis of Goater et al. (2003) and new information presented in their paper, moving Armadini and Aediini from Catocalinae to Acontiinae, and move Tytini from Catocalinae into subfamily Metoponiinae of the Noctuidae;
Following Fibiger and Hacker (2002, 2004) removing the tribe Eublemmini from Eustrotiinae and treat it as a basal subfamily of the Erebidae;
Araeopteroninae Fibiger was proposed as a new subfamily of Erebidae;
A new family, Micronoctuidae Fibiger, was proposed;
Reinstating Pantheinae as a subfamily of Noctuidae near Acronictinae on the basis of molecular research by Mitchell et al. (1997, 2000) and Yela and Kitching (1999);
Removing the tribe Eriopini from Hadeninae (s. l.) and treating it as a subfamily of Noctuidae;
Removing the tribe Phosphilini from Hadeninae (s. l.) and move it to Psaphidinae next to Nocloini;
Revising the classification of Hadeninae sensu Kitching and Rawlins (1998) into subfamilies, tribes and subtribes;
Reviewing the classification of Noctuinae on the basis of Lafontaine and Fibiger (2004) and the subtribes of Noctuini.

Quadrifine Noctuidae assemblage or Family Erebidae of Fibiger & Lafontaine (2005)

On the basis of morphological data reviewed by the authors, and that of Mitchell et al. (1997, 2000), subfamilies Herminiinae, Hypeninae, Calpinae, Catocalinae, Stictopterinae and Euteliinae were united into the family Erebidae and treated as the sister group to the arctiid families. In addition, Rivulinae, Boletobiinae, Hypenodinae, Phytometrinae, Scolecocampinae, and Erebinae were reinstated as subfamilies of Erebidae. Eublemminae was transferred from Noctuidae (s.s.) and treated as a subfamily of Erebidae.

They recognized the necessity of constructing a separate family-group taxon comprising the entire quadrifines. The need to recognise Erebidae as a family separate from Noctuidae began in the 1990’s when the results from molecular research provided increasingly strong support for a close relationship between Arctiidae and quadrifine subfamilies of Noctuidae (Weller et al. 1994.; Mitchell et al. 1997, 2000), indicating that Noctuidae (s.l.) was paraphyletic with respect to Arctiidae and Lymantriidae.

Kühne and Speidel (2004) proposed conserving the subfamily name Catocalinae over the rarely used name Erebinae on the basis of ICZN Article 23.9.1. This Article requires that a name has not been used since 1899. Kühne and Speidel argued that Erebinae had not been used as a senior synonym of Catocalinae since 1899. However, it has, in conjunction with subfamily Catocalinae (e.g., Richards, 1933, 1935, 1936, 1939; Forbes, 1954), and it is thus available and takes priority over Catocalinae. Erebidae corresponds to a small part of Calpinae as defined in Goater et al. (2003), where relatively few genera were listed. Nye (1975) wrote, “Those genera without mid tibial spines [those with spines he called Catocalinae] are retained as Ophiderinae. An immense amount of research is required into this large and uncatalogued subfamily of over 1100 genera, which must be reclassified particularly where they merge into the following two subfamilies”.

Fibiger & Lafontaine restricted Calpinae to the fruit-piercing moths and their relatives, as defined by Kitching and Rawlins (1998). They also mentioned that three groups of genera had been particularly problematic in their placement within the quadrifid subfamilies. They were the Parascotia Hübner group (Boletobiinae Grote), the Phytometra Haworth group (Phytometrinae Hampson), and the Pangrapta Hübner group (Eublemminae, tribe Pangraptini Grote). Pangraptini was used for Zethes Rambur by Beck (1999-2000) but Zethes is in Catocalinae: Ophiusini (Goater et al. 2003) and only superficially resembles the genera of Pangraptini. In the most recent North American list (Franclemont and Todd 1983), these three groups were placed in Rivulinae, Hypeninae, and Catocalinae respectively, following the work of Richards (1933) on the progressive development of the tympanum. In contrast, in Europe the first two groups were included in Rivulinae by Fibiger and Hacker (1991) but were moved to Catocalinae by Beck (1999-2000). All three groups have a primitive hindwing venation (vein M2 about 1/3 up the cell parallel to M3), like Rivulinae, Hypenodinae, Herminiinae and Hypeninae, and differ from Erebinae + Calpinae + Catocalinae, which have vein M2 converging toward M3 and adjacent to it at the apex of the discal cell. However, they lack the diagnostic larval characteristics of Rivulinae, Hypenodinae, Herminiinae and Hypeninae described by Beck (1999-2000), so Fibiger & Lafontaine followed Beck and used the names Boletobiinae and Phytometrinae, but as subfamilies of Erebidae, not as tribes of Catocalinae. Finally they treated Pangraptini as a tribe of Eublemminae.

How is the family Erebidae of Fibiger & Lafontaine (2005) characterized according to the apomorphic character states? (Apomorphic character states in bold):

Larvae with only primary setae;
Larvae with three SV setae on A1;
Silk pore of the larval spinneret concealed by an apical depression edged laterally by flaps (illustrated in Crumb 1956, and Beck 2000) (silk pore open in Noctuidae and secondarily open in some Erebidae);
Adults with vein M2 in the hindwing as thick as M3 (a character state shared with some groups of basal Noctuidae sensu stricto (some Eustrotiinae, Pantheinae, and Raphiinae);
The erebid subfamilies Erebinae, Calpinae, Catocalinae, Euteliinae, and Stictopterinae (and Arctiidae and Lymantriidae) have vein M2 in the hindwing arising in the lower part of the cell adjacent to the M3-CuA1 stem;
Scales on the ventral half of the frons deciduous, quickly falling off to leave the area bare in most specimens (see the figure of scale-less clypeofrons). This character is an easy way to recognise moths of most species in the family. However, the frons is scaled in the more basal subfamilies (Rivulinae, Boletobiinae, Hypenodinae and Araeopteroninae) (see the figure of scaled frons of more primitive subfamilies), so the bare frons is likely a derived condition (=apomorphy) of the clade formed by Eublemminae, Herminiinae, Hypeninae, Erebinae, Calpinae, Catocalinae, Stictopterinae and Euteliinae;
The erebid subfamilies Erebinae, Calpinae, Catocalinae, Euteliinae, and Stictopterinae (and Arctiidae and Lymantriidae) have a direct articulation between the tegumen and vinculum (i.e. pleural sclerite fused to anteroventral margin of tegumen, unlike most Noctuidae);
Erebidae have no basal abdominal brushes (and associated levers and pockets), which is a derived character found in most of the subfamilies of Noctuidae (absent from the basal lineages of the Noctuidae and secondarily lost many times);
In the tympanal area pocket 4 is open (V-shaped), as in Arctiidae, not closed (with a small opening at the top) as in Noctuidae. A closed pocket is interpreted as an independent modification in some Nolidae and some Erebidae (i.e., Eublemminae: Pangraptini, Scolecocampinae and Stictopterinae) (Richards 1933);
The sclerite defining the posterior margin of the tympanum proper is a bar or flat plate in Erebidae (a non-sclerotized fold in Arctiidae; an epaulette with raised, rounded nodules (= nodular sclerite) in most Noctuidae) (Speidel et al. 1996);
The two counter-tympanal bullae are small and separated in Erebidae (large and fused medially in the Noctuidae);
Erebidae have a heavily sclerotized scaphium, and often also a less heavily sclerotized subscaphium (Noctuidae have a membranous scaphium and a more-or-less sclerotized subscaphium);
Posterior apophysis in females of Erebidae stick-like, often expanded where it fuses with base of anal papillae (lacking the elongated or diamond-shaped preapical sclerite near the base of the anal papillae in Noctuidae).

Systematics and evolution of the cutworm moths (Lepidoptera: Noctuidae): evidence from two protein-coding nuclear genes by Mitchell et al. (2006)

The classification of Mitchell et al. (2006) was the last of three studies on Noctuoidea based on two nuclear genes elongation factor-1α (EF-1α) and dopa decarboxylase (DDC, 700-1100 bp ), but the 2006 study included twice as many taxa (146 species), as the previous two (i.e. Mitchell et al. 1997, 2000), with an emphasis on trifines, giving the authors a more definitive interpretation of the results in terms of proposing changes to the classification than they had had in the two previous studies. Lafontaine & Fibiger (2006) summarized major findings of the study of Mitchell et al. as follows:

The close relationship between a clade comprising Arctiidae and Lymantriidae and one containing five subfamilies of quadrifine Noctuidae (i.e. Calpinae, Catocalinae, Strepsimaninae [on the basis of Phobolosia anfracta (Hy. Edwards), which is in Scolecocampinae in Fibiger and Lafontaine 2005], Rivulinae [on the basis of Mycterophora Hulst, which is in Boletobiinae in Fibiger and Lafontaine 2005], and Eublemmini [Eublemminae in Fibiger & Lafontaine 2005];
Placement of Rivulinae s.s. between Lymantriidae and Arctiidae as the sister group to Arctiidae in the strict consensus tree, although this position is unstable because Rivulinae are the sister group to Lymantriidae + Arctiidae in the maximum-likelihood tree (Fig. 4 in Mitchell et al.);
Placement of Nolidae as the sister group to the rest of quadrifine Noctuidae (including Arctiidae and Lymantriidae);
Placement of a clade groupings Euteliinae and Stictopterinae above Nolidae as sister group to the remainder of the quadrifine noctuids.

Lafontaine & Fibiger (2006) also discussed about the three high bootstrap support of the groupings of subfamilies as follows:

A large clade that included Hypeninae, Aganainae, Herminiinae, Calpinae, Catocalinae, Strepsimaninae [Scolecocampinae], Rivulinae [Boletobiinae], Eublemminae, Lymantriidae, Rivulinae and Arctiidae, termed the L.A.Q. clade (i.e. lymantriids, arctiids, quadrifine noctuids) by Mitchell et al.;
A clade supporting the L.A.Q. clade plus the euteliiine-stictopterine clade (55%-66% support);
A clade grouping the Euteliinae with the Stictopterinae (99% support).

Revised higher classification of the Noctuoidea (Lepidoptera) by Lafontaine & Fibiger (2006)

The classification of Lafontaine and Fibiger (2006) was mainly based on an evaluation and comparison of the similarities and differences between two most recent comprehensive classifications (Fibiger & Lafontaine 2005 and Mitchell et al. 2006), combining the morphological and molecular results of both studies.

They recognized the fundamental problem of making direct comparisons between two classifications using different names for the same groups of genera. To compare effectively the two classifications, they listed the taxa used by Mitchell et al. and then gave the equivalent taxon in the classification of Fibiger & Lafontaine (see table 1 and table 2). They stated there were two main difficulties in converting the consensus trees generated by Mitchell et al. into a classification. First, it could be difficult to treat nuclear gene sequences as uniquely derived character states (I never really understood what this “difficulty” was!). Second, the trees were based on only 146 species and are strongly biased towards the trifine Noctuidae (= Noctuidae s.s.) and especially towards Noctuinae s.l. (Xyleninae, Hadeninae and Noctuinae s.l.). The subfamilies and tribes of quadrifine noctuoids and those of the other families of quadrifid noctuoids (72 taxa in Fibiger and Lafontaine 2005) were represented by only 30 species from 20 family-group taxa and only nine of these represented the type species of the taxa, so some subfamilies and tribes are represented by genera not necessarily closely related to the type species. As a result, Mitchell et al. generally did not attempt to change the names of higher taxa on the basis of so few exemplars, but listed these taxa in their current status.

They recognized seven principal areas where the classification of Fibiger and Lafontaine (2005) differed from that of Mitchell et al. (2006) (those are related to the quadrifines have been discussed below):

The Arctiidae-Lymantriidae clade in the classification of Mitchell et al. was nested within a large clade comprising the noctuid subfamilies Hypeninae, Aganainae, Herminiinae, Calpinae, Catocalinae, Scolecocampinae (as Strepsimaninae), Boletobiinae (as Rivulinae), Eubleimminae and Rivulinae (s.s.). This clade was strongly supported by the molecular results of Mitchell et al. and is referred to as the L.A.Q. clade (Lymantriidae, Arctiidae, quadrifine Noctuidae clade). Within the L.A.Q. clade, the relationships among subfamilies were weakly supported with the exception of a well-supported clade uniting Aganainae and Herminiinae. In the classification of Fibiger & Lafontaine, the Arctiidae-Lymantriidae clade was the sister group to Erebidae (= quadrifine Noctuidae). The monophyly of Erebidae, however, was based on the smooth, dorsally sclerotized scaphium and the apical flaps on the larval spinneret; both character were secondarily lost in some lineages of the Erebidae, leading to the possibility that the arctiid-lymantriid clade might be derived from within Erebidae (the dorsal surface of the scaphium is membranous in the arctiid-lymantriid clade and in the trifine Noctuidae). The possibility that the arctiid-lymantriid clade might be derived from within Erebidae is further strengthened by two derived character states shared by the arctiid-lymantriid clade with only part of Erebidae, namely a clade consisting of Erebinae, Calpinae, Catocalinae, Aganainae, Cocytiinae, Stictopterinae and Euteliinae (the erebine-euteliine clade). These characters are the fully quadrifine hind wing venation, where vein M2 in the hind wing has moved to the lower part of the cell adjacent to or fused with the base of M3 and CuA1 (apomorphic position of M2), and the fusion of the pleural sclerite with the posterior margin of the tegumen, creating a simple elbow-like articulation between the tegumen and the vinculum. Theses two character states, shared by most subfamilies of quadrifine Noctuidae and the arctiid-lymantriid clade, were treated as convergent by Kitching & Rawlins (1998), but their presence in these two clades, in combination with the lack of any consistent character state to define family Noctuidae (s.l.), suggests that Erebidae are more closely related to the arctiid-lymantriid clade than to Noctuidae s.s. This interpretation led Fibiger & Lafontaine to treat the quadrifine noctuids as family Erebidae and as more closely related to the arctiid-lymantriid clade than to the trifine noctuids (Noctuidae s.s.). The difficulty with this classification, however, was that the derived character states that the arctiid-lymantriid clade shares with Erebidae are found in only part of Erebidae, namely the erebine-euteliine clade. The possibility that Erebidae could be paraphyletic with respect to the arctiid-lymantriid clade was discussed by Fibiger & Lafontaine (2005), with the implications this would have for the classification. These morphological characters, in combination with the well-supported L.A.Q. clade from the molecular results of Mitchell et al., led the authors to accept that Erebidae as defined by Fibiger & Lafontaine is a paraphyletic group and that a new classification was needed to reflect this. The adults of Arctiidae and Lymantriidae with smaller palpi have a fully scaled frons, unlike other members of the erebine-euteliine clade, whereas the frons is denuded in genera of Lymantriidae with larger labial palpi. This suggests that the general lack of a denuded area on the lower part of the frons in the arctiid-lymantriid clade may be related to the smaller size of the labial palpi, and that it is the long, upturned labial palpi that are responsible for denuding the frons.
Aganainae have been, and continue to be, an enigmatic group that has both noctuid and arctiid characteristics. Aganaines share three character states with Arctiidae: a bar-like retinaculum in males, a prespiracular tympanal hood, and loss of the adenosma (ventral prothoracic gland) in the larvae. However, Aganainae lack the two derived character states of Arctiidae (the metathoracic microtymbal organs and dorsal eversible pheromone glands associated with the anal papillae in females) used by Kitching and Rawlins (1998) to define Arctiidae, but absence of these two arctiid synapomorphies does not exclude the possibility that Aganainae are the sister group of Arctiidae, with the association supported by the three character states listed above. The quadrifine noctuid characteristics of Aganainae (bare lower frons, long pointed uncus, dorsally sclerotized scaphium, one SV seta on T2 and T3 in the larva, larvae with primary setae only) are all plesiomorphic character states of the L.A.Q. clade and are not relevant to a possible association of Aganainae with Arctiidae. Mitchell et al. associated Aganainae with Herminiinae with strong (>90%) support but, as stated above, such an association is not supported by morphological characters. Therefore, the association of Aganainae with Herminiinae must await confirmation by other data sources, such as molecular studies using more nuclear genes, longer gene sequences and more taxa (two genera of aganaines and two herminiines only were included in Mitchell et al. 2006). Aganainae have a prespiracular tympanal hood, a character state shared with Herminiinae (and Arctiidae), but the bar-like retinaculum in males, loss of the adenosma, and fully quadrifine hindwing venation all suggest a closer association of Aganainae with Arctiidae than with Herminiinae.
Nolidae combine character states of the trifine Noctuidae and Arctiidae. The tympanal structure of Nolidae is similar to that of the trifine noctuids, especially Eustrotiinae, whereas the bar-like retinaculum of Nolidae is similar to that of Arctiidae. Also, the larvae of Chloephorinae and Risobinae have only primary setae and two SV setae on A1, as in the trifine Noctuidae, whereas the larvae of Nolinae have secondary setae on verrucae, as in Arctiidae. The family was treated as the sister group to the remainder of the quadrifid Noctuoidea (i.e. Strepsimanidae, Arctiidae, Lymantriidae, Erebidae, Micronoctuidae and Noctuidae) by Fibiger and Lafontaine. In Mitchell et al. Nolidae were the sister group of a large clade consisting of the euteliine-stictopterine clade plus the L.A.Q. clade, but with only weak molecular support. The best-known derived morphological character state of Nolidae is the boat-shaped cocoon with a vertical exit slit. Holloway (2003) gave seven additional characters to support the monophyly of Nolidae, but most have not been surveyed throughout Nolidae, and the derived versus primitive condition of many has not been determined. Nolidae have a bar-like retinaculum in males, like Arctiidae, and this, in association with the fully quadrifine hind wing plus the fused pleural sclerite, suggests a placement of Nolidae within the L.A.Q. clade near Arctiidae, in spite of the basal position of the family in Mitchell et al., where it is far removed from the L.A.Q. clade. As in Arctiidae, the scaphium is unsclerotized and the frons is fully scaled in most groups, although some genera with larger labial palpi (e.g. Iscadia Walker) have a bare lower frons. Nolidae share a derived character state with Arctiidae that is not in Lymantriidae, namely the bar-like retinaculum in males. The isolation of some highly divergent long-branch groups, and the relatively few taxa of quadrifine noctuids (17 species) included in the molecular study, might contribute to taxa such as Nolidae being misplaced in the cladogram of Mitchell et al.
Lafontaine & Fibiger believed that the stictopterine-euteliine clade was also misplaced in the Mitchell et al. classification, where it is treated as the sister group to the entire L.A.Q. clade of subfamilies. Subfamilies Stictopterinae and Euteliinae are part of a smaller clade (i.e. Erebinae, Calpinae, Catocalinae, Aganainae, Cocytiinae, Stictopterinae and Euteliinae) that has a fully quadrifine hind wing and a fused pleural sclerite. The arctiid-lymantriid clade also has these last two character states but has lost the denuding of the frons characteristics of a larger clade (Eublemminae, Herminiinae, Scolecocampinae, Hypeninae, Phytometrinae, Aventiinae, Erebinae, Calpinae, Catocalinae, Aganainae, Cocytiinae, Stictopterinae and Euteliinae), although it is retained in some genera of Lymantriidae. It is possible that the arctiid-lymantriid clade is more closely related to the erebine-calpine-catocaline-aganaine-cocytiine part of the clade than it is the stictopterine-euteliine part of the clade. This may explain, at least in part, the divergent position of the stictopterine-euteliine clade in Mitchell et al.
Rivulinae s.s. were shown as the sister group to Arctiidae ( Fig. 2 in Mitchell et al.) or Arctiidae + Lymantriidae ( Fig. 4 in Mitchell et al.), but with weak molecular support. The instability in the placement of Rivulinae was probably the reason they were not raised to family status by Mitchell. et al. In Fibiger & Lafontaine, Rivulinae were treated as the most basal subfamily of Erebidae because of the primitive tympanal structure, fully scaled frons, partially quadrifine hind wing in the adult, and lack of a ventral prothoracic adenosma in the larva. This last character might suggest an association with Arctiidae. However, the free pleural sclerite and lack of a fully quadrifine hind wing suggest a basal position of Rivulinae within the quadrifine Noctuidae rather than a position within the erebine-euteliine clade. A free pleural sclerite is typical of trifine Noctuidae but is also present in the more basal subfamilies of quadrifine Noctuidae (e.g. Rivulinae, Boletobiinae, Hypenodinae, Araeopteroninae, Eublemminae, Herminiinae, Scolecocampinae, Hypeninae, Phytometrinae and Aventiinae). The long, barbed setae of the larvae give them an arctiid look, but the hairy appearance is based on the unusual length of the setae rather than on multiple setae. As for the barbs themselves, all noctuid setae are minutely barbed but the barbs are conspicuous in Rivulinae because of the enlarged size of the setae.

Classification of Lafontaine & Fibiger (2006)

Lafontaine & Fibiger modified their previous classification (Fibiger & Lafontaine 2005) as follows (those which are related to quadrifines have been mentioned):

Moving Arctiidae and Lymantriidae into Erebidae (sensu Fibiger & Lafontaine 2005) and treating them as subfamilies and as the sister groups to erebine-calpine-catocaline-cocytiine clade;
Treating the Aganainae as a plesiomorphic sister group to a redefined Arctiinae;
Moving Nolidae into the quadrifine Noctuidae and treating it as a subfamily and as the sister group of the arctiine-aganaine clade;
Including the stictopterine-euteliine clade within the larger L.A.Q. clade in the subgroup with a fully quadrifine hindwing and a fused pleural sclerite, and treating it as a highly derived sister group to the clade consisting of Erebinae, Calpinae, Catocalinae, Cocytiinae, Nolinae, Aganainae, Arctiinae and Lymantriinae.

According to the phylogenetic tree of Lafontaine & Fibiger the following apomorphic character states were proposed:

Scaphium sclerotized dorsally (character 2) and vesica globular and multi-pouched (character 3) supported a clade consist entire quadrifine Noctuidae assemblage including: Rivulinae, Boletobiinae, Hypenodinae, Araeopteroninae, Eublemminae, Herminiinae, Scolecocampinae, Hypeninae, Phytometrinae, Aventiinae, Erebinae, Calpinae, Catocalinae, Cocytiinae, Stictopterinae, Euteliinae, Nolinae, Aganainae, Arctiinae, Lymantriinae and Strepsimaninae;
Character 4 ( lower frons denuded of scales) supported a clade consisting of the eublemmine-strepsimanine assemblage ( see the clade);
Characters 5, 6 (hindwing fully quadrifid ( bases of M2 and M3 approximate= apomorphic state) & pleural sclerite fused to tegumen, respectively) supported a clade consisting of the erebine-strepsimanine assemblage ( see the clade);
Character 7 ( tympanal hood prespiracular (spiracle below hood in Nolinae)) supported a clade consisting of the noline-lymantriid clade ( see the clade);
Character 8 ( retinaculum 2 times as long as wide (bar-like)) supported a paraphyletic clade consisting of the noline-aganaine-arctiid clade on the one hand, and Herminiinae on the other hand ( see the clade) (this character seems to be convergent rather than an apomorphic).