Larval Characters of Taxonomic Significance of New Zealand Ennomines (Lepidoptera: Geometridae)

Transactions of the Royal Society of New Zealand : Zoology, Volume 1, Issue 16, 25 November 1961, Page 215

Larval Characters of Taxonomic Significance of New Zealand Ennomines (Lepidoptera: Geometridae)

J. S. Dugdale

By

[Received by the Editor, October 20, 1960.]

Summary

The general external larval morphology of New Zealand Ennomines is described and briefly compared with that of Indian and Nearctic species. The abdominal chaetotaxy is compared with that of other Geometrid sub-families, and the published homologies of the SV and subprimary L setae are compared. Thirty species of Ennomines from six genera were examined, and the characters were considered so consistent that they are taxonomically valid. Reasons are given, based on larval and genitalic characters to show that Selidosema dejectaria (Walker) is incorrectly placed in the genus Selidosema. This paper places New Zealand Ennomines into the Ennomini ( Sestra, Azelina, Gargaphia, Declana ), Erannini ( Hybernia) and Boarmiini [Selidosema). The larval and genitalic characters of these genera are outlined. In erecting taxonomic groups on larval structure, the structure of all instars must be used, and this grouping must be correlated to grouping on adult structures.

Introduction

This study arose out of the need for a sound basis on which to erect field guides for the identification of arboreal defoliators by the Forest Biology Survey. Of the 56 New Zealand Ennomines, eight species, previously restricted to indigenous hosts, now defoliate exotic conifer species; other Ennomine species are also found in exotic plantations, associated with relict areas of indigenous forest or scrub, or with the invading understorey of indigenous shrubs and ferns. Recognition of these latter species is important as they may act as alternate or alternative hosts for parasites and disease organisms affecting the conifer-feeding species.

Previous descriptions of Ennomine larvae by Purdie (1884), Hudson (1928, 1939, 1950), Clark (1936) and Chappell (1920) are confined to descriptions of colour pattern, occasionally illustrated, and sometimes with a brief mention of outstanding structures. Recent.work on Indian Geometrid larvae by Singh (1953, 1956) and Nearctic larvae by McGuffin (1950, 1956, 1958 a, b) has taken into account the morphology of the larva rather than colour pattern. Singh (1953) showed that identification by colour alone is impracticable. This paper is part of an attempt to bring knowledge of New Zealand Ennomines up to the level of current work on Indian and Nearctic Geometrids and on Nearctic Noctuids (Hard-

wick, 1958) and Olethreutids (McKay, 1959). Larvae of the following species were examined:

Larentiinae. Asthena pulchraria Dbldy. (all instars) ; Chloroclystis paralodes Meyr. (all instars); C. semialhata (Walk.) (all instars); Hydriomena callichlora (Butl.) (all instars) ; H. similata (Walk.) (instars 3-5); H. deltoidata (Walk.) (instar 5) ; Tatosoma fasciata Philp. (all instars) ; T. tipulata (Walk.) (instars 3-5); Venusia undosata (Feld.) (all instars) ; V. charidema Meyr. (instar 5) ; Xanthorrhoe cymozeucta (Feld.) (instars 3-5) ; X. rosearict (Dbldy.) (all instars) ; Oenoghromatinae : Samana falcatella Walk, (instars 3-5); Dichromodes nigra (Butl.) (instar 5); Epirrhanthis alectoraria (Walk.) (all instars) ;E. ustaria (Walk.) (all instars); E. veronicae Prout (all instars). Sterrhinae: Leptomeris ruhraria Dbldy. Ennominae. Selidosema aristarcha Meyr. (all instars) ; S. dejectaria (Walker) (all instars) ;S. fenerata Felder (all instars) ;S. fascialata (instar 5); S. indistincta (Butler) (all instars) ; S. leucelaed Meyr. (all instars) ; S. modica Philpott (instars 2-5); S. monacha Huds. (instars 2-5); S. panagrata (Walker) (all instars) ; S. pelurgata (Walker) (all instars) ; S. productata (Walker) (all instars) ; S. prototoxa Meyrick (instars 2-5) ; S. rudiata (Walker) (all instars) ; S. suavis (Butler) (all instars) ; Gargaphia muriferatct Walker (instars I—3, 5) ; Sestra flexata (Walker) (all instars) ; Sestra humeraria Walker (all instars); Azelina fortinata (Walk.) (instars 2-5); A. gallaria (instars 3-5) ;A. variabilis (Warr.) (all instars) ; Hyhernia indocilis (Walker) (instars 3—5) ; Declana atronivea (Walker) (all instars) ; D. egregia (Felder) (instar 5); D. feredayi Butler (instars I—3, 5); D. floccosa Walker (all instars); D. glacialis Hudson (all instars); D. griseata Hudson (instars 4-5) ; D. hermione Hudson (instars 2-5) ; D. junctilinea (Walker) (all instars) ; D. leptomera (Walker) (all instars) ; D. niveata (Butler) (instars 3-5). All adult determinations were made by the author and most were checked against specimens in Dominion Museum and Entomology Division, Nelson, collections.

Larval and most of the adult material is in the collection of the Forest Biology Survey, Forest Research Institute, Rotorua; some adult material was loaned by Entomology Division, Nelson. Most of the larval material was collected by the Forest Biology Survey, and the writer wishes to thank his colleagues for their assistance in the preparation of this paper.

External Morphology. I. Chaetotaxy

The terminology for abdominal setae is a modification of Hinton’s (1946) system; setae of the anal proleg are named according to Singh (1953) and the anal shield setae according to McGuffin (1958 b). Singh’s notation for the abdominal segments (Al, A2, etc. for first, second abdominal segment, etc.) is adopted for brevity. The terms “ primary ” and “ subprimary ” refer to setae which appear in instar I and instar II respectively. Secondary setae are those which appear in instars HI—V.

Proprioceptor and tactile setae are present on Lepidopterous larvae in definite patterns (Hinton, 1946) and the Geometridae are no exception to this rule (McGuffin, 1958 b). The proprioceptors, responsive to intersegmental movement, are situated on those areas overlapped during segmental contraction. They are not considered in this paper. The tactile setae, responsive to contact with surrounding objects, are situated more in the mid-section of each segment and are longer than the proprioceptors. Using the homologies proposed by Hinton, each typical Ditrysian abdominal segment bears five groups of setae. Dorsally, there are the dorsal (D) setae, D ls D 2 , both primary. More or less directly ventrad of Di is the subdorsal group, containing a large SDi and a small or rudimentary SD 2 , both primary. The lateral group consists of 2 primary setae Li and L 2 , and two subprimary setae L 3, L 4 (Geometridae) or one subprimary seta L 3 (other families ). The subventral group consists of two primary setae SVi, SV 2 and one subprimary

seta SV 3 . The ventral group is unisetose and primary. On segments A 7 and A 8 the SV group is usually bisetose, and on AB, Di is often ventral of D 2 , and the numbers of setae in the L and SV groups are reduced.

Homologies of abdominal setae for Ennomines (Enno.), Larentiines excluding Asthenini (Lar.); Asthenini (Asth.) and Oenochromatines (Oeno.) and according to the present modified

H : nton homology (MH) and to McGuffin (Me) and Singh (S).

S, subprimary; P, primary. Primary/subprimary inconsistencies in italics. (* This seta absent in Alsophila Hubn.; ** this seta absent in Ozola Walk. Singh does not describe Brephos, or Asthenine Larentines.)

The anal shield setae are termed by McGuffin Di on the disc of the shield, D 2 at the apex, SDi marginal; between SDi and D 2 , also on the margin is Li (Fig. 2). These anal shield setae are primary in the Ennominae. Singh’s homology for anal proleg setae in Geometrids recognises three setal series: an anterior series (CPi, CP 2 , etc.) an outer lateral series (Li. 3), and a posterior marginal series (GDx, CD 2 , etc.). A modification of Singh’s terminology is used in this paper; his Li is subprimary, and is here termed L 3; his L 3 becomes Lx of the anal proleg (Fig. 2). Singh’s terminology for the anal shield is not used because McGuffin’s terminology, while in no sense homotypic, uses no subprimary names for primary setae. McGuffin did not attempt to name the anal proleg setae.

Setal patterns common to larvae of the Tortricidae, Olethreutidae, Oecophoridae, Gelechiidae, Heliodinidae, Glyphipterygidae and Carposinidae are shown diagrammatically in Fig. le; in the Psychidae ( Oiceticus omniuorus Fered.) this pattern lacks SV 2 on Ax (Fig. Id). One difference between this arrangement and the chaetotaxy of the Lyonetidae, Plutellidae and Tineidae (Fig. If) lies in the position of the primary L setae; these are close together or on a common pinaculum in the first group, and wide apart in the second group of families. Fig. 1c represents diagrammatically the abdominal setal pattern common to Rhapsa Walk., Ariathisa Walk., Erana Walk., Melanchra Hubn. and Persectania Hamps. (SV 2 absent on Ax in Rhapsa ), and Figs, la, b, icpresent the basic Ennomine and Larentiine

abdominal chaetotaxy respectively (SV2 absent in Asthenine Larentiines). The most obvious difference between Noctuid and Geometrid chaetotaxies is the presence in the Geometrids of the additional subprimary lateral seta, here termed L 4.

Seta L 3 is often the most posterior (Hinton, 1946). The larva of Plusia chalcites Esp. (Noctuidae), a “ looper ”, resembles the typical Geometrid larva in that the A r , proleg is posteriorly placed on the segment, instead of being directly under the spiracle. L 3 in P. chalcites arises between two furrows, and the same condition is seen in Geometrid larvae.

The additional subprimary L seta, which is on the anterior part of the segment (usually on a line ventrad of SDi) has no clear homologue in other families. Hinton scarcely dealt with Geometrid abdominal chaetotaxy, which is complicated not only by the seta L 4, but by the variable occurrence of SV 3 and SV 3 . Both Singh and McGuffin base their homologies on Hinton’s system, but when these homologies are compared (Table I), disagreement is seen in the SV and subprimary L groups. These workers erected their homologies on penultimate and ultimate instar larvae, and this approach has resulted in systems in which primary setae on one segment are homologised with subprimary setae on another. Only by comparing both first and second instar larvae of a species is it at first possible to distinguish between primary and subprimary setae.

In Table I, the homologies of McGuffin and Singh are compared with the modified Hinton homology used in this study. Although both workers also considered the Sterrhinae, Brephinae and Geometrinae, these could not be included in the table as little material of Sterrhinae was available and there are no Geometrinae or Brephinae in New Zealand. The table indicates the confusion of homology, not only between segments, but between groups of a subfamily—e.g., L 3 is called SV 3 in the Larentiines except the Asthenines, where it is regarded as SVi.

In the Nearctic fauna, McGuffin’s (1958 b) figures for Larentiines also demonstrates the loss of SV 3 on all abdominal segments. His figures of Alsophila pometaria Harr. (Oenochromatinae) (p. 91, Figs. 133, 134) resembles those of Deilinea and Sicya sp. (Ennominae), (p. 91, Fig. 138, p. 90, Fig. 131), and does not agree with his key to the Geometrid subfamilies on p. 11.

The absence of SV 3 in Larentiines is used as a taxonomic criterion by Singh (1953, p. 11); his other criteria are the chaetotaxy of the anal shield, anal proleg and the A 6 proleg. His subfamilies Hemitheinae and Geometrinae correspond to McGuffin’s Geometrinae and Ennominae respectively.

The seta L 4 is of general occurrence throughout the family, and the primitive setal pattern is likely to be that found in Epirrhanthis Hubn. (Oenochromatinae), where SVi, SV 3 and SV 3 are present on Ai as well as on A 2-- The Ennominae retain SV 3 on A l} but have lost SV 2 ; on segments A 2 -s all three SV setae are present. On the example figured by McGuffin and Singh, the Geometrinae have lost SV 3 on Ai, and the Brephinae SV 3 and SV2 on Ai; the SV group is complete on other segments, Singh divides the Sterrhines into two groups, but the homology of the setae on Ai is difficult to determine from his figures, as seta Vi appears to be either absent or placed more laterally than ventrally. SV 3 is present on A 2.5. The Larentiinae have lost SV 3 on all abdominal segments, and the group Asthenini have lost SV 2 on Ai. While Epirrhanthis in the Oenochromatinae may represent the primitive chaetotaxy, the Nearctic genus Alsophila, and the Indian genus Naxa resemble the Ennomine pattern; the Indian genus Ozola has Ai with SVi, A2 with SVi and SV 2 , and A 3 with SVi, SV 2 , SV 3 . The New Zealand genera Dichromodes and Samana have the complete Epirrhanthoid chaetotaxy.

While for most species examined the chaetotaxy is bilaterally symmetrical, in Selidosema suavis larvae the number of SV 3 setae varies from side to side and from segment to segment. The seta SV 3 is single in other Ennomine species, but is

represented by two, three, or four setae, arranged more or less horizontally, in S. suavis. In a sample of 56 larvae, all of the same instar from one locality, collected at the one time, 20 were bilaterally symmetrical on all segments, and 36 were asymmetrical. The numbers of SV 3 setae present on symmetrical larvae were: (Ai) 2+ 2, (A 2 ) 2+ 2, (A 3.5) 3+3. Asymmetrical larvae had the arrangements (Ai) 2+ 1, (A 2 ) 2+ 3, (A 3.5) 3+4. Some larvae had SV 3 quadrisetose on A 4 and A 5. SV 3 was lacking on one side of one larvae on Ai. Symmetry and asymmetry of the SV 3 individual numbers on the segments is shown in Table 11.

In one example, SVx was lacking on one side of A 4, but other setal groups were stable in all examples. By segments, the range of the SV 3 numbers in the sample was: Ax : 0-2 setae; A 2 : 1-3 setae; A 3 : 3-4 setae; A 4 : 2-4 setae; A 5 : 1-4 setae. Selidosema rudiata showed a similar condition, SV 3 being double on A 4 in one larva.

The criteria for delimiting the Indian Ennomine larvae are (Singh, 1953) :

(a) Anal proleg setae CD 2 above the level of L 2 . (b) SV setal group trisetose on A 2 . 5 . (c) Hypoproct strongly produced to a fine point. (d) Anal shield with only seta Di on the disc. (e) Proleg on A e with four or more lateral (“external”) setae, two of these setae being far dorsad of the ventral margin of the lateral sclerotised plate on the proleg. For Nearctic Ennomine larvae (McGuffin, 1956, 1958 a, b) : (a) No rudimentary prolegs on A 1.5. (b) SV setal group trisetose on A 2 „ 5 . (c) The setae SVi, SV 3 and Vi in vertical alignment on Ai. (d) Four to eight annulets (transverse sclerotised bands which are muscle attachments) on each segment from A 1.5. (e) Anal shield with level of SDi anterior to level of D x .

The characters common to Nearctic, Indian and New Zealand Ennomine larvae are:

(a) Anal shield with SDi anterior to Di; (b) Hypoproct produced to a fine point, strongly developed; (c) Four to eight annulets on each of the first five abdominal segments; (d) The SV se.tal group is trisetose (up to six-setose) on A 2 _5.

The New Zealand Ennomines show disagreement with Singh’s and McGuffin’s criteria in the following examples:

The genus Declana has seta CD2 below the level of L2 on the anal proleg (cf. Singh: a.); Selidosema dejectaria has three lateral setae on the A 6 proleg— SVi, SV 2 and SV3 (cf. Singh: e) ; Declana atronivea, D. egregia, D. niveata, D. floccosa, D. griseata and D. leptomera possess functional or rudimentary prolegs on A 5 (cf. McGuffin: a) ; Declana, Sestra and Gargaphia do not have setae SV3, SVi and V x in vertical alignment (cf. McGuffin: c).

The three setae L 3, SVi and Vi are in vertical alignment on A 3 in Hyhernia indocilis and Selidosema spp., except S. dejectaria, where SVi is on a line anterior to L 3 and Vi. The three setae are not in vertical alignment in Declana spp., Sestra spp. or Gargaphia spp.; SVi is always anterior to L 3 and is anterior or posterior to Vi ( Sestra , Declana spp.) or directly dorsad of Vi (Gargaphia).

The seta SD 2 is present as a microscopic seta or as a sclerotised puncture, and is directly ventrad of SDx on Ax- 4 , and antero-dorsad of, or directly anterior to SDx

on A g and A 6 in Hybernia indocilis and Selidosema spp. (except S. prototoxa and S. dejectaria, in which SD 2 is in the same position on A 5 and A e as on Al-4, directly ventrad of SDi). SD 2 is directly ventrad of SD X on A x . 6 in Declana floccosa, D. niveata and Sestra spp. SD 2 is postero-ventrad of SD X in other Declana spp. on Ai_ 6 j and in Gargaphia SD 2 is posteroventrad of SD X on A x . 5 , and directly ventrad of SDi on A 6 -8-

Setae SV 3 , SVi, and Vi are in vertical alignment on Ai in Selidosema spp. and in the Nearctic genera Plagodis Hubner, Anagoga Hubner, Hyperitis Guenee, Srmiothisa Hubner, Deilinea Hubner and Itame Hubner (McGuffin, 1950, 1956, 1958 a). These setae are not aligned vertically in Declana spp., Sestra spp. or Gargaphia, SV3 being anterior ( Sestra, Gargaphia) or posterior ( Declana spp.) to SVi. SV3 is absent on Ai in some Declana atronivea and in Hyhernia indocilis.

In most Ennomines, SV3 is about one half the size of SVi, but may be equal in length, as in late instar larvae of Gargaphia, Declana floccosa, D. niveata and Selidosema suavis.

The position of seta CD 2 in relation to L 2 on the anal proleg is typical (i.e., CD 2 is dorsad of L 2 ) in Hybernia, Selidosema, Azelina and Gargaphia; in Sestra, CD 2 and L 2 are level, and in Declana CD 2 is ventrad of L 2 .

The lateral setae on the A 6 proleg belong in three classes: primary (SVi, SV 2 ); subprimary (SV 3 , which may be double or triple) and secondary (only in Declana floccosa, D. niveata, Azelina spp. and Gargaphia) . The number of subprimary plus secondary setae for each genus is: Selidosema: 2; {S. dejectaria: 1) Hyhernia: 2; Sestra: 4; Declana (except D. floccosa and D. niveata) : 3 ; D. floccosa: 9; D. niveata: 8; Gargaphia: 10. In those species with more than 4 subprimary plus secondary setae, the number of setae increases with each instar.

First instar larvae of the genus Azelina (Fig. 8e) have groups of setulae homologous to the setae Di, Li, L 2 and SVi; D 2 , SDi are unisetose. Second instar larvae are setulose over wide areas of each segment and accurate homotypy is impossible.

11. Mandibles.

Both McGuffin and Singh described Ennomine mandibles. Dissimilarity between individuals of the same species is demonstrated by Singh (1953: p. 87, PI. V, figs. 40, 41, 43) in his description of Semiothisa pluviata Fab. One form has dentate mandibles, a second has smooth margined mandibles, and a third form has smooth mandibles with a strong “ internal tooth ”. The adults cannot be differentiated. In the New Zealand Ennomines the mandibles bear two setae in the scrobe; there is a coarsely toothed outer part and a finely toothed inner (mesal) portion on the cutting margin. There are no differences that are conspicuous or constant between genera.

The development of the mandible, figured for Selidosema suavis is characteristic for other New Zealand Ennomines. The outer “ teeth ” are coarse and caninelike, and the first and third teeth have a ridge extending from the apex to the base of the mandible. The greatest growth of the mandible from instar to instar is in the development of the mesal, fine toothed region (Fig. 4). The number of mesal teeth increases with each instar. In some species all the teeth are triangular, in others they are rounded; in Selidosema pelurgata, the last instar mandibles have the cutting margin reduced to a wavy edge. In all species, the lateral tooth, which projects outwards instead of towards the midline, is well developed on early instar mandibles, being comparatively greatest in size in the first instar. In Selidosema spp. the lateral tooth is vestigial in the last instar; it is relatively large in last instar larvae of the other Ennomine genera. Observations on the eclosion of Selidosema suavis larvae from eggs showed that the lateral tooth is not used for ripping the egg membrane; the first outer tooth is employed.

111. Ocelli arrangement and the antennal pit.

The arrangement of ocelli in relation to the margins of the antennal pit or indentation is of one type in Selidosema (except in S. dejectaria) and in Hybernia, and of another in the other genera. The antennal pit is roughly triangular, the base being the outer margin of the mandible base, and the other two sides formed by the parietal lobe of the epicranium. The apex of the pit is also different in the two groups. In Selidosema and Hybernia, ocellus 4is separated from the margin of the pit by a distance equal to or greater than the ocellus diameter. The apex of the pit is suddenly narrowed and slightly but distinctly bent. In S. dejectaria, and in species of Declana, Sestra, Azelina and Gargaphia ocellus 4 reaches the pit margin, or is separated from it by a distance equal to less than the diameter of the ocellus. In S. dejectaria, the antennal pit apex is broadly rounded; it is evenly acute in other genera.

IV. Crotchet arrangement; claw structure.

Both Singh and McGuffin have placed emphasis on the crotchet arrangements in penultimate and ultimate instar larvae for the classification of genera within the Ennominae. Ultimate instar crotchet arrangements of larva of New Zealand Ennomines present a varied picture; in Selidosema, some species have complete mesoseries, others incompletely broken mesoseries, or completely broken mesoseries. An incompletely broken mesoseries is one in which the mid-section of the series is represented by a row of blunt, peg-like, very short crotchets. Crotchet arrangement in the first instar larva is uniform for a genus, and the New Zealand genera fall into three groups.

In the first group, the crotchet arrangement is primarily an incompletely broken mesoseries ( Hybernia; Selidosema spp. except S. dejectaria). Ultimate instar larvae in some species have complete mesoseries.

In the second group, the crotchet arrangement is primarily a completely broken mesoseries; the gap is narrowed in successive instars, being obliterated in the ultimate instar ( S. dejectaria).

In the third group, the crotchet arrangement is primarily a complete mesoseries, with all crotchets equal in size ( Azelina, Declana, Gargaphia, Sestra spp.).

Last instar Azelina and Gargaphia have completely broken mesoseries, and Sestra last instar larvae have incompletely broken mesoseries. In Declana spp. the last instar crotchet arrangement may be different on each proleg pair.

In all species examined, the functional crotchets are uniordinal in instar I, irregularly biordinal in instar II and biordinal in later instars. No comparisons can be made with Nearctic and Indian genera, as the first instar crotchet arrangements were not examined by either Singh or McGuffin, except for Deilinea Hubner (McGuffin, 1958 a). The three main types of crotchet arrangement are shown in Fig. 6.

The tarsal claws of the foreleg have been used by McGuffin (1958 b) as indications of relationship, or lack of it, between groups of Larentiine genera. He has postulated primitive and advanced types of claws (“ thoracic claws have evolved from a simple type with an acute notch to a specialised type with a well rounded notch ” 1958 b, p. 22). Claws vary within a genus in the New Zealand Ennomines, but generally, the claw has an acute notch, that of Selidosema and Hyhernia being nearly a right-angle, while that of Azelina and Sestra is narrowly acute. Declana has a small bulge at the angle, often giving a double notched appearance.

V. Pigmentation Pattern.

The pigmentation pattern of the third to ultimate instar larva is of little taxonomic use for distinguishing between genera as either the pattern varies between individuals of the same species and between species of the same genus, or species of widely different genera have the same pattern. The first instar patterns (Fig. 8) are less variable and there are marked differences in this character between genera and groups of genera. The genera Hyhernia and Selidosema (except S. dejectaria) have pallid dorsal and lateral areas; there is a dark broad subdorsal stripe on each meson. The subventer and venter may be dark or pale; the dorsum may be darkened in transverse bars. In all species examined, the prothorax bears only the subdorsal band. In Azelina, Gargaphia, Sestra and Declana the body bears either a single ( Azelina, Gargaphia, Sestra) or double ( Declana) dorsal stripe, and subdorsal, lateral (spiracular), subventral and ventral stripes. Generally these stripes are thin, and the areas between are pallid; in some Declana spp. with very dark first instar larvae the dorsal and sub-dorsal areas are infuscate, the lateral area is pallid or white, and the subventral and ventral areas are infuscate; dorsal, subdorsal and/or lateral stripes are, however, present on the prothorax, or, as in D. atronivea small elongate pallid irregular patches separate the dorsal from the subdorsal stripe.

In Selidosema dejectaria the dorsum, subdorsum, subventer and venter are dark, and the lateral area is white, or pallid, with reddish flushes. It resembles the first instar of Boarmia bistortata Goeze (= Ectropis bistortata Goeze) figured by Schimitschek (1958).

VI. Body Shape, ornament, “ resting pose ”.

Body shape in larvae of first to third instars is invariably cylindrical. In the later in stars the species assumes a characteristic shape, either cylindrical or ventrally flattened. Segments are of uniform thickness, or greatly swollen either around or just posterior to the spiracle; larvae which are cylindrical have a twig-like “ resting pose ”, except the larvae of Azelina, Gargaphia and Sestra which lie appressed to the fern rachis and jump off when disturbed, coiling and uncoiling with great rapidity. Larvae with a twig-like resting pose are usually dark on the venter and have projections resembling twig stumps on the second abdominal segment. In many of these species the first two pairs of forelegs are appressed, while the third pair is held out at an angle of 90° to the body. Declana atronivea and D. egregia rest either coiled or extended; their colour pattern and body form are disruptive either way. D. glacialis in instars I-111 rests coiled, and drops to the ground without a suspensory thread of silk (as do all instars of Azelina, Gargaphia and Sestra); later instar larvae lie appressed to the host plant.

The larvae which lie appressed to the host plant (except D. glacialis) are flattened ventrally, and have the venter pallid. The body outline is broken by

a disruptive colour pattern and/or by the presence of laterally placed long white setae (i S . suavis ) or foliose papillae. In nearly all species the tubercles on the body are variable; particularly this is so in species with appressed larvae. Hudson’s figure (1928) of Selidosema leucelaea portrays a larva type which has been rarely found in the four years that the Forest Biology Survey has collected this species from indigenous and exotic hosts.

Larvae of Declana atronivea, D. egregia, D. leptomera, D. junctilinea and D. feredayi have a lateral papilla anterior to the A e spiracle; in D. feredayi this appears to butt on to a scobinated mound on the posterior margin of A 5. The papilla is not evident in D. griseata (instars 3-5) or in any instars of D. floccosa, D. hermione, D. niveata or D. glacialis.

Key to Subfamilies of New Zealand Geometridae

Larvae of the Geometrid subfamilies can be distinguished by the following key, which is a combination and interpretation of keys by McGuffin (1958 b) and Singh (1953); subfamilies not occurring in New Zealand are in parentheses. Second to ultimate instars only are considered.

1. SV 3 present on (if absent, then hypoproct long, acute) 2 SV 3 absent on A x , often absent on A x _ 5 ; hypoproct short, blunt or anal plate setae D x , SD X , in a transverse row (see couplet 5) 3

2. Hypoproct short; SV 2 present on A t N.Z. Oenochromatinae Hypoproct long; SV 2 absent on A x Ennominae

3. Abdominal prolegs on A 3 SV 2 present on A x ( Brephinae ) Abdominal prolegs on A 6 and A lO only

4. SV 3 present on A 3 often on A 2 _ 5 5 SV 3 absent on A x . 5 Larentiinae (a) SV 2 absent on A x Asthenini (b) SV 2 present on A 1 Other Larentiine groups

5. Anal plate with setae D 1? SD X approximately in a transverse row ...... Sterrhinae (a) SV 2 , SV 3 absent on A x , SV 3 absent on A 2 Group A (Singh) (b) Only SV absent on A t , SV SV, SV 3 present Group B (Singh) on A 2 . 5 ‘ _ Anal plate with seta D t on a line distinctly posterior to SDj; SV 2 absent on A 1? present on A 2 . 5 ( Geometrinae)

The subfamilies Oenochromatinae, Ennominae and Larentiinae are well represented in the New Zealand fauna. The Sterrhinae is represented by a single immigrant, Leptomeris rubraria Dbldy., present also in Australia.

The larvae of the New Zealand Ennomines may be identified to their genera by the following key, which is a summary of the Sections I-VI above.

Key to the New Zealand Genera of Ennomine Larvae

1. First instar larvae with the proleg crotchets in an incompletely broken mesoseries; each meson of the body with a single broad dark subdorsal band. Second to ultimate instar larvae with setae V t , SV 1 and L 3 in vertical alignment on A 3 A 6 proleg with two subprimary setae, no secondary setae ...... 2

First instar larvae with the proleg crotchets in a cornequal mesoseries; each meson of the body, or at least the pro thorax, with two to four bands (dorsal, subdorsal, lateral, subventral). Second to ultimate instar larvae with setae V ls SV X and L 3 not in vertical alignment on A 3 A fi proleg with three subprimary setae and often several secondary setae 3

First instar larva with the proleg crotchets in a completely broken mesoseries. Second to ultimate instars with SVj on a line anterior to Vj on A 3 A 6 proleg with one subprimary seta, and no secondary setae - ( Selidosema ) dejectaria Walk.

2. First instar larva with black head capsule. Second to ultimate instars with SV 3 absent on A ls A 2 Hybernia indocilis Walk

First instar larva with brown or red head capsule. Second to ultimate instar larva with seta SV 3 present on Al, A2 Selidosema Hubn.

3. First instar larvae with the dorsal band red or brown, double; where the larva is dark on the dorsum and subdorsum then a lateral papilla is present on A 6. Second to ultimate instar larvae with all setae strong; prolegs may be present on A 5; seta GD 2 on anal proleg slightly below level of seta L 2 ; larvae defoliate woody angiosperms Declana Walk.

First instar larvae with dorsal band red, broad, single, double only on prothorax; no lateral papilla on A 6. Second to ultimate instar larvae with seta Lj on A 6 or all setae fine; no prolegs on A 5; seta CD 2 never below level of seta L 2 on anal proleg. Larvae defoliate ferns 4

4. First instar larvae with a longitudinal red band between seta and SV 1 on the thoracic segments. Second to ultimate instar larvae with seta L x on A 6 _ 8 extremely fine; crotchets arranged in an incompletely broken mesoseries; A 6 proleg with 2-3 subprimary setae Sestra Walk.

First instar larvae with no continuous red band between setae Lj and SVj on the thoracic segments. Second to ultimate instar larvae with either seta L 1 on A 8 only, fine, or all setae fine; crotchets arranged in a broken mesoseries; A 6 proleg with more than 3 subprimary setae 5

5. First instar larvae with SVj and SV 2 on widely separated; no setal members replaced by setulose groups. Second to ultimate instar larvae with only L x on A s fine; A 6 proleg with 5-9 subprimary and secondary setae Gargaphia Walk.

First instar larvae with SVj and SV 2 on A 1 on a common pinaculum, and setae D ls Lj and L 2 represented by groups of setulae. Second to ultimate instar larvae with large setulose areas on the head capsule and body, no setae long or strong; A 6 proleg with 20-40 setulae Azelina Guenee

It is idle to assume that the above differences between genera or groups of genera are taxonomically valid unless there are corresponding differences in adult and pupal structure. The Ennomine pupa has been studied by Forbes (1945) and he distinguishes two types —(a) the “ Boarmines ” with a bifid cremaster, vestigial hooked setae, or these absent, and a “ flange plate ” present on at least A 5; (b) the “Ennomines” with the cremaster simple, bearing four pairs of hooked setae, the terminal pair(s) longer and stronger than the others, flange plate absent, and the femurs concealed. Forbes divides the “ Boarmines ” into (a) the “ Bistonine ” type with the dorsal (transverse) groove present and femurs concealed, and (b) the “ Paraphia” type with no dorsal transverse groove and with the femurs exposed. The New Zealand Ennominae have concealed femurs and both the dorsal transverse and the lateral grooves are present (Fig. 7c). The genus Selidosema (including S. dejectaria) has a pupa of the “ Boarmine ” type; the flange plate is present although not obvious unless a close inspection is made. The cremaster is bifid (Fig. 7b). In the genera Azelina, Sestra and Declana the flange plate is absent, and there are four pairs of hooked setae, the apical pair being greatly longer and stronger than the others (Fig. 7a). These genera are of the “ Ennomine ” type. Pupae of Hybernia and Gargaphia were not available.

Specific differences are small, but may be of taxonomic use; there is variation between species in the position of the antenna, haustellum and leg-apices in relation

to each other, but the differences are slight. Selidosema panagrata pupae can be readily identified by the presense of a reniform spiracular groove on A 5. In all the genera examined, the transverse dorsal groove and the lateral groove delimiting the cremaster are of variable shape; the transverse groove is invariably irregularly crenulate and minutely pubescent.

The pupal cremaster (Fig. 7a, b) bears a strong resemblance to the anal shield, and on position only, the setae of the pupal cremaster and the larval anal shield are considered homologous. In the “ Ennomine ” type of pupa, setae SDi, D x , L x and D 2 are present; in the New Zealand genera the D 2 setae form the apical prong. In the “ Boarmine ” type, the setae SD X , D x , and L x are lost, only D 2 remaining to form the bifid cremaster apex.

As do the pupal characters, so do the adult genitalic characters bear out the consistency of the characters used to classify the larvae of New Zealand Ennominae. The genitalia of the Ennomine spp. listed in the introduction were examined except for the female of Declana egregia; in addition the genitalia of the following species were examined (*denotes female not examined) :

Selidosema adusta Phlp.*; S. argentaria Phlp.; S. fluminea Phlp.*; S. lupinata (Feld.)*; S. omhrodes Meyr.*; S. terrena Phlp.*; Declana callista Salm.*; Azelina nelsonaria Feld. The male genitalia of Hybernia indocilis, Selidosema suavis, S. productata, Sestra flexata and Declana griseata are shown in Fig. 9 a-e; the female genitalia of Declana leptomera and Selidosema suavis are shown in Fig. 10a, b. These six species show most of the characteristic features of their genera or groups of genera. Terminology of the structures is that advocated in Tuxen’s (1956) glossary.

Pierce (1914) distinguishes eight tribes in the Ennominae, each distinguishable from the other by differences in the genitalia: Ennomini, Macarini, Ourapterygini, Bistonini, Boarmiini, Erannini, Gnophini and the Abraxini. Of these tribes, the Boarmiini, Ennomini and possibly Erannini are represented in the New Zealand fauna. The characters selected by Pierce to distinguish these three tribes are:

(a) Ennomini : Gnathos dentate; valves generally broad, with setose areas on their inner surfaces; furca (where present) forked; uncus strong; coremata (a bush of long hairs) on A 9; females with a dentate signum on the bursa copulatrix.

(b) Boarmiini: uncus strong; gnathos strong; valves of variable shape with spiny armature; female with ovipositor retractile; signum dentate.

(c) Erannini : foreshortened uncus with two apical points; valve spinose, juxta a thin bifurcate plate.

The genera Declana, Sestra, Azelina and Gargaphia have similar male genitalia; this group is not excluded by Pierce’s definition of the Ennomini. The genus Selidosema (including S. dejectaria) have genitalia which are typical of Pierce’s Boarmiini, which includes European species of Selidosema. The genus Hybernia, while lacking the double-apexed uncus, otherwise resemble Pierce’s figures of European Erannini.

The genitalic characters of the New Zealand Ennomini, Boarmiini and Erannini are listed below.

Ennomini : Males: uncus elongate, sparsely haired beneath, base small; socii present or absent (see below) ; gnathos strong, either billhook shaped, or with expanded, spinose apices joined by a transverse bar, gnathos extending to near costal process; valves long, broad, with the cucullus as a setose band extending alongside the costa; no valvular or saccular processes; inner costal process (“ transtilla ”) strongly sclerotised, often massive, scobinate, or spinose, usually joining in midline or greatly curved outwards (in which case uncus is short, socii absent, cucullus is terminal and a true transtilla is present). Juxta plate-like, furca forked,

or (rarely) minute, generally massive, scobinate; furcal brushes generally present. Females with the ovipositor broad or slender, incapable of much protrusion; anterior and posterior vaginal lamellae fused, reaching laterally to near the apophysial arms of the A g tergite, strongly or weakly sclerotised; A 8 sternite often sclerotised; ductus bursae strongly sclerotised, smooth or grooved, straight or twisted or looped; colliculum present or absent; signum variable, dentate, scobinate or star shaped.

In the genus Declana, excluding D. niveata, D. floccosa and D. callista, the uncus is elongate, the juxta is well developed and the furcal apices are of varied form. Furcal brushes are present, and the inner costal process is directed towards the midline. In D. niveata the socii are present, as in other Declana species except in D. floccosa and D. callista. In these three species, the inner costal process is directed outwards along and/or across the valve; the cucullus is terminal. In all Declana species the sacculus is strong.

In Azelina and Sestra species the sacculus is weak and short. Azelina species have the furca plate-like and the inner costal processes elaborate, and Sestra species have the furca elongate, scobinate or spinose and the inner costal processes triangular and weak. Gargaphia muriferata has the uncus spathulate, the socii almost indistinguishable, and the apex of the furca bearing 4-6 long, curved, flat bristles, each as long as the valve and each with a crotchet hook at the apex. The valve is oval, the costa is expanded, and the cucullus is reduced to a small bristly area by the apex of the costa.

Boarmiini : Males: uncus short, apex hooked, bifid or trifid, bare beneath; base of the uncus swollen; socii present as indistinct hairy patches in some species; the costal process weak, never joined in midline, simple; gnathos arms short, bent, joined at apices, scobinate or smooth. Valves acute or blunt at apex; costa strong, ampulla spinose, saccular process present, spinose or setose; cucullus usually a strip or patch of stiff setae by costa on apical area of valve. Juxta elongate, anellus furcate or blunt tipped; furcal area and furcal brush undeveloped. No coremata. Females : ovipositor capable of protrusion for a distance equal to twice the length of A 8; anterior vaginal lamella usually weak or absent, (except in S. dejectaria) ; colliculum present, by antrum; ductus bursae narrow, straight, weakly sclerotised (except in S. deject aria ) signum star-shaped or absent.

Excluding S. dejectaria, the New Zealand Selidosema species have male genitalia with a prominent costa which ends either before the valve apex (S. suavis (Fig. 9b), S. adjusta, S. fenerata, S. argentaria) or at the valve apex (e.g., S. productata, Fig. 9c). S. panagrata has rounded valves; in many species there are elaborate ampullae and/or saccular processes.

Erannini : The sole New Zealand species, Hybernia indocilis, has the following characters. Male: uncus short, with apex hooked, gnathos short, elbowed; socii present; valvae elongate, bent; inner costal process absent; cucullus apical, setose; sacculus strong; no valvular or saccular processes; juxta plate-like, diamond shaped; furca and furcal brushes undeveloped. Female: Anterior vaginal lamella present as a sclerotised semicircle; ovipositor capable of extrusion for a distance equal to at least /s the length of the abdomen at rest; colliculum present, split; ductus bursae slim, not heavily sclerotised, widening gradually and slightly to join the elongate bursa copulatrix. Signum absent.

In the absence of similar information on Australian Ennominae it is impossible to relate closely any New Zealand genus to the Australian fauna.

Both the larva and adult of Selidosema dejectaria present characters that are not common to other New Zealand Selidosema species. The male genitalia of S. dejectaria lacks socii, the valvae are oblong and there is a hollow finger-like backward-projecting saccular process. The costal process is finger-like and projects

into the midline. The cucullus is represented by a broad setulose area covering most of the valva, and the base of the costa is bordered by a dense group of blunt rods. The apex of the sacculus projects at right angles to the valve. The female genitalia resembles the Declana type, having the ductus bursae broad, heavily sclerotised and grooved, and the anterior and posterior vaginal lamellae sclerotised and elaborate. The ovipositor resembles the Selidosema type in that it is protrusible. The first instar S. dejectaria differs from other Selidosema larvae as outlined in the preceding sections in the pigmentation pattern, chaetotaxy, crotchet arrangement and form of the antennal pit. There are no observable basic differences in pupal structure.

Abbreviations Used in Figs. 1-10

(a) Setae.—D l 5 D 2 , dorsal setae; L l 4, lateral setae; SDj.2, subdorsal setae; SV 1 _ 3 , subventral setae; V, ventral setae; CD, setae on posterior margin of anal proleg; CP, setae on anterior margin of anal proleg.

(b) Other Structures.—Anell, anellus of the juxta (male genitalia); ant. ap., anterior apophysis (female genitalia); ant. p., antennal pit (larva); antr., antrum (female genitalia); amp., ampulla (male genitalia); co., ventral condyle (mandibles); G., costa (male genitalia); coll., colliculum (female genitalia); cor., coremata (male genitalia); crp. b., corpus bursae (female genitalia); cu., cucullus (male genitalia); du. bur., ductus bursae (female genitalia); du. sem., ductus seminalis (female genitalia); dors., dorsal; fu., furca; gn., gnathos (male genitalia); hp., hypoproct (larva); j., juxta (male genitalia); l.ant., lamella anterovaginalis (female genitalia); 1. post., lamella posterovaginalis (female genitalia); 1.g., lateral groove (pupa); oc., ocellus (larva); pap.a., papilla analis; p.ap., posterior apophysis (female genitalia); pp., paraproct (larva); sig., signum (female genitalia); si., sacculus (male genitalia); sl.p., sacculus process (male genitalia); soc., socii (male genitalia); st., sternite; terg., tergite; t.g., transverse groove (pupa); t., transtilla (male genitalia); vent., ventral; vu., valvula (male genitalia).

References

Chappell, A. V., 1929. Biological notes on N.Z. Lepidoptera. Trans. & Proc. N.Z. Inst. 60: 2, pp. 259-262.

Ci-ark, A. F., 1935. The Winter Moth (Hybernia indocilis Walk.). N.Z. Journ. Sci. & Tech. 17: 3, pp. 541-549.

Forbes, W. T. M., 1945. The Ennomine pupa (Lepidoptera: Geometridae). Journ. N.Y. ent. Soc. 53: 3, pp. 177-210.

Hardwick, D. F., 1958. Taxonomy, Life History and Habits of the Elliptoid-Eyed Species of Schinia (Lepidoptera: Noctuidae), with notes on the Heliothidinae. Canad. Ento. Supplement 6; 116 pp., accompanies Vol. 90.

Hinton, H. E., 1946. On the homology and nomenclature of the setae of Lepidopterous larvae and some notes on the phylogeny of the Lepidoptera. Trans. R. ent. Soc., Lend. 97: pp. 1-37.

Hudson, G. V., 1926. Butterflies and Moths of New Zealand. Ferguson and Osborne, Wellington.

Osborne., Wellington.

MgGuffin, W. G., 1950. Descriptions of Larvae of Forest Insects ( Plagodis, Anagoga, Hyperitis (Lepidoptera: Geometridae) ). Canad. Ento. 82: 10, pp. 205-210.

Canad. Ento. 88: 1, pp. 6-16.

on other stages. Canad. Ento. 90: 4, pp. 216-222,

Ento. Supplement 8. 104 pp., accompanies Vol. 90.

McKay, M. R., 1959. Larvae of the North American Olethreutidae (Lepidoptera). Canad. Ento. Supplement 10, 338 pp., accompanies Vol. 91.

Pierce, F. N., 1914. Genitalia of British Geometridae, Oundle. Purdie, A., 1884. Larvae of N.Z. Geometrina, etc., N.Z. Journ. Sci. 1: 3, pp. 60-3. Schimitsghek, E., 1958. Boarmia bistortata Goeze, als Larchenschadling {B. bistortata

Goeze as a defoliator on Larch). Z. Angew. Ent. 40: 1, pp. 37-51. Singh, Balwant, 1953. Immature stages of Indian Lepidoptera: No. 8; Geometridae. Ind. For. Rec. (Ent.) 8 (7), pp. 67-158 (Pt. 8).

ponents of various groups of setae. Ind. For. Rec. (Ent.) 9: 6, pp. 131-163. Tuxen, S. L. (ed.) 1956. Taxonomists’ Glossary of Genitalia in Insects, ed. S. L. Tuxen, Copenhagen; esp. pp. 97-111.

J. S. Dugdale, N.Z. Forest Service, Forest Research Institute, Rotorua.

MH Enno. Lar. Asth. 1 Oeno. Me s Me s Me s Me s 4 (S) l 3 (S) L, SV 3 sv 3 p 3 l 3 SV 3 SV 3 l 3 L 3 sv 1 1— SV, L Dors. (S) l 4 p 3 Ld l Ld l — P 3 Ld Al (S) sv 3 sv, SV 2 — — — — l 4 _** (P) SV 2 — — SV 2 sv 2 — — —* _** (P) sv. sv 2 SVi SVi SVi SV 2 — sv 2 SV 1 ■„ \.— (P) V r (P) V t V t v v Vi ViVi V v Vi Vi V r Vi — - —Vi -V,Vi Vi Vent. Vent. 1 l 3 sv l 3 sv 3 L 3 5F — SV 3 L Dors. A2-5 l 4 p 3 Ld l 3 Ld La — L Ld sv 3 L sv — — — — l 4 _** } sv 2 sv 2 sv. sv 2 sv 2 sv 2 — SV, sv SVi SVi SV 3 sv 4 SVi sv, — sv sv n V-, ■ V 1 Vi Vi Vi Vi ? Vi Vi Vent.

Table I.

Seg. No. of setae No. larvae No. larvae Normal Normal symm. asymm. asymm. arr. symm. arr. meson. symm. asymm. asymm. arr. symm. arr. 0 0 1 A1 1 2 7 35 14 13 2 + 1 2 + 2 0 0 0 A2 1 2 3 0 41 0 7 15 8 2 + 3 2 + 2 A3 3 4 43 2 11 11 3+4 3 + 3 2 0 1 A4 3 34 15 3 + 4 3 + 3 4 6 16 1 0 1 A5 2 0 3 3 + 4 3 + 3 3 34 14 4 6 13

Table II.—SYMMETRY AND ASYMMETRY IN THE SV 3 GROUP ON Al-5 OF SELIDOSEMA SUAVIS (N = 56).