Click Beetles: Genera of the Australian Elateridae (Coleoptera)

By AA Calder

This is the first monograph on the Australian genera of Elateridae - click beetles. The book deals with 74 genera, among which 14 are newly recognised.

The volume documents the entire Australian fauna and provides lavish illustrations of representative species, and typical examples of the male and female genitalia for each genus. The phylogeny of the genera is analysed and there is a checklist of all described species and appropriate bibliographic and type locality details are given.

• Language: English
• Publisher: CSIRO PUBLISHING
• Release date: Jan 1, 1996
• ISBN: 9780643105874

Book preview

Chapter 1

INTRODUCTION

Generalities

The Elateridae is a diverse group of beetles found throughout Australia in all vegetation types. It contains some 670 described species in 70 genera previously arranged in as many as 18 subfamilies (see Neboiss 1956). The family name dates from the time of Leach (1815) and is based upon the Greek word for a hurler or driver— (elater). In the English speaking world the adult’s ability to click has led to several common names: Click beetle, Skip-jack, Spring-beetle, Clicker, or Blacksmith, while Swammerdam calls them Grasshopper or Locust-beetles and in German they are known as Schnellkäfer or Schmiede.

Adults are primarily nocturnal hiding during the day under stones on the ground, in leaf litter, in dead tree stumps or under the bark of gum trees (Eucalyptus). These nocturnal species are readily attracted to light where they can be collected in considerable numbers. Several species are diurnal and can be found upon flowers, foliage, tree stumps or even clustered at the tips of grass stems in forest clearings. Many diurnal species when in danger will fall to the ground in the first instance to avoid capture. Unusually, adults of the negastriine Rivulicola are riparian often being encountered crawling over sand on the banks of rivers and creeks where it is thought their larval stages occur.

Adult elaterids are characteristically elongated and usually have well developed, pointed prothoracic hind angles. They are usually recognised by their ability to jump into the air, holding their antennae and legs close to the body and making an audible clicking noise in the process. The legs taking no part in the jumping mechanism. Prior to jumping the body is arched, this movement being facilitated by the loose articulation between the prothorax and mesothorax. Normally, when a beetle is not preparing to jump, the prosternal spine fits part way into the mesosternal cavity and the prothorax and mesothorax fit tightly together. The jump-clicking mechanism involves a jack-knifing movement whereby the prosternal spine slides rapidly down a smooth track into the mesosternal cavity causing a sudden movement of the prothorax relative to the hind body. The muscles involved in this movement are able to build-up tension due to a friction hold formed by a small notch on the dorsal side of the prosternal spine resting on the anterior margin of the mesosternal cavity. The tension is released when the tip of the prosternal spine slips over the lip of the mesosternum and slides into the mesosternal cavity causing the body to jack-knife. The principal muscles that produce the jump are the dorsal longitudinal intersegmental muscles, extending from the pronotum to the inflexed anterior margin of the mesonotum, which Larsen (1966) has designated M4. These are the largest muscles in the prothorax and account for one third of the weight of the head and prothorax. The main antagonistic muscle involved is the corresponding medio-dorsal one (M2a and b of Larsen) (Evans, 1972). The jump of a click beetle is considered by many authors to be a means for beetles that have fallen on their backs to right themselves, whereas Crowson (1981) thinks that it could equally be considered as a defensive mechanism.

The larvae are more or less characteristic being elongate, cylindrical to flattened, usually evenly sclerotised dorsally and ventrally, with or without an apically forked ninth abdominal segment and with a heavily sclerotised head with strong mandibles and well developed maxillae, labium and densely setose oral filter that is associated with extraoral digestion. The slender, cylindrical body form with tough cuticle of many soil dwelling species has resulted in the collective term of wireworm. They are generally yellow, brownish-yellow or dark brown to black in colour. The French know the larval stage either as vers fils de fer or vers jaunes, while the Germans refer to the larvae as Drahtwürmer. Larvae are for the most part to be regarded as omnivorous relying on other insect larvae and various invertebrates, as well as feeding upon organic and other vegetable matter. The larval stages can be collected from under the bark of trees, in rotten wood or from the soil. As identification is based solely on the adult form it is often necessary to rear wireworms through to the adult stage. Alternatively if fully mature adults can be found in their pupal cells removal to the laboratory for eclosion is quite successful in obtaining the adult. Species from both rotten wood and soil can be successfully reared by placing a single larva (to prevent cannabilism) in a container with sufficient quantity of substrate. It is necessary to occasionally add a few drops of water to maintain the moisture level of the original larval substrate, being most careful not to allow the contents to become too dry or sodden. The larval diet of both rotten wood and soil dwelling larvae can be supplemented by occasionally adding either moist dog or cat food wrapped in tissue for the larva to feed upon. Conoderus species collected from backyard compost heaps have been successfully reared in such conditions. The author has also reared several Agrypnus species by placing the second or later instar in containers whose bottoms have been filled with plaster of Paris, into which the larva invariably burrowed. The duration of the larval stage is not fixed as such but is highly dependent on the living conditions and availability of suitable food. The larvae of Cardiophorinae are soil predators and have an unusual appearance being soft, flaccid and elongate with apparent triple segmentation of the abdominal segments. Cardiophorines are also unusual in possessing deeply cleft mandibles and the terminal abdominal segment is furnished with anal papillae.

Grassland and pasture are the natural habitat of the larval (wireworm) stage of numerous click beetle species, thus it is not surprising that some have become agricultural pests. Recorded pest species belong to the two speciose genera Agrypnus and Conoderus, as well as Arachnodima, Heteroderes and Hapatesus. Agrypnus variabilis (Candèze) has long been a pest of the establishing phase of crops such as sugar cane (McDougall, 1934a, 1934b), maize (Forrester et al., 1984), sorghum and sunflowers (Gunning & Forrester, 1984). Two wireworm species from South Australia are known to attack cereal crops (Allen, 1968). These have now been identified as Arachnodima opaca Candèze which damages barley crops on the Yorke Peninsula, and A. ourapilla which attacks germinating wheat seeds and seedlings on the Eyre Peninsula. Arachnodima bribbarensis (Calder) and A. xenikon (Calder) also attack germinating wheat seeds and seedlings in southern New South Wales (Calder, 1986). In the case of A. xenikon the area under cultivation had carried improved pasture for the previous 10 years. Unidentified species of Conoderus have been reported eating the eyes of cane setts in central Queensland (Bates, 1925; Jarvis, 1916) and an undescribed species of Heteroderes attacks germinating sugar cane in the Bundaberg district of Queensland (Calder, 1990). Hapatesus hirtus Candèze causes sporadic damage to the Victorian potato crop (Neboiss, 1962; Smith, 1980).

Elaterids are defined as possessing the following character states. ADULT: Head: antenna 11-segmented, occasionally 12-segmented. Labrum exposed. Thorax: hind pronotal angles well developed, projecting. Procoxae globular, not projecting, with highly reduced and concealed trochantins. Procoxal cavities open both internally and externally. Posterior margin of prosternum elongated to form a spine. Mesosternum with a well developed cavity, that with the prosternal spine form the characteristic jump-clicking mechanism. Mesocoxae usually widely separated, rarely close together. Hind coxal plates well developed. Tarsi each with 5 tarsomeres. Abdomen with 5 ventrites, basal 4 (sternites 3–6) connate.

LARVA: Head deeply pigmented, prognathous and protracted. Stemmata either absent or one on each side of the head. Antenna 3-segmented, segment 2 with conical sensory papillae apically. Frontoclypeal suture absent. Labrum fused with frons to form the nasale. Nasale unidentate or tridentate. Mandibles stout, usually with retinaculum, mola absent; mandibles of Cardiophorinae deeply divided, dorsal lobe toothed. Maxillolabial complex (ventral mouthparts) retracted, maxilla either with a single mala or both galea and lacinia. Submentum triangular or rectangular. Maxillary palp 3 or 4-segmented. Labial palp 2-segmented. Hypopharyngeal sclerome absent or present. Hypostomal rods absent. Ventral epicranial ridges present.

Thorax with 3 subequal segments. Legs well developed, 5-segmented, with claw-like tarsungulus. Pretarsal setae absent, reduced or 2 in number.

Abdomen with 9 segments dorsally (Cardiophorinae have secondary segmentation on segments 1–8). Segment 9 variable. Urogomphi present or absent; if present urogomphus usually divided into an inner and outer prong, rarely undivided. Caudal notch variable in size. Segment 10 located on segment 9 ventrally, sometimes with sclerotised teeth around the anus; anal papillae present in Cardiophorinae.

Spiracle biforous, without closing apparatus and with ecdysial scar.

Family definition

As a group adult elaterids are extremely homogeneous in external appearance. This has led previous workers to rely almost exclusively on the external ‘gestalt’ of the beetle rather than critically analysing character states when assigning a new species to a genus and subfamily. The family Elateridae is not an easy one to define owing to the apparent correlation of the evolution in the adult of a unique startle defense mechanism - the ‘jump-clicking’ apparatus which has resulted in a very distinctive and easily recognisable body form. This has also severely constrained the pathways of adult variation and increased the incidence of homoplasy within the group. Thus many authors have experienced considerable difficulty in fixing the limits of the family. However, the vast majority of elaterids cannot be confused with any other family due to the following combination of characters - an exposed labrum, projecting hind pronotal angles, a long prosternum, globular procoxae with highly reduced concealed trochantins, well developed hind coxal plates and first four ventrites (sternites 3-6) connate.

The limits of the family Elateridae are open to question, particularly with respect to the few related families whose members share the clicking mechanism, or have lost it secondarily. The families with which confusion is most likely to arise are Cerophytidae, Eucnemidae and especially Throscidae. Both van Emden (1932) and Gardner (1936) suggested that the throscid subfamily Thylacosterninae (as Balginae) be placed in Elateridae on the basis of the larvae of Pterotarsus Guérin-Méneville (as Lissothyreus Bonvouloir) and Cussolenis Fleutiaux. Böving & Craighead (1931) described the larva of another throscid genus, Drapetes Dejean, and proposed its transfer to the elaterid subfamily Oestodinae, a move that was supported by Crowson (1961) on the basis of both larval and adult characters. Burakowski (1975) and Costa et al. (1988) considered both Drapetes and Lissomus to be members of a separate family Lissomidae, with links to both Elateridae and Throscidae. However, the Oestodinae (Hyslop, 1917) along with the Protelaterinae (Schwarz, 1902e) and Lissominae (Laporte, 1840) were considered to form a monophyletic group of Elateridae by Calder et al. (1993).

In Australia, several taxa formerly associated with other elateroid families, are considered to be elaterids in this work. Firstly the Thylacosterninae (=Balginae), which includes Cussolenis, resemble eucnemids and were once included in that group. Secondly, the Lissominae in its original sense including Drapetes, Paradrapetes Fleutiaux and Lissomus which were transferred from the Throscidae. This subfamily now includes Austrelater and Osslimus as well. Thirdly, two doubtful elaterids, formerly included in the Eucnemidae, Anischia Fleutiaux (Anischiinae) and Subprotelater Fleutiaux (Subprotelaterinae). Anischia, is somewhat throscid-like with clavate antennae and was once included in the Cerophytidae based on the absence of hind coxal plates. Subprotelater is decidedly eucnemid-like with antennal grooves present on the hypomeron similar to those found in eucnemids but it also has a sclerotised, exposed labrum and a free ventrite 5, which are typical of a majority of Elateridae. These two genera have not been treated in this study.

Historical review

W.S. Macleay (1826) described the first two Australian elaterids– Elater xanthomus and Elater nigroterminatus based upon specimens collected by Captain P.P. King during his survey of the northern coastline of Australia. Elater xanthomus has since been transferred to the genus Anilicus Candèze, and Elater nigroterminatus has been synonymised with Melanoxanthus melanocephalus (Fabricius, 1781), a species not usually met with in Australia, although a closely related new species is commonly collected on Cape York Peninsula.

Between 1829 and 1888 twelve authors named numerous click beetles. Eschscholtz (1829) described two species of Monocrepidius (now Conoderus) and a Ludius (now Dicteniophorus). Guérin-Méneville (1830, 1838) described two species collected on the voyage of the Coquille: Elater scapularis (now Conoderus) and Adelocera grisea (=Agrypnus caliginosus (Boisduval). Boisduval (1835) described 10 species, 4 of which were subsequently synonymised, collected on the voyage of the Astrolabe under the command of Dumont D’Urville. Hope (1834, 1842a, b, 1845) described 6 species, erecting the genus Macromalocera (1834) for two species of unusual elaterids whose males had elongate antennae that exceeded the apex of the elytra, giving them the appearance of cerambycids. Gory (1836) described the tetralobine Tetralobus australasiae. Erichson (1842) described 15 species collected in Tasmania including 5 Conoderus (as Monocrepidius), and two new genera, Crepidomenus and Atelopus (a preoccupied name that was later replaced by Acroniopus Erichson, 1843). LeGuillou (1844) described 3 Conoderus species (as Monocrepidius). Germar (1844, 1848) described 4 species mostly from the environs of Adelaide. Blanchard (1853) named 2 species: Monocrepidius cinereus now synonymised with Conoderus leluti LeGuillou and Agriotes quadripunctatus, which is undoubtedly a Paracardiophorus judging by the illustration given, both collected at Raffles Bay, Cobourg Peninsula, Northern Territory. Newman (1857) described 2 species Dorcastoma jansoni (now synonymised with Aphileus lucanoides Candèze) and Alaus gibboni (=Paracalais) from specimens collected by James Gibbon at Moreton Bay, Queensland. Boheman (1858) described Crepidomenus luteipes collected at Sydney during the voyage of the Eugenies. W.J. Macleay (1872, 1888) in two papers described 71 species from material collected at Gayndah, Queensland and from the vicinity of King Sound (Derby), north-west Western Australia respectively.

The first serious attempt to deal comprehensively with the taxonomy of the Elateridae was that of Candèze. During the period from 1857 to 1863 he monographed the group in four volumes entitled Monographie des Élatérides and a single supplementary volume (1874) in which numerous genera and species were described, many of which were Australian. A classification of the family was proposed and a key was also provided for the identification of the known genera.

Then followed a period of 50 years in which five main authors described numerous Australian species, but neglected to provide any means of identifying the genera, other than by reference to closely related taxa. None of these authors provided a key to the genera, an essential step in providing a means of identification for the serious student. Blackburn (1889-1912) described 54 Australian species and two unusual genera from the arid interior of the continent — Antoligostethus and Nullarborica. Schwarz (1895-1907) named 93 species and Lea (1908-1930) 10 species. Elston (1924, 1927) revised the known species of Agrypnus (as Lacon) and (1928) the genus Pseudotetralobus. Carter (1935) described two new species of Glypheus and provided a key to all known species and in 1939 described some 50 new species including a revision of two large groups: Melanoxanthus and Paracardiophorus providing keys to all the species of these groups then known to him.

Recent work on the Australian elaterid fauna has been conducted by four authors. Neboiss (1956, 1961) provided the only relatively complete check list of described Australian elaterids this century and produced revisionary work on the following genera: Hapatesus and Toorongus (Neboiss, 1957); Aphileus (Neboiss, 1959); Elatichrosis and Lingana (Neboiss, 1960); Paracalais and Austrocalais (Neboiss, 1967); Ophidius and Yalganus (Neboiss, 1975). Hayek (1973, 1979) reviewed all species of the subfamily Agrypninae excluding the Tetralobini, redefining all included genera on the basis of new characters and assigning all described species to an appropriate genus. A key to the included genera was provided. Gullan (1977) revised the genus Anilicus Candèze and erected the genus Augenotus to accommodate Anilicus quadriguttatus (Erichson). Calder (1983) recorded the presence of Melanotus in Australia. In later publications the subfamily Crepidomeninae was revised (Calder, 1975) and two new genera erected (one of which, Orodina, is synonymised in this study), the genus Anthracalaus was noted from Australia for the first time since the original description of an Australian species by Fleutiaux (Calder, 1990) and Anthracalaus was revised (Calder & Hayek, 1992) on a world-wide basis in which two new species from Western Australia and Queensland were described. The Australian Pityobiinae were reviewed (Calder, 1992) and the genera Parablax and Wynarka (formerly in Crepidomeninae) and Parasaphes (formerly in Denticollinae (as Hemicrepidiinae) were transferred. In 1993, Calder et al. described the unusual elaterid Austrelater, which has a larva similar to those of lissomines, and included it in an expanded Lissominae along with Drapetes, Oestodes LeConte, Sphaenelater Schwarz, Protelater Sharp and Lissomus Dalman and, by implication, the related genera Anaspasis Candèze, Paradrapetes Fleutiaux and Hypochaetes Bonvouloir.

Higher classification

The first major attempt at a classification of the Elateridae was by Candèze (1857, 1859, 1860, 1863) who divided the family into eight tribes (Agrypnides, Melanactides, Hémirhipides, Chalcolépidiides, Oxynoptérides, Tétralobides, Élatérides vrais, and Campylides), which were also used by Lacordaire (1857) in his Genera des Coléoptères. This classification was based on characters derived from the metasternum, prosternum, mesepimeron, mandibles and antennae. The Élatérides vrais contained a majority of the click beetles then known and encompassed both the presently defined Denticollinae and Elaterinae. In his introduction to the classification of the family Élatérides Lacordaire lamented: ’La classification de la famille présente des difficultés excessives et peut-étre insolubles. Dans celle qui suit, et dont je suis loin d’être satisfait, les huit groupes ou tribus que j’ai era devoir établir passent insensiblement de l’un à l’autre.’ Even today this state of affairs has not changed markedly, with the higher classification still in a state of flux and considerable confusion surrounding the definition of higher taxa.

Since the publication of Genera Insectorum (Schwarz 1906b, c; 1907a), in which 28 tribes were recognised largely based upon the previous work of Candèze but with a number of subtribes formerly included in the Élatérides vrais raised to tribal rank, the only major efforts at a higher classification of the elaterids have been by Hyslop (1917), Fleutiaux (1947), Nakane & Kishii (1956), Crowson (1961), Gur’jeva (1974), Dolin (1975) and Stibick (1979).

Hyslop (1917) proposed a classification of the Elateridae based upon larval characteristics which differed radically from previously existing systems based on adults. Three subfamilies: Agrypninae (as Pyrophorinae and which also included members of the present day Denticollinae), Elaterinae and Cardiophorinae were recognised, which roughly corresponded to three larval types. A fourth subfamily was listed by Hyslop, the Physodactylinae, but no larval characters were known.

Fleutiaux (1947), studying the south-east asian fauna and using adult characters, recognised 23 subfamilies, including the aberrant monotypic subfamily Anischiinae. These subfamilies were largely based upon Schwarz’s tribes which Fleutiaux had raised to the rank of subfamily. Nakane & Kishii (1956), using male genitalic characters derived from a study of the Japanese fauna, recognised 16 subfamilies that included the newly erected Negastriinae. This subfamily was based upon species with outcurved prosternal sutures, a large prosternum, a closed mesocoxal cavity and characteristic aedeagus.

Crowson (1961) used the number of sclerites in the apical part of the hind wing, along with supplementary adult and larval characters, to divide the Elateridae into 6 subfamilies: Pyrophorinae (=Agrypninae), Cardiophorinae, Elaterinae, Pityobiinae, Corymbitinae (=Denticollinae pars) and Oestodinae.

Gur’jeva (1974), on the basis of thoracic characters alone, divided the Elateridae into ten major subfamilies (Agrypninae, Pityobiinae, Negastriinae, Tetralobinae, Oxynopterinae, Diminae, Athoinae, Oestodinae, Elaterinae, Cardiophorinae). This study was followed up by Dolin (1975) who used characters of hind wing venation together with structural features of the larval forms to propose the same ten subfamilies which differed only in the disposition of a few genera.

Stibick (1979) relying heavily on the characters of the adult, with some input from larval morphology, divided the family into 12 subfamilies. These included eight of the subfamilies recognised by Gur’jeva (1974) and Dolin (1975) except that Tetralobinae was regarded as a tribe of Agrypninae (as Pyrophorinae) and Diminae a subtribe of the Denticollinae-Denticollini. Stibick also recognised the subfamilies Melanactinae, Hypnoidinae, Aplastinae and Melanotinae (a tribe of the Elaterinae).

I have chosen to base the classification on eight subfamilies Lissominae, Thylacosterninae, Agrypninae, Denticollinae, Pityobiinae, Elaterinae, Cardiophorinae and Negastriinae which are based upon those of the Russian authors. Any attempt at a tribal classification of the Australian Elateridae must await a study on a worldwide basis.

The genera are arranged in alphabetical order under each subfamily. Each of these subfamilies is not as easily defined as I would have preferred. The Denticollinae and Elaterinae are defined on the basis of wing venation characters and the morphology of the frontal region of the head. The shape of the frontal region of the head has long been used to define these subfamilies, and was found to be a useful if not difficult character when applied to the Australian fauna. Even though the shape of the frontal region of the head is difficult to ascertain with precision, in all elaterines included in this study the anterior region of the head is convex, whereas the Australian denticolline genera studied can have both a flattened anterior region or a more or less convex anterior region.

Among the Australian genera listed in the Neboiss Checklist a number have been transferred from the Elaterinae to the Denticollinae and vice versa. Carterelater (for Elastrus flavipes Macleay), Dicteniophorus, Doloporus and Drymelater (for Agonischius species) formerly included in the Elaterinae are transferred to the Denticollinae on the basis of the apex of the hind wing membrane possessing a single sclerotisation, although all but Drymelater have a head with a convex frontal region. Anilicoides, Antoligostethus, Diadysis, Lingana and Nullarborica formerly included in the Denticollinae are transferred to the Elaterinae on the basis of the frontal region of the head being convex and the apex of the hind wing membrane possessing three sclerotisations arranged in the form of the Greek letter epsilon [Ψ]. Patriciella has been moved from the Elaterinae to the Cardiophorinae on the basis of a closed mesocoxal cavity, cordate scutellum, short prosternal spine, pronotum with sublateral basal incisions and hind wing venation.

At the start of this project a number of Australian genera had been misplaced and many species incorrectly assigned to Holarctic genera creating a major source of confusion when attempting to identify Australian elaterids. In this study Australian species assigned to Limonius Eschscholtz have been transferred to Microdesmes Candèze, the sole Australian Agriotes Eschscholtz species to Paracardiophorus Schwarz, species of Elater Linnaeus to Conoderus or Paracardiophorus and species of Elastrus, Ctenicera Latreille and Agonischius Candèze assigned to new genera.

The purpose of this study is firstly, to study and redescribe in full all previously described Australian click beetle genera and describe any new genera; secondly to assign all described species to existing or new genera, the limits of which are defined and thirdly to provide a key to all described genera.

Chapter 2

MATERIAL AND METHODS

Materials

Material on which this revision is based is housed in numerous Australian and overseas institutions. These are listed below, together with the abbreviations used in the text. The Australian Museum, Sydney (AMSA)

Australian National Insect Collection, CSIRO, Canberra (ANIC)

Australian National Insect Collection (Macleay Museum Collection), CSIRO, Canberra ANIC (MACL)

Natural History Museum, London (BMNH)

Canadian National Collections, Biosystematic Research Institute, Agriculture Canada, Ottawa (CNC)

Deutsches Entomologisches Institut, Eberswalde-Finow (DEIE)

Institut Royal des Sciences Naturelles de Belgique, Brussels (IRSNB)

Macleay Museum, University of Sydney (MACL)

Museum für Naturkunde der Humboldt-Univesität zu Berlin (MNHB)

Muséum National d’Histoire Naturelle, Paris (MNHP)

Museum of Victoria, Melbourne (MVM)

Queensland Museum, Brisbane (QMB)

South Australian Museum, Adelaide (SAMA)

University of Queensland, Brisbane (UQIC)

Western Australian Museum, Perth (WAMP)

Dissecting techniques and illustrations

A specimen of which the genitalia or hindwings were to be removed was first relaxed by being placed directly into very hot water for up to 40 minutes, so that it could be removed from the pin or card. Usually if the whole specimen was to be disarticulated it was put directly into 10% KOH overnight for clearing prior to dissection. For dissection of the genitalia directly from a specimen that was not destined for dismemberment, in the case of a female the whole abdomen was removed from its junction with the metathorax and placed in cold 10% KOH overnight to be cleared. The internal reproductive organs were then easily exposed and removed from the abdominal cavity similar to the method described by Becker (1956). For the male genitalia, the terminal four or five segments were dissected directly from the specimen and placed in 10% KOH overnight. After maceration, the preparations were washed in distilled water and acidified ethanol before storage either in 80% ethanol or in small plastic microvials containing glycerol pinned beneath the appropriate specimen.

For illustration the aedeagus was temporarily mounted in glycerol on a slide and drawn with the aid of a camera lucida attached to a Wild M-8 stereomicroscope. For small preparations a Leitz Laborlux 12 research microscope with drawing prism was used.

For the female organs it was discovered that when they were transferred from the potassium hydroxide into water they swelled up. The genital tract and its associated structures was therefore drawn before they were stored. Since the genital tract and associated structures are mostly membranous, only the sclerotised spines and plates show up against a white background. The spermatheca, colleterial glands and the spermathecal gland were only visible against a dark background or reflected light adjusted to give the structures a bluish appearance.

Measurements

In the following generic descriptions several measuremaents are used. The ratio of the minimum distance between the eyes to the maximum width across the eyes is given to indicate relative eye size. This is the ocular index of Campbell & Marshall 1964. An index of eye prominence is also given for newly described species for newly erected genera. This index (EI) is obtained by subtracting the interocular head (frons) width from the maximum width of the head across the eyes and dividing the result by the maximum head width. Thus the larger the index the greater the degree of eye prominence. The pronotal index (PI) is obtained by dividing the length of the pronotum measured along the midline by the maximum width of the pronotum before the hind angles, and multiplying the quotient by 100. Thus indices larger than 100 indicate that the pronotum is longer than wide. The length of the elytron (EL) is measured along the suture of the right elytron from the humeral margin to apex on a plane where both base and apex are in focus. Body width is measured across the base of the closed elytra. The total length is EL plus the distance from the base of the pronotum to the frontal margin of the labrum along the midline, no matter what the degree of retraction of the head into the prothorax. The length of the aedeagus is measured along the midline from the apex of the median lobe to the anterior-most margin of the basal piece (phallobase) and the width is measured across the base of the parameres at a point level with the articulation of the parameres with the median lobe or in the case of the negastriines the maximum width. In the new species descriptions, for each measurement a range is given followed by the mean in parantheses. The number of specimens measured is also indicated.

Taxonomic procedure

Characters of specimens of the type species of each descibed genus were used to provisionally define the limits of that particular genus and the available specimens were then sorted into their major generic groupings. A preliminary survey of internal genitalic structures, particularly of the female, was then made and the generic groupings refined further by adding or deleting taxa when necessary. The species composition of previously proposed genera has been assessed by examining as many type specimens as possible and where not available specimens that have been compared with syntypic series or holotypes. The criteria were provided by as many external morphological features as possible (including the ‘classical’ features), but especially by those provided by the female genitalia. Even though I have been able to examine specimens of nearly all described species, in a work of this nature, it is inevitable that generalisations about the presence of some characters have to be made, particularly with regard to the genitalia. Thus it is inescapable that the following generic descriptions will, in the fullness of time, be emended.

Chapter 3

MORPHOLOGY

In this section I will only give comments on selected characters used in the generic descriptions. I do not intend this to be a comprehensive discourse on elaterid morphology, but rather an indication of what characters are considered to be of some importance at the generic level.

Head

Most Elateridae possess a head capsule that is more or less convex anteriorly with the mouth-parts directed antero-ventrally (Figs 293, 294). However, in the Denticollinae the anterior portion of the head capsule is flattened dorso-ventrally causing the mandibles to be directed more anteriorly, giving such elaterids a characteristic prognathous appearance (Fig. 192). This is the classic character for the separation of the denticollines from the elaterines. In the pityobiines Parablax, Parasaphes, Tasmanelater, Wynarka and Xuthelater the frontal region of the head is strongly produced fowards to completely cover the base of the labrum (Figs 266, 267). This condition is more marked in the males.

Figs 1, 2.

Frontal view of head: 1, Conoderus sp. (Wanniassa, ACT). 2, Agrypnus caliginosus (Boisduval) (Wanniassa, ACT). Scale lines: 1, 1.0 mm; 2, 0.1 mm.

The head is always narrower than the pronotum and in many specimens it is withdrawn into the thorax. The eyes are mostly uninformative for identification purposes except were they are enlarged and strongly protuberant (as in Austrelater) (Fig. 18) or not protuberant at all and almost not visible from the front (as in Agrypnus) (Fig. 2).

Above each eye there is usually a supra-antennal carina which when extended mesally to meet midway across the front of the head produces a frontal carina between and above the antennal sockets (Fig. 1). In genera where there is no carina above the antennal insertion the frontal carina is said to be obsolete.

The area between the frontal carina and the labrum has been referred to as the epistome by numerous authors (eg. Candèze) or nasale by Quate & Thompson (1967) and Hayek (1990). This area is particularly well developed in genera such as Melanotus (Fig. 350) and Rousia (Fig. 230). In many genera the frontoclypeal region rather than being wide and at an angle to the plane of the labrum is narrow and gradually slopes to the base of the labrum as in Agrypnus (Fig. 2).

The antennae of most elaterids are inserted into crescent-shaped sockets that are flush with the surface of the head capsule (Figs 2, 3). However, in the lissomines (Drapetes (Fig. 42) and Osslimus) and thylacosternines (Cussolenis (Fig. 33), the antennal insertions are enlarged to form large saucer-shaped depressions that are carinate both dorsally and ventrally but not along the outer eye margin. This usually results in the front of the head having an area the shape of an hourglass. A similar condition is seen in the species of Parasaphes (Fig. 266) and Rousia (Fig. 230).

The mandibles are usually bidentate although a number of genera have simple unidentate mandibles (Fig. 169). The bidentate mandibles can be divided into those whose apex is broad, with teeth perpendicular to the plane of movement, as in Austrelater (Fig. 19) and related lissomines, and those representing the majority of elaterids, whose teeth are usually oblique to the plane of movement with only one apical tooth and a second tooth (sometimes weakly developed) located some distance from the apex on the incisor edge.

The antennae are normally II-segmented and serrate or subserrate. Antennae are considered subserrate if the outer apical angle is broadly rounded rather than acute, in which case the antennae are distinctly serrate, being shaped more like saw teeth. In the genera Ascesis, Austrelater, Cussolenis, Dicteniophorus, Drymelater, Paranilicus, Pseudotetralobus, Stichotomus, Tetrigus, and Wardulupicola the antennae are pectinate. In the genera Pseudotetralobus and Wardulupicola, the antennae are also 12-segmented. In some cases (Austrelater, Dicteniophorus, Drymelater, Pseudotetralobus and Wardulupicola) which have dimorphic sexes, the female has subpectinate or serrate antennae. The length of the antennae and the relative size of the second, third and fourth antennomeres has been noted in the generic descriptions, but it should be treated with caution as it is extremely variable. The antennae of the males is without exception longer than those of the female. Another character that is of some use in delimiting genera is the appearance of the first, second, third and fourth antennomeres. The surface of these segments is usually smooth and devoid of punctures and setae and the subsequent segments are densely clothed with short setae. Whether the vestiture begins on the third or fourth segments is useful in characterising some genera. The median longitudinal carina on the flat surface of the basal antennomeres varies from an indistinct carina that is only visible when the specimen is held at an angle to the incident light under the microscope to a clearly visible, strongly raised, narrow ridge. The species of Paracrepidomenus have only the third antennomere with such a carina, in Conoderus species only the fourth antennomere is carinate, and in Enischnelater both the third and fourth antennomeres have a median longitudinal carina on the flat surface. In Pseudaeolus and Simodactylus this median carina extends from 4 to 8 and 4 to 10, respectively.

Thorax

The prothorax varies from being longer than wide to wider than long and is useful for delimiting some genera. Genera having the pronotum wider than long include Drapetes, Osslimus, Agrypnus, Aphileus, Trieres, Cardiotarsus, Paracardiophorus, Diadysis, Hapatesus, Anchastus, Antoligostethus, Nullarborica and Tyiwiphila.

The integument is usually provided with similar sized punctures, however in Heteroderes (Figs 96, 97) and Aeoloderma the pronotal punctures fall into two different size classes. This gives the pronotum a distinctive appearance and is termed double punctation by many authors.

The lateral carina separating the pronotum from the hypomeron varies from being complete and extending the entire length of the prothorax to being completely absent. In genera such as Rivulicola (Fig. 414), the cardiophorines Cardiotarsus (Fig. 390) and Paracardiophorus, Carterelater, Corystelater, Stichotomus and Wynarka the lateral carina does not extend the full length of the prothorax, being effaced anteriorly. In Patriciella (Fig. 407) and some species of Elatichrosis the lateral edge of the prothorax is acarinate.

The base of the prothorax occasionally has two sublateral longitudinal incisions sometimes associated with carinae. This is particularly useful in determining such genera as Aeoloderma, Drasterius, Rivulicola (Fig. 415), all three cardiophorine genera (Figs 391, 392), Glypheus, Paracrepidomenus, Enischnelater, Litotelater and Neboisselater.

A characteristic attribute of the genera Agrypnus, Trieres and Lanelater (Fig. 3) is the presence of an elongate groove along the pronotosternal suture that opens into an antennal pocket beneath the hypomeron. This accommodates the antenna when at rest. A similar pronotosternal groove and associated antennal pocket is found in Cussolenis (Fig. 35), Drapetes (Fig. 43) and Osslimus.

Figs 3, 4.

Pronotum, ventral view: 3, Lanelater sp. (Carnarvon, WA). 4, Anthracalaus australis (Brisbane, QLD). Scale lines: 1.0 mm.

A number of elaterids have a very short prosternal spine, which is defined as that portion of the prosternal process that is posterior to the procoxae. The spine is shorter than the procoxae in Austrelater (Fig. 21), Rivulicola (Fig. 414), Antoligostethus and Nullarborica. In Drapetes, Cussolenis, Cardiotarsus (Fig. 390), Macromalocera, Microdesmes and Melanotus the prosternal spine is approximately as long as the procoxae. It is sometimes difficult to distinguish between the two conditions.

Depressions are sometimes present on the hypomeron and/or mesosternum, especially in Agrypnus, but this character appears to be of significance only at the specific level.

In most elaterids the anterior margin of the scutellum is well defined, sharply angulate and steeply declivous to the prescutum that lies on a lower plane than the disc of the scutellum. However, in the genera Parablax, Wynarka, Parasaphes, Xuthelater and Tasmanelater the anterior margin of the scutellum is not angulate but rounded and gradually slopes ventrally to the prescutum instead. In Ophidius, Rangsia and Yalganus the scutellum forms a cylindrical column whose apex forms an oval shape, the disc usually being strongly convex as well.

The mesosternal cavity is deep in nearly all Elateridae but in Austrelater and Tyiwiphila the cavity is very shallow and the walls of the cavity are not well defined. In a number of genera, all agrypnine (Anthracalaus, Austrocalais, Lanelater, Trieres, Agrypnus, Paracalais and Macromalocera), the floor of the mesocoxal cavity has a medial, elongate groove or depression that is densely lined with short, golden setae. This character may prove to be of considerable value in assessing synapomorphic relationships in future studies.

Figs 5–7.

Mesothorax ventral view: 5, mesocoxal cavity open to both mesepimeron and mesepisternum. 6, mesocoxal cavity closed. 7, mesocoxal cavity open to mesepimeron only. CX2, mesocoxal cavity. HY, hypomeron. EM2, mesepimeron. ES2, mesepisternum. PR prosternum. S2, mesosternum. S3, metasternum. [after Platia 1994]

The mesocoxal cavity provides numerous useful characters at the generic level. In particular the sclerites that form the margin of the mesocoxal cavity differ. Three main types have been recognised in the Australian fauna and these have been used in the generic key. In a majority species the mesocoxal cavity is open to both the mesepimeron and mesepisternum (Fig. 5), or in other words both the mesepimeron and mesepisternum form part of the margin of the mesocoxal cavity as well as the mesosternum and metasternum. In genera such as Agrypnus, Trieres, Rivulicola and the cardiophorine genera the mesocoxal cavity is laterally closed (Fig. 6), or in other words only the mesosternum and metasternum form the margin of the cavity. In Cussolenis, Drapetes, Osslimus, Anchastus, Diadysis, Melanotus, Hapatesus, Crepidomenus and Glypheus the mesocoxal cavity is only open to the mesepimeron (Fig. 7) or in other words the mesosternum, mesepimeron and metasternum form part of the margin of the mesocoxal cavity.

Hind wing venation

The principal veins of the hind wing are indicated by the names adopted by Kukalová-Peck & Lawrence (1993). These names do not agree with those adopted by Crowson in his works, as well as many subsequent authors on the group. However, perusal of the above mentioned paper will point out the differences and the veins have been labelled on one wing of each subfamily illustrated in this work (Figs 50, 102, 114, 233, 239, 249, 301, 310, 343, 365, 393, 417).

Venational characters useful in the separation of some genera include the presence or absence of CuA1 (referred to as the apparent cross vein between MP4 and CuA2 in the generic descriptions) and the position of CuA1 in relation to the branching of MP4 and MP3. CuA1 is present in the lissomine, thylacosternine, denticolline and pityobiine genera, both present and absent in agrypnine and elaterine genera, and absent in cardiophorine and negastriine genera. The MP4-MP3 juncture is proximal to CuA1 in the lissomine and elaterine genera and it is either proximal or distal to CuA1 in the agrypnine and denticolline genera and distal to CuA1 in the thylacosternine Cussolenis.

The wedge cell (sometimes referred to as the anal cell) is absent in the lissomine Drapetes, all Agrypninae, Cardiophorinae and Negastriinae, and the elaterine Melanoxanthus. The type species of Arachnodima, some species of Agrypnus and Paracardiophorus are either brachypterous or lack hind wings altogether.

The apex of the hind wing membrane is that section distal to the radial cell that contains the apical sclerotisations and lacks veins. The length of apical wing membrane without venation (distal to the end of RA1+2) is given as a proportion of the total wing membrane, that is measured from the humeral plate to the apex of the wing membrane. This proportion ranges from less than one-third the length of the wing membrane in most elaterids to at least one quarter the length in Cardiophorinae and more than one-third in Negastriinae.

Legs

The legs of Elateridae provide a number of useful characters for defining genera. The mesotrochantin can be either visible or concealed. It is concealed in those genera that have the margins of the mesocoxal cavity formed by the mesosternum and metasternum only, and the states are possibly correlated. In nearly all of the elaterid genera studied the mesotrochanter is quite short in length forming at most 0.3 X the length of the femur. However, in Drapetes and Osslimus the trochanter is elongate and forms at least 0.5 X the length of the femur. In Patriciella the trochanter is enormously developed, being quite globular in appearance.

Figs 8, 9.

Apex of hind femur, lateral view: 8, Agrypnus caliginosus (Boisduval) (Wanniassa, ACT). 9, Anthracalaus australis Fleutiaux (Brisbane, QLD). Scale lines: 0.1 mm.

In a number of genera the presence of two apical spurs on the tibia (Fig. 9) is a useful aid to identification. In the new genus Rousia the apices of the fore and mid legs are furnished with a small comb (Fig. 232) that is diagnostic for the genus.

Elaterid legs show substantial structural variation among groups. In some genera the tarsomeres are simple and are distally oblique or truncate (Fig. 11) as in Agrypnus for example. However other genera have complex tarsomeres that are furnished with either spongiose pads that resemble brushes, setose fleshy lobes or lamellae. Crepidomenus and Glypheus exemplify genera whose tarsi are widened distally as well as possessing distinctive spongiose pads (Figs 158, 159, 194, 195). Anchastus (Figs 302, 303) is the only Australian genus with a densely setose fleshy lobe being present only on tarsomere 3. Cussolenis (Fig. 37) and Drapetes (Fig. 44) have tarsomeres with well developed lamellae on the first four tarsomeres. Conoderus (Fig. 10) on the other hand only has tarsomere 4 lamellate, the lamella either well developed and often extending beneath the next distal