The larva of Chiron senegalensis and comments on its relationships with other Scarabaeoidea (Coleoptera: Chironidae)

After a brief presentation of some historical, taxonomical and biological data on the Chironidae, the third instar larva of Chiron senegalensis Hope & Westwood, 1845 is described and illustrated on the basis of larval characteristics. The systematic posi­ tion of the genus Chiron Mac Leay, 1819 within the Scarabaeoidea complex is discussed.


INTRODUCTION
Although it is now clearly established that the Chi ronidae forms a very peculiar and morphologically iso lated group within the Scarabaeoidea (Huchet, 2000), its relationships with other groups remains uncertain or even enigmatic.A recent discovery in Senegal gave us the opportunity to describe the larva of Chiron senegalensis Hope & Westwood, 1845 and, using larval characters, to discuss the affinities of Chiron Mac Leay, 1819 with the main groups of Scarabaeoidea.

HISTORY AND TAXONOMY OF CHIRONIDAE
There have been many controversies in the past about the systematic position of species currently belonging to Chiron (Huchet, 2000).Fabricius (1798) was the first to describe a species, as Scarites cylindrus.Few years later, that species was placed either by Illiga (1801) in the genus Passalus Fabricius, 1792, at that time still in the Lucanidae, or by Fabricius (1801) in Sinodendron Hellwiy, 1792.Latreille (1817: 292), in a paper dealing with Passalus, suggested that a new combination was required for the species described by Fabricius.The genus Chiron was proposed two years later by Mac Leay (1819), who considered this taxon to be close to the Platypodidae.Latreille (1825Latreille ( , 1829)), taking into account the structure of the labrum and mandibles (well sclerotised and visible beyond the clypeus in dorsal view), considered Chiron to be closely related to Aegialia Latreille, 1807.This opinion was accepted by subsequent authors, albeit with some reservations.Until recently, Chiron was considered to belong to the Aphodiidae or Aegialiidae (Balthasar, 1963) and the only genus in either a tribe or subfamily.Paulian (1976) was the first author to treat Chiron as forming a distinct family.
In a recent paper, Huchet (2000) reviewed all these suggestions and showed that the adult morphology of Chiron (mouthparts, mandibles, genitalia...) was suffi ciently distinctive to warrant the family status suggested by Paulian.Two new genera recently proposed by Huchet (2000) have been reviewed (Huchet, 2002a, b).Adults of Chironidae are easily recognised by their cylindrical shape, 9-segmented antennae, labrum produced beyond the clypeus, prominent mandibles and mesothorax distinctlypedunculate(Fig. 1).

BIOLOGY AND DISTRIBUTION
Little is known about the biology of these insects.Adults are nocturnal* and are most often collected at light traps, sometimes in large numbers.In Mauritania (P.Bruneau de Miré, pers.com.) adults of Chiron senega lensis and Theotimius grandis (Gory, 1830) (Col.Chi ronidae) were found together in the vicinity of Nyayes (temporary ponds) under rotting trunks of Acacia nilotica (L.) (Leguminosae), in waterlogged soil.Field observa tions of one of us in Senegal during the rainy season (J.B.H.) indicate that, during the day, adults and larvae of Chiron senegalensis live in subterranean burrows in the vicinity of temporary ponds.The adults appear to be probably gregarious since they were often found in adja cent tunnels near the soil surface (approx.depth: 1-2 cm).Larvae were collected in the same place but living in sub oval cells, approximately at 3-5 cm beneath the soil sur face.Based on an examination of the hind gut contents of adults, R. A. Crowson (pers. com., 1998) concluded that species of Chiron feed on terrestrial filamentous algae, and have habits similar to Bledius Leach, 1819 or Heterocerus Fabricius, 1792 species.
The Chironidae includes 3 genera (Huchet, 2000) and about 25 species mainly distributed in Africa south of the Sahara; with some species in Madagascar, Pakistan, India (as far as Assam), Bangladesh and Sri Lanka.The pres ence of species of Chiron in the Mediterranean Basin (Tunisia, Sicily and Corsica) is very likely the result of accidental introductions.

Description of the larva
Melolonthoid larva, strongly C-shaped but not hump backed (Fig. 16).Length: 7.5-10 mm.Body milky-white, the head ivory.Mandibles brown, the apex distinctly darker.Setation on body very fine and sparse except for the short dorsal stout setae arranged in transverse rows on dorsa of abdominal segments I-V (Fig. 27).
Labrum transverse, symmetrical, coarsely triangular, trilobed anteriorly.The median lobe with 4 setae on the anterior margin; each of the lateral lobes with 5-6 unequal setae (the postero-lateral ones very short), a dozen short setae arranged in a transverse row in antero-median posi tion and, more proximally 2 median macrosensilla dis tinctly separate.
Mandibulae (Figs 7-12) strongly sclerotised, sub pyramidal, asymmetrical, both with a ventral process well developed.Scissorial area of left mandible with 3 scissorial teeth (S1-3), and little separation between the apical (S1) and median tooth (S2); S2-S3 separated by a scisso rial notch, the third posterior tooth well developed, acute.Molar area with 2 lobes, distal lobe slightly developed, the proximal lobe large, sub-quadrate, concave with the antero-external angle distinctly emarginate.Scissorial area of right mandible with 2 scissorial teeth (S1-2) sepa rated by a scissorial notch.Molar area with 2 lobes, distal lobe rather reduced, the proximal lobe larger with base clearly bilobed, the internal basal lobe more developed proximally.Each mandible with 4 long dorso-external setae, the first three setae arranged in a distinct groove, the fourth seta more distal and placed at the base of the S2-S3 groove.Baso-medial angle with a brustia con sisting of 2 short curved setae.No stridulatory organ.
Maxillae (Figs 22,23): Lacinia and galea slightly scle rotised, distinctly separate.Lacinia larger than galea, the internal angle acute, without uncus, the internal margin rounded and bearing nine setae (6 strong beveled setae and 3 long and acute basal setae).On dorsal view, these setae are bordered with an inner row of 5 fine setae; an inner row of 4 fine setae on ventral side.Galea with a single strong terminal uncus; a row of 4 stiff dorso-apical setae and 3 latero-extemal setae in ventral view.Dorsal part of stipe with 2 strong setae; the ventral part with 2 long latero-basal setae and 2 similar setae near the proximal margin of galea.Cardo longer than wide.Maxil lary articulating membrane bearing sparse setae and a definite transverse squamiform area.No stridulatory organ.Maxillary palpi 4-segmented; palpomeres 1-3 sub equal, distal palpomere sub-fusiform with a terminal sen sory area bearing unequal microsensillae arranged in a row.Palpifer scarcely distinct.
Labium with a distinct sub-trapezoidal postmentum.Prementum divided into a distal and a proximal prementum.Labial palpi 2-segmented, the proximal palpomere cylindrical, longer than wide, distal palpomere scarcely shorter, sub-conical with apical sensory area bearing few short sensillae.Distal part of prementum transverse, with 2 long median setae and 2 similar setae located between the palpi, the proximal part sub-trapezoidal with 2 long setae arranged postero-laterally.
Hypopharynx (Figs 3, 21) strongly convex, sub trapezoidal, with large base and anterior part narrower, the lateral lobes rounded.The 2 oncyli small, distinctly fused together in a J-shape hypopharyngeal sclerite and located behind the dorsal surface of labial glossae.Glossae with 4 distinct macrosensilla in the middle, sur rounded on each side by a longitudinal row of 3-5 stiff setae and linked by a transverse row of approximately 8 spine-like setae.Central lobe of glossae bearing 2 para median short setae, the lateral lobes with 2 unequal long setae on each side.Antennae 4-segmented; antennomeres 1-3 well devel oped; the first antennomere cylindrical, very elongated, slightly pubescent, twice length of second, the third antennomere with 4-5 apical setae arranged in circle and with a distinct dorsal sensory cone; distal antennomere markedly shorter than the first three ones (Figs 5,6,19), sub-conical, latero-dorsally inserted on the third segment and appearing as an appendage of this segment, with an apical sensory area bearing several microsensillae.
Prothorax without projecting anterior processes or sclerotised plates; dorsum of prothorax divided into 2 scarcely distinct annulets bearing few sparse setae; dorsa of meso-and metathorax divided into: prescutum, scutum and scutellum; scutellum of metathorax bearing a row of approximately 14 short stout setae.
Legs (Figs 13,14,24,25,26) well developed, quite strong, all four-segmented.Trochanter and femur nearly fused.Prothoracic pair slightly longer than meso-and metathoracic pairs.Prothoracic pair with a distinct tibia and an elongated tarsus bearing a strong claw and 2 strong setae on each side (Fig. 25).Meso-and metatho racic pairs shorter and squatter; tibia and tarsus fused in a tibiotarsus, with the claw inserted on a weak bulge.No stridulatory organs.
Spiracles cribriform (Fig. 15), the respiratory plate reniform bearing approximately 3-9 confluent "cavities" arranged in about 20-30 semicircular rows; concavities on the respiratory plates of thoracic spiracles facing pos teriorly, those of abdominal segments facing ventrally.Respiratory plates of spiracles not constricted and with no large surrounding bulla.
Raster (Figs 29,30) with teges bearing approximately 28 short spike-like setae on each side oriented laterally, without sclerites at base; septula narrow, scarcely distinct; barbula consisting of rather numerous long setae; campus occupying anterior third part of A10; anal opening transverse, slightly curved, surrounded by fleshy lobes: A single upper lobe and 2 tangent lower lobes.
Remarks.Among the material studied there were many living larvae that showed distinct localised lesions due to parasitic fungi, mostly on legs, antennae and palpi.Larvae that had died of fungal infection were not found in the field.

Main differential characteristics of Chiron larvae
The following larval characters strongly indicate the outstanding position of Chiron: the shape of the lacinia, very large and without terminal uncus, does not corre spond to that of any other taxon in the Scarabaeoidea.In Chiron the maxillae, mandibles and legs bear no stridulatory organ or stridulatory area.Several other families also lack a stridulatory apparatus on different anatomical parts.For example, a maxillary stridulatory area is absent in Lucanidae; a mandibular stridulatory area is present in most Scarabaeoid larvae but not in Trogidae; and a stridu latory apparatus on legs is a variable feature of different families of Scarabaeoidea (Browne & Scholtz, 1999).In the present case, the total absence of such organs in Chiron could be interpreted as an important distinctive character.
The 3rd and 4th antennomeres are intimately fused antero-laterally, hence they are not clearly distinct.A similar feature (secondary fusion of the 3rd and 4th antennomeres) is also recorded for some Hybosoridae (Scholtz, 1990).
Last, the morphology of the legs is quite different from that of other Scarabaeoidea larvae: very strong or even squat legs (especially the meso-and metathoracic legs, the anterior ones being more slender), with a terminal uncus quite short and strong, bearing two spine-like setae on each side (in Aphodiidae, Aegialiidae, Aulonocnemidae and also in Trogidae, the legs are always elongated, with fine and prominent claws).The structure of legs may be an adaptation for burrowing (some Melolonthidae and Cetoniidae larvae have similar shaped legs).In Chiron, however, the legs are longer relative to body size.Another remarkable feature is the distinct reduction of the meso-and metathoracic legs.In Chiron, the prothoracic tibiae are elongated and bear a distinct tar sus, each with a strong apical claw.A notable regression occurs in the meso-and metathoracic legs where the tibiae and the tarsi appear to be reduced and intimately fused, forming a tibiotarsus.In fact, the terminal claw is inserted on a weak bulge, which bears 2 strong sensory setae laterally (tarsungulus).Finally, the femur and tro chanter of all legs are distinctly fused and the femur appears to be reduced.Similar features exist in some Geotrupidae (Paulian, 1939).
Other groups of Scarabaeoidea have reduced legs, espe cially the Passalidae (metathoracic legs unsegmented and reduced to a stridulatory organ) and even some Geotrupi dae, where all legs are equally 3-segmented.This does not imply close relationships between these taxa but emphasises some major trends in the reduction of legs in Scarabaeoidea.

Taxonomic relationships
A complete phylogenetic study will be carried out later to confirm or invalidate our conclusions on the relation ships of Chiron within the Scarabaeoidea based on the larval morphology of Chiron senegalensis.
In order to determine how other authors have placed Chiron1 among the other members of the superfamily, we have referred to publications with dichotomic keys (Hayes, 1928;Boving & Craighead, 1931;Van Emden, 1941;Janssens, 1947;Ritcher, 1966;Paulian & Baraud, 1982).Depending on the larval characters used by these authors, the results differ (however, in most cases, not all characters are used) and are therefore inconclusive.
Depending on the key, the taxonomic placement of Chiron appears to be: close to Aphodiinae, according to the larval characters used by Van Emden (1941), Ritcher (1966) and Paulian & Baraud (1982); between Trogidae and Ceratocanthidae (Acanthoceridae sensu auct.), according to the larval characters used by Boving & Craighead, (1931); between Trogidae and Aphodiinae, following Janssens (1947); with affinities with Scarabaeidae, but also intermediate between Lucanidae and Tro gidae (with these latter, in regard with the trilobed anal segment), according to the key ofHayes (1928).
Although Chiron share several features with some of the above taxa, they are not clearly related to any of these groups.The main larval characters it has in common with these groups are most often present in Laparostict2 Scara baeoidea: body cylindrical (not hump-backed3), maxilla with galea and lacinia separated, antenna 4-segmented, tormae fused medially, legs 4-segmented with a well developed claw.According to these characters, Chiron larvae can be placed unequivocally among the less spe cialised Scarabaeoidea.This is in concordance with adult morphology since many characters support this conjec ture: trilobed-type aedeagus (D'Hotman & Scholtz, 1990), 8 pairs of functional spiracles, presence of the 1 The study of larval material of other genera of Chironidae (e. g.Theotimius Huchet or Amphiceratodon Huchet) could confirm or even refute the conclusions here presented.2 The former separation of Scarabaeoidea into Laparosticti and Pleurosticti is controversial since the adults of some of the families (e.g.Glaphyridae, Pachypodidae or Aclopidae) cannot be related to either of them.3 The typical hump-backed body is one of the main characteristics of Scarabaeidae larvae (dung beetles) associated with their habit.A similar larval character occurs in some Geotrupidae.
The fact that some major taxa were initially removed and thus not considered (Scarabaeidae, Geotrupidae...) does not imply a posteriori a relationship with Chironidae since some of the distinctive characters used in the keys (e.g.hump-backed body) are the result of secondary adaptations.
The relationships with Aphodiinae is based on the labrum being trilobed anteriorly in both groups.This character is often present in other taxa (some Pleocomidae, Scarabaeidae -particularly some Coprinae -, Aulonocnemidae or even Lucanidae); as it is not always present in some groups, its taxonomic value is question able.The second feature is the number and disposition of the anal lobes (trilobed).This last character also occurs in Trogidae and Lucanidae but in those groups, the anal slit is generally Y-shaped.In Chiron, the lower anal lobe is divided into two distinct lobes, as recorded in some Aphodiidae (Eupariinae or even Psammodinae) (Ritcher, 1966).If we accept that the shape of the lower anal lobes determines the relationships between Chiron and Apho diidae or Aegialiidae, the latter cannot be retained as a related group (one single lower anal lobe in Aegialiidae).Lastly, the 2-lobed condition appears to be a variable character in the Aphodiidae (e.g. 4 lower anal lobes in Aphodius (Colobopterus) erraticus (Linnaeus, 1758): Jer ath, I960; Ritcher, 1966: 27) and is insufficient support for a close relationship between Aphodiidae and Chironidae.
Aphodiidae and Aegialiidae or even Aulonocnemidae (Paulian & Lumaret, 1974) are distinct homogeneous groups with clear larval affinities, particularly some fea tures of the epipharyngeal structure: remarkably well developed epitorma and strongly sclerotised tormae.Moreover, the shape of the epipharynx is more or less constant in these families.The importance of the epipharynx for distinguishing between the major groups of Scarabaeoidea has not to be stressed today.In the pre sent case, the peculiar triangular shape of the epipharynx, the absence of epitorma and the structure and slight sclerotisation of the tormae in Chiron, is a very distinctive and unique feature in the Scarabaeoidea.A similar weakly developed epitorma and the lack of a stridulatory apparatus on the legs have been pointed out on Trogidae but, according to Scholtz & Lumaret (1991), the larvae of this family have many other distinctive and primitive characters (antennae 3-segmented, ocelli always present, galea 2-segmented or even spiracles with a closing appa ratus), which probably do not support a close affinity with Chiron.While sharing several features with other Scara baeoidea families considered to be phylogenetically primitive (Scholtz, 1990), Chiron has a combination of distinct characters that are peculiar to the taxon.The structure of the epipharynx (triangular shape without epi torma) as well as the shape of the lacinia, the modifica tions to the legs or even antennae strongly indicate the very isolated position of this genus and are sufficiently distinctive to justify a particular status within the Scara baeoidea.
Nevertheless, from the point of view of larval morphol ogy, the systematic position of Chironidae within the superfamily remains still uncertain.This is often the case in small mono-or paucigeneric families (e. g.Aclopidae, Pachypodidae, Belohinidae).A complete cladistic ana lysis based on larval and adults characters will be carried out in order to determine the relationships between Chi ronidae and other Scarabaeoidea; at this time the choice of possible sister groups has not yet been completely made.