4. Elateroidea Introduction, Phylogeny labrum in the larva, and Brachypsectridae were considered to be even more diffi cult to separate on adult features. The superfamily Cantharoidea was John F. Lawrence, Ladislav Bocak, Milada Bocakova, considered to be the most likely group to be merged Rolf G. Beutel and Jyrki Muona with Elateroidea. Lawrence & Newton (1982) followed Crowson The constitution of Elateriformia has varied over in considering Artematopodidae, Brachypsec- time, as discussed in detail by Beutel & Leschen tridae, Elateroidea and Cantharoidea to form a (2005) (see 1–14). The series was fi rst proposed by monophylum, and Lawrence (1988) formally rec- Crowson (1960) for his Dascilliformia (Crowson ognized an expanded Elateroidea to include all of 1955) minus the family Dascillidae (which was these groups. The position of Rhinorhipus Lawrence combined with Scarabaeoidea to form the series at the base of the elateroid clade was considered to Scarabaeiformia) and the families Eucinetidae, be tentative because of lack of information on the Clambidae and Scirtidae, which were placed in a larva, combined with the fact that there are six free superfamily Eucinetoidea. This classifi cation was Malpighian tubules (instead of four as in all other also used in Crowson (1981) except that a series Euci- members of the group). Furthermore, in clado- netiformia was recognized and Rhipiceridae was grams produced by Lawrence et al. (1995), Rhinor- added to Scarabaeiformia-Dascilloidea, based on hipus usually formed a clade with Dascillus Latreille Crowson (1971). In all of Crowson’s classifi cations, (Dascillidae), Sandalus Knoch (Rhipiceridae) and the superfamily Elateroidea included Perothopidae, Dystaxia LeConte (Buprestidae or Schizopodidae) Eucnemidae, Throscidae, Cebrionidae, Elateridae, and was never placed within the elateroid-cantha- and Cerophytidae, although the last was omitted in roid group. error from the 1981 work. In cladograms produced by Beutel (1995) and Lawrence & Newton (1982) followed Crowson’s based on larval characters, Elateroidea (sensu lato) classifi cation in most respects. They did not defi ne was always monophyletic, but this was true of series as such, but considered all Eucinetiformia, neither Elateroidea (sensu stricto) nor Cantharoidea. Scarabaeiformia and Elateriformia as belonging to Most cantharoid families plus Brachypsec- an “Elateriform lineage”. Although Scarabaeoidea tridae formed a clade sister to Cerophytidae + was tentatively included in this “lineage”, some Throscidae + Eucnemidae, while Cantharidae doubt was expressed about the relationship formed a clade with Artematopodidae and Elat- of the group to Dascilloidea. The families nor- eridae. The non-monophyly of the Cantharoidea mally included in Elateroidea and Cantharoidea, was also supported by Bocakova et al. (2007) in plus the Artematopodidae and Brachypsectridae cladograms based on nuclear and mitochondrial were considered to form a monophyletic group. gene sequences. While Elateroidea (sensu lato) was In the fi rst cladistic analysis of Elateriformia, strongly supported in all cladograms, the soft-bod- Lawrence (1988) excluded Scarabaeoidea alto- ied groups usually placed in Cantharoidea never gether, while Eucinetoidea were included as an formed a monophyletic group. The major clusters outgroup in some analyses. The monophyly of were formed by 1) Lampyridae (including Ototreti- Elateroidea + Cantharoidea + Artematopodidae + nae) + Cantharidae, 2) Elateridae (including Drilidae Brachypsectridae was confi rmed in analyses based and usually Omalisidae) + Phengodidae (including on both adult and larval characters, with the family Rhagophthalmidae), 3) Lycidae and 4) Eucnemi- Rhinorhipidae (known from adult characters only) dae. The positions of the genera Drilonius, Telegeusis, at the base of this clade. Trixagus and sometimes Omalisus varied with type Elateroidea was restricted by Crowson (1955) of alignment and analysis: 1) Drilonius, Telegeusis to those taxa the adults of which have more or less and Trixagus formed a clade with Chelonariidae rounded procoxae with concealed trochantins, no and outside Elateroidea; 2) Drilonius and Telegeusis transverse metakatepisternal suture, contiguous formed a clade sister to Elateroidea and Trixagus was metacoxae, hind wing with an apically truncate sister to Elateroidea minus Drilonius and Telegeusis; wedge cell, acutely projecting hind pronotal 3) Drilonius was in Eucnemidae, Telegeusis sister to angles, head without a distinct frontoclypeal Elateroidea minus Eucnemidae, and Trixagus sister suture, trilobate aedeagus with freely articulated to Lycidae; or 4) Drilonius and Telegeusis formed a parameres, and 4 free Malpighian tubules, while clade sister to remaining elateroids, and Trixagus larvae lack a free labrum or epicranial stem and and Omalisus formed a clade within Eucnemidae. have simple, non-channeled mandibles. Artema- Similar results were published by Sagegami-Oba topodidae (then in Dryopoidea) were considered et al. (2007) and Bocak et al. (2008). to be separable from elateroids on little more The Elateroidea, as here delimited, exhibit than exposed trochantins in the adult and a free several major evolutionary trends which deserve 36 John F. Lawrence, Ladislav Bocak, Milada Bocakova, Rolf G. Beutel and Jyrki Muona further mention: 1) development of a type of defen- Leptolycinae (Lycidae) is not a result of the long evo- sive behavior known as “clicking” in adults of the lutionary history, but a consequence of relatively families Cerophytidae, Eucnemidae, Throscidae recent modifi ed function of the endocrine system. and Elateridae, 2) reduction in sclerotization These events potentially led to homoplasious mod- of the cuticle, often accompanied by chemical ifi cations of morphology. The resulting similarity defense mechanisms and aposematic color patterns of soft-bodied or neotenic lineages is therefore dif- in adults of various families formerly included in fi cult to interpret in morphology based analyses. Cantharoidea, and 3) retention of larval features Crowson (1972) postulated that some neoten- (neoteny) in adults of at least some of these families; ous groups, specifi cally the Southeast Asian lycid 4) the evolution of bioluminescence in both adults genera Duliticola and Lyropaeus, are members of and larvae; 5) the occurrence of an elateroid type of primitively neotenous lineages and that fully meta- ecdysis associated with biforous spiracles and the morphosed winged forms re-developed from neo- loss of the spiracular closing apparatus in larvae; tenic ancestors. Similar scenarios of evolutionary and 6) consolidation of the larval maxillae and ‘re-imaginalisation’ were proposed for Lycidae by labium to form a maxillolabial complex. Kazantsev (2005), and equally for the closely related The cuticular and muscular modifi cations which Lampyridae by Cicero (1988). Bocak et al. (2008) make the clicking maneuver possible have been hypothesized that soft-bodiedness represents a discussed by Evans (1972, 1973) for Elateridae, fi rst level of incomplete metamorphosis. Soft-bod- but precursors of these conditions are exhibited ied adults of both sexes are known in Telegeusidae, by members of various families of Dascilloidea, Omethidae, Cantharidae, Lycidae, Lampyridae, Buprestoidea and Byrrhoidea. The evolution of a Phengodidae, Rhagophthalmidae, Drilidae, and pro-mesothoracic interlocking device involving Omalisidae. Some adult females within these projections and concavities or crenulate edges at families are neotenic, i. e., they maintain appar- the posterior end of the prothorax, anterior ends of ently juvenile features resulting in incomplete the elytra, scutellum and/or mesanepisterna, com- metamorphosis and, in extreme cases, the lack of bined with a mesoventral cavity for reception of the adult stages. The neotenic development of females prosternal process, allow these beetles to combine is obligatory in all Omalisidae, Drilidae, Phengodi- mobility with structural integrity, by the unlock- dae and Rhagophthalmidae, and in many lineages ing or locking of this device. The transformation of Lampyridae and Lycidae. The modifi cations of this condition to form the clicking mechanism include females with vestigial wings, but adult- involves the enlargement of the prothorax, increase like thorax (Omalisidae, Lampyridae part), wing- in the mass of the M4 muscle (Larsén 1966), reduc- less females (Lampyridae part) or females with only tion of the size of the exposed portion of the pro- mouthparts and head adult-like (Drilidae, Lampy- coxa, enclosure of the trochantin and (except in ridae part). Lineages affected by neoteny to the Cerophytidae) its fusion to the notum, enlarge- highest degree are found in Lycidae where females ment and deepening of the mesoventral cavity lack both pupal and adult stages and retain a larvae- combined with the formation of a prosternal rest like morphology after the last ecdysis (Wong 1996). and an oblique slide at the anterior end of the cav- Some neotenic lycids reach body sizes of fi ve centi- ity. Based on the topology given by Bocakova et al. meters and more and are frequently referred to as (2007), this condition could have arisen indepen- ‘trilobite larvae’ due to their appearance (Gravely dently from three to fi ve times in the Elateroidea. 1915; Mjöberg 1925). The corresponding males are Vahtera et al. (2009), however, suggested that, given regularly fully metamorphosed and only seldom the complexity of the clicking mechanism, it could brachelytrous (Alyculus