Revision and phylogenetic analysis of the subfamily Platerodinae ( Coleóptera : Lycidae )

Genus-group taxa of Platerodinae are revised and valid taxa are redescribed. The validity of Plateros Bourgeois, 1879 is reinstated. Libnetomimus Kleine, 1927 is made a junior synonym of Libnetis Waterhouse, 1878. Calleros Gorham, 1881, Calloplateros Pic, 1923, Costatoplateros Pic, 1949, Ditoneces Waterhouse, 1879, Libnetomorphus Pic, 1921, Microplateros Pic, 1921, Planeteros Gorham, 1883, Tolianus Pic, 1921, Melampyrus Waterhouse, 1879, and the subgenus Cautirodes Pic, 1921 are consid­ ered to be junior synonyms of Plateros Bourgeois, 1879. The subgenus Pseudeuplectus Pic, 1922 is synonymized to Cavoplateros Pic, 1913, and Pseudoplateros Green, 1951 is made ajunior synonym of Falsocalleros Pic, 1933. Macrolibnetis Pic, 1938 formerly classified with Platerodini is synonymized to Platerodrilus Pic, 1921. Samoaneros Blair, 1928 is considered to be ajunior objective synonym of Melaneros Fairmaire, 1877, which is excluded from Platerodinae and is kept incertae sedis in Lycidae. Fernandum Pic, 1924 and Subdihammatus Kleine, 1926 are transferred to the subfamily Leptolycinae. Teroplas oculatus sp. n. and Microlycus mexicanus sp. n. are described. Neotype of Plateros brasiliensis (Lucas, 1857) and lectotype of Microlycus minutus Pic, 1922 are desig­ nated. In order to understand relationships within the subfamily, included genus-group taxa were cladistically analysed.


INTRODUCTION
With about 900 recognized species, Platerodinae is the second largest subfamily of Lycidae after Metriorrhynchinae.Before this revision Platerodinae comprised 29 valid genus-group names (Bocák & Bocáková, 1990a, 1992;Bocáková, 1997a;Miller, 1997).Most of them, especially those of M. Pic, have never been mentioned since they were proposed.After examination of extensive material, predominantly from south east Asia, it has become clear that many of these taxa are typologically defined groups of species ofvarious relationships.
Most of the confusion in taxonomy of Lycidae was induced by taxa of a famous French author, Maurice Pic, a taxonomist of Coleoptera, Heteroptera, Diptera, Hymenoptera, and some other groups.He mostly pub lished in Mélanges Exotico-Entomologiques and L'Echange where he was usually the single contributor.It is generally known that his descriptions were extremely brief.The problem of M. Pic's taxa is that they are not adequately defined and are compared often with distant taxa.Within Lycidae M. Pic proposed 44 genus-group taxa (Kleine, 1933), 29 of them are monotypic.These 44 taxa together contain only 139 species.
M. Pic's collection is predominantly composed of pri mary types, and in fact, does not contain other identified material or large paratype series.The reason is not extraordinary diversity, but absolute chaos.M. Pic described a single species of Pyrochroidae 17 times (D.K. Young, pers.comm.), and described one series of a lycid species collected by J. B. Corporaal in a single loca tion in Sumatra as 3 distinct species.
Confusion in genus-group taxa is similar to that at the alpha taxonomic level.Only by using prefixes Falso-, Flabello-, Gracilo-, Macro-, Micro-, Pseudo-, and compa-ring genera with representatives of other subfamilies (or sometimes even families) M. Pic proposed 30 mostly monotypic genera in Lycidae (other Pic's monotypic genera bear different names).None of these names was later cited, and most of them are indistinguishable from previously described genera.It is possible to say that around 50% of his genus-group taxa represent nothing more than another species.His genus Flabelloporrostoma based on a supernumerary third antenna, glued on a label under the holotype and which does not belong to the specimen, is a peculiar case.Now, the question is how to solve this situation with many valid genus-group names for few real taxa.R. Kleine, another specialist in Lycidae, ignored all Pic's work, but this resulted in further synonyms.Other authors simply did not use Pic's taxa.However, these approaches cannot persist infinitely.
The cleanest solution would be to perform a phyloge netic analysis of the whole group at alpha-taxonomic level, and to synonymize taxa as a result of this analysis.Unfortunately, we have to work with very restricted data.Some old genera are based on a single female appearing no different from related species of other genera.It often happens that males, which have characters not expressed in females, are almost consistently classified in different genera than females.Moreover, we are often not able to associate males and females of relevant species, and even in the future it could be complicated.The possible ways to associate the sexes are either to capture a couple of specimens in copula, or to use DNA techniques to pro vide an unequivocal identification.Both the possibilities require collection of additional material.The difficulties are: (1) Capturing of mating specimens is a rare occasion.
(2) Most of species on which the taxa are based were col lected in lowlands of tropical regions that are now heavily deforested and, therefore, finding additional specimens is also a rare occasion.For example, when collecting in the Philippines I found only about 5% overlap at species level when compared with Baker's material from 1920's.When studying the very extensive material of the Smith sonian Institution expeditions to Sri Lanka from 1970-1977, Bocak & Bocakova (1990b) found only five described species, but 19 new species.(3) The use of molecular methods is dependent on the quality and quan tity of DNA that can be extracted.These depend on pres ervation methods.A particular problem with pinned specimens from tropical collections is fungal growth (Townson et al., 1999).Also specimens killed with ethyl acetate and subsequently air-dried (which is the most common method to kill beetles) produced low yields of fragmented DNA (Dillon et al., 1996), and therefore the use of these techniques on M. Pic's type material will always be limited.
Regardless of these facts, I coded available characters of type species of genus-group taxa to solve relationships of the platerodine genera.The consequences of coding unknown character states of a sex (depending on which sex was available) are elucidated in the phylogenetic analysis section.
The subfamily Platerodinae was established by Kleine (1928) in his paper on Indian Lycidae.Later on, Kleine (1933) downgraded this subfamily to tribal status.The Platerodini of Green (1951) united Platerodini and Dictyopterini and his subtribe Platerodi included also some genera formerly placed in Dictyopterini (Lopheros LeConte).Nakane (1969) followed Green (1951) and left genera Lopheros and Eropterus Green within Platerodini, which was accepted by Bocak & Bocakova (1990a).Recently Miller (1997) removed Eropterus and Lopheros from Platerodini to Erotini, falsified monophyly of Erotinae and raised Platerodini to subfamilial status again on the basis of larval characters.Unfortunately, the only genera of Platerodinae he examined (besides removed Lopheros and Eropterus) were representatives of the Plateros -Calleros -Ditoneces lineage that are here understood as a single genus Plateros.The other Platerodinae genera were beyond the scope of his paper and therefore, he kept them intact in the subfamily.
Despite several workers reporting immature stages of lycid beetles, our knowledge remains poor and cannot be generally used to infer the phylogeny of Lycidae.The main objective of this paper is to use evidence of adult morphology to conduct a phylogenetic analysis of genus group taxa placed within Platerodinae, in order to under stand relationships within it, and to propose a natural tribal classification.Additionally, the resulting taxonomic and nomenclatural problems are addressed.

MATERIAL AND METHODS
About 4000 specimens of Platerodinae and other subfamilies were examined.Type specimens of type species of all genus group taxa were examined, except those of Microlyropaeus notaticollis Pic and Planeteros ochropterus Gorham (long-term loans), and except types of Sculptocalleros discithorax Pic, Plateros (Mimolycinella) basicornis Pic, and Libnetisia atri-cornis Pic which were not found in the Natural History Museum in Paris and seem to he lost.
Specimens were examined and illustrated using a Zeiss Technival stereoscopic microscope, with the magnification up to 125x, and illustrated using either the drawing tuhe or the ocular microgrid.
All measurements are in millimetres.Somatic morphology measurements were taken using an ocular micrometer.Eye diameter and interocular distances are taken from the span of eyes.When the eye outline was not circular the diameter was measured at the widest point, the interocular distances at the narrowest point.The hody width was measured at humeri in dorsal view.Male and female genitalia were dissected after having heen hoiled in 10% KOH solution, memhranous parts of female genitalia were then stained in chlorazol hlack and put into microvials containing glycerine and pinned helow the respective specimens.
The numerical cladistic analysis was accomplished using Hennig86 computer program, version 1.5 (Farris, 1988), and applying the implicit enumeration option (Lipscomh, 1994).The tree length and consistency (Kluge & Farris, 1969) and retention indices were calculated.Autapomorphies and the synapomorphy of Platerodinae (character 6) were excluded to avoid increasing the indices.Then a strict consensus tree was constructed using the "nelsen" command of Hennig86.To show the character opti mization, Clados version 1.9 (Nixon, 1995) was employed using default settings for all parameters, including homoplasy (hom 0) that is to indicate as homoplasious only those changes that des ignate an independent origin of a state.Tree Gardener version 2.2 (Ramos, 1997) was used for editing the data matrix and quick processing.Support for individual clades was assessed using one of the computer programs from the package of Random Cladistics 4.0.3,Heyjoe version 3.0 (Siddall, 1996)

Taxa
Before this revision commenced, Platerodinae included 29 genus-group taxa.Seven of these taxa are excluded from the present analysis as terminal units for the fol lowing reasons.Two taxa (Fernandum, Subdihammatus) are transferred to suhfamily Leptolycinae hased on the presence of characters defined hy Bocák & Bocáková (1990a).Another two taxa (Melaneros, Macrolibnetis) are excluded from Platerodinae, heing left incertae sedis in Lycidae, and three taxa (Sculptocalleros, Mimolyci nella, Microlyropaeus) could not he examined (see ahove).The genus Microlyropaeus was only defined on the hasis of puhlished information (Kasantsev, 1997a), and consequently many of its characters could not he assessed.Because the inclusion of a taxon with many character states coded as "?" resulted in unresolved polytomy, Microlyropaeus was also excluded from the analy sis, even though it is supposed to belong to Libnetini.
Therefore, together with Plateros represented 5 times by different species and with 3 outgroup taxa, the prelimi nary cladistic analysis involved 30 terminal taxa (Table 1).
A modified data matrix consisting of 13 terminal taxa and including 3 outgroup taxa was then assembled and analysed.The following genus-group taxa were deleted: Pseudoplateros, Calleros, Calloplateros, Costatoplateros, Ditoneces, Cautirodes, Libnetomorphus, Microplateros, Planeteros, Tolianus, Melampyrus, Libnetomimus.Plateros lanceolatus, P. jizushanensis, P. kubani, P. guineensis, P. cordatus were also excluded from the matrix, and therefore the genus Plateros was represented by a single taxon with the coding representing all known species.The reasons for deleting these taxa were as fol lows: (a) Pseudoplateros and Libnetomimus are bound (Fig. 1a) to Falsocalleros and Libnetis respectively (see Table 1.Data Matrix comments on Falsocalleros and Libnetis).(b) Most of these taxa, when they were proposed, were known from a single or few specimens.When more specimens and spe cies were examined, the characters these taxa were based on were often found to be transitional states.The extent of expression used for erecting of genus forms was only one extreme in the continuous variability of the character (see comments on Plateros -comments on synonymization of individual genera).

Characters, polarity and coding
The following section describes the main features of lycid morphology and how they are allocated into the characters and character states used in the cladistic analy sis.The explicit character and character state definitions are also given.
The data matrix is composed of 45 adult characters.The male and female genital morphology provides the largest suite of characters for the study of Platerodinae.Polarity of characters was determined by outgroup com parison method (Watrous & Wheeler, 1981).Because the sister group relationships cannot be identified with cer tainty, three real outgroup taxa (Dictyoptera, Taphes, Lygistopterus) were coded and employed directly in a simultaneous, unconstrained analysis to avoid the need to assign a priori polarity to characters.A missing data code, "?", was used when the character state was unknown (usually when only one sex was examined).Missing char acter states of taxa that do not exhibit the character in question, were represented by a dash (-).Multistate char acters were treated as unordered.A data matrix of 30 ter minal taxa by 45 characters was constructed shown in Table 1.
In the modified data matrix composed of 13 terminal taxa (see above) characters 37-44 had to be switched off.These characters were uniform for remaining taxa in the analysis (taxa coded by 0), or were present in both states in some taxa (coded by a dash "-"), i.e. all taxa were coded by 0 or -.Characters 2 and 37 encode in fact one multistate character and after exclusion of above given taxa their coding for remaining taxa was identical.There fore character 37 had to be excluded from the matrix, oth erwise character 2 was assigned a priori double weight.

Characters and scoring of states
Below are listed all the characters used for the cladistic analysis and the coding of states.
Figs 2-4: 2 -The first of three minimum-length trees accomplished by Hennig86 "ie" command with bootstrap percentiles for all monophyletic groups; 3 -"Jackboot" values for all monophyletic groups of the first of minimum-length trees accomplished by Hen-nig86 "ie" command; 4 -The third of three minimum-length trees accomplished by Hennig86 "ie" command.The tree of the same topology was obtained by successive weighting.

Phylogenetic analysis
The analysis of data matrix given in tab. 1 using com mands "mh*; bb*;" yielded over 3058 most parsimonious cladograms (overflow) of 55 steps length, a CI of 63 and RI of 81.When the strict consensus tree was generated using the nelsen command, the resultant cladogram (Fig. 1a) was of 75 steps length, a CI of 46 and RI of 63.The consensus tree is much longer than the most parsimonious trees because many branches collapsed due to ambiguous support.The matrix encompasses many unknown char acter states coded by "?" (as only one sex of many taxa is known), and therefore the consensus tree contains a large unresolved polytomy of terminal taxa belonging to tribes Libnetini and Platerodini.Only Libnetomimus and Pseu doplateros can be synonymized as a result of the analysis of the whole data matrix (Fig. 1a).
Afterwards, all the synonymized taxa were excluded from the matrix as well as relevant characters (ch 37-44), and the genus Plateros was represented only by one taxon, similarly as remaining taxa in the matrix.The reason for deleting these taxa is explained in previous section.
The reduced matrix included taxa Dictyoptera, Taphes, Lygistopterus, Lyponia, Microlycus, Calolycus Teroplas, Falsocalleros, Plateros, Cavoplateros, Dihammatus, Libnetis, Libnetisia, and characters 0-36 of the matrix given in tab. 1.The analysis of reduced data matrix using implicit enumeration option yielded 3 most parsimonious cladograms of 37 steps length, a CI of 62 and RI of 68 (trees 1 and 3 -Figs 3, 4; 2nd tree was of the same topology as in fig.4, except the positions of outgroup taxa Lygistopterus and Taphes which are reversed).The strict consensus tree using the nelsen command yielded in two steps longer cladogram of CI-58, RI-64 (Fig. 1b), where the position of genera Calolycus and Microlycus is unresolved.If the successive weighting was applied, the same procedure yielded in single cladogram of the same topology as in Fig. 4, where genera Calolycus and Micro lycus form a monophyletic group.Platerodinae are supported by 3 synapomorphies (ch 6, 11, 15) two of them (ch 11, 15) have each a reversal.Nevertheless, the coding of character 15 for Microlycus is not certain because its coxites are only slightly coalescent.Thus, it would be possible to treat character 15 as multistate, coded by state 2 for Microlycus.Then also this character would have no reversal in Platerodinae.
The interrelationships of Platerodinae genera have remained out of interest of previous authors, and the ini tial placement of most genera (Kleine, 1928(Kleine, , 1933) ) has never been explicitly justified.Bocak and Bocakova (1990a) examined only the tribal classification, and they proposed the subfamily as a tribe Platerodini and estab lished here three subtribes: Lyponiina, Libnetina and Platerodina.But examination of type specimens of all genus-group taxa was beyond the scope of that paper.
Bocak & Bocakova (1990a) proposed sister-group relationships of the genus Lyponia and Platerodini in the restricted sense, and considered Libnetini to be a sister group of Lyponia -Platerodini lineage.According to the present phylogenetic hypothesis (Fig. 1b) Lyponia has a more basal placement and is the sister-group of Libnetini and Platerodini.There are several apomorphies sup porting Lyponiini as a separate clade: shape of the phallo-base and phallus, prolonged basal part of coxites and an attachment of valvifers.The clade Platerodini -Libnetini -Microlycus -Calolycus is defined by 3 synapomorphies, especially the shape of mesosternum, different shape of reticulate cells, as well as by the prolonged trochanters, which seems to be a homoplasy shared with Taphes.
Microlycus and Calolycus have a basal placement in the strict consensus tree, placed after Lyponia but before Libnetini and Platerodini.The monophyly of the clade Microlycus and Calolycus is uncertain.They form two successive branches in the first most parsimonious tree.In the 2nd and 3rd of the three most parsimonious trees (Fig. 4), as well as in the single tree obtained after successive  Dihammatus, formerly placed in the Platerodini lineage (Bocak & Bocakova, 1990a) is considered the basal member of Libnetini.The evidence supporting the closer relationship of Dihammatus with Libnetini, instead of with restricted Platerodini, is based on female genitalia characters (presence of paraprocts -ch 14, inner basal margin of coxites emarginate and distant -ch 17).The close relationship between Dihammatus and other Libnetini, proposed here, is not so evident from the adult external morphology (presence of reticulate cells -ch 3, and secondary costae on elytra -ch 0 -absent in Libnetis and Libnetisia both being plesiomorphies).No data on larvae and pupae of these genera are available, but they may be valuable for better resolution of the phylogenetic relationships.
The monophyly of Platerodini is supported by one synapomorphy -absent spiculum gastrale (ch 19).This character has evolved several times within Lycidae, and may be a result of parallel evolution also in Platerodini.It is the only character supporting placement of the genus Teroplas at the base of Platerodini and therefore, the posi tion of Teroplas, having long paramerae and many autapomorphies in the structure of phallus, within Platerodini is rather tentative.
The sister-group relationship of Falsocalleros and the Plateros -Cavoplateros lineage is clearly supported by a male genital character (stick-formed phallus provided with a dorsobasal pointed projection -ch 12).The close relationship of the speciose Plateros and the American Cavoplateros is well supported by entirely absent paramerae and the shape of the phallobase.

Tests for reliability of the clades of the cladogram
Various tests have been published for assessing confi dence in the clades of a cladogram.In testing the reduced data matrix composed of 13 terminal taxa (when the synonymized taxa were excluded) I have used two randomi zation procedures -bootstrapping and jackknifing.Because these methods are affected by the inclusion of uninformative characters (Carpenter, 1996), these were excluded from the matrix (ch 6 and 20 -35).The new modified matrix was analysed using Heyjoe version 3.0in M. Siddall's (1996) Random Cladistics package version 4.0.3.For bootstrapping the keyword "weight" was used to re-weight characters randomly, while the sum of weights must equal the original number of characters.For thejackknife M. Siddall's "jackboot" (Siddall, 1996) was applied using the "rcd" command (random character dele tion, where deleted characters are not replaced by others).1000 replicates were generated in both the analyses -the bootstrap and thejackknife, the most parsimonious clado gram for each replicate was then found, and the degree of conflict among them assessed by means of a majority rule consensus tree.The ten monophyletic groups present in the tree are given bootstrap (Fig. 2) and jackboot (Fig. 3) percentiles on the tree.The clade Microlycus -Calolycus present only in the second and the third most parsimo nious trees (which only differed in the position of out group taxa Lygistopterus and Taphes) had low bootstrap values (bootstrap -50%,jackboot -42%).
Diagnosis.Head transverse, almost prognathous, occipital foramen of characteristic shape .Tentorial arms long.Mandibles stout, maxillary palpi pro vided with large securiform, rather oblique terminal palpomere.Terminal segment of labial palpi rather dilated apically, labrum strongly transverse.Male antennae pecti nate to filiform, female antennae more or less serrate to filiform.Pronotum subquadrate, with a median impressed line on the disc, scutellum almost square.Elytra slightly widened posteriorly, each elytron with 9 longitudinal costae, reticulate cells strongly transverse in subgenus Ponyalis, subquadtrate in Lyponia s. str.and Weiyangia Bocak.Male genitalia without paramerae, distal portion of phallus often provided with a pair of lateral spines, phallobase annuliform.Female genitalia elongate, valvifers laterally bent, tightly attached to coxites, styli small.Female terminal abdominal sternum provided with short spiculum gastrale.
Distribution.Northern part of the Oriental Region (northern India, Thailand, Laos) and eastern part of the Palaearctic Region (Japan, China, Taiwan).
Comments.Bocak (1999) has recently given a more detailed description of Lyponia in the revision and phylo genetic analysis of the genus.
Comments.Kleine (1927) established the genus Libnetomimus on the basis of the presence of a lamella on antennomere 7.Because of this character appears several times in different species groups of Libnetis (lamella can be present on different antennomeres) it cannot be sup posed to be a synapomorphy, but a multiple parallelism.
Comments.Unfortunately, it was not possible to examine the type species of the genus Libnetisia atricornis because its type specimen was not found in M. Pic's collection (MNHN).The only Libnetisia type specimen found was that of Libnetisia atronotata.There fore, the concept of Libnetisia as here understood corre sponds to it.
Diagnosis.Lycids resembling Dihammatus in having short antennomere 3 that is of the same length as 2. On the other hand, Kasantsev (1997a) examined the structure of male genitalia that are similar to those of Libnetis. Distribution.Sumatra.
Comments.Unfortunately, it was not possible to examine the only known specimen of the genus because its lectotype has been loaned for several years.Therefore, Microlyropaeus Pic could not been included into the phy logenetic analysis.Kasantsev (1997a) examined its lectotype and supposed Microlyropaeus to have rather inter mediate position between the genera Dihammatus and Libnetis.Consequently, here it is tentatively classified within Libnetini incertae sedis.
Composition.This, formerly monotypic genus, cur rently comprises two species.
Redescription.Body dark brown to black, lateral belts of pronotum and humeral portion of primary costa 3 yel low.
Distribution.Central America.
Etymology.The name "oculatus" refers to large eyes of the male.
Composition.This formerly monotypic genus cur rently comprises two species.
Comments.Genera Falsocalleros and Pseudoplateros are the only Platerodini taxa having primary elytral costae elevated (ch 1).Also other characters are in such congru- ence that even species identity is suspected, however, this cannot be solved until both sexes from the two species are known.Therefore, only the synonymization of Pseudo plateros to Falsocalleros is made here.Pic, 1933 (Figs 5, 26, 80, 135) Falsocalleros particularis Pic, 1933: 110.Diagnosis.Elytra with strongly reduced secondary costae, transverse costae irregular, primary costae 1 and 3 apically reduced.

Comments on synonymizations of individual genera.
Plateros Bourgeois, 1879 is a large, widely distributed genus.After the examination of considerable adult mate rial from all zoogeographic regions, and after several regionally limited species revisions (Bocáková, 1997a(Bocáková, , 1997b) ) of the genus Plateros (as Melaneros), it was apparent that maintenance of several genera as distinct from Plateros is arbitrary.The fusion of Plateros, Calle ros, Ditoneces into one genus is also maintained by larval characters (Miller, 1997) -the tripartite larval dorsum seems to be a synapomorphy of the genus Plateros as here understood.Examination of type specimens of all genus-group taxa of Platerodini resulted in many nomenclatural acts.
Calleros Gorham, 1881 was proposed for "small Lycidae allied to Plateros, but differ considerably both in general appearance and, especially, in the long thin anten nae".Holotype of the type species -Calleros puniceus Gorham, 1881 has antennae reaching slightly over elytral midlength, which is usual in other Plateros.The other Gorham's distinguishing character -brightly coloured red thoraces and elytra, also occurs in Plateros from other zoogeographic regions.Calleros puniceus has a rather shorter antennomere 3, but the length is within the range of variation of the other Plateros.Because I found no other distinguishing character, Calleros Gorham, 1881 is synonymized.
Calloplateros Pic, 1923 was proposed for the single species Calloplateros particularis Pic, 1923 which has a slender median longitudinal carina on pronotum.Pic (1923) noticed that Calloplateros is very close to Plateros.If we accept the concept of Calloplateros, the genus Plateros would become paraphyletic having no apomorphic character distinguishing it from Callo plateros, and therefore Calloplateros is synonymized.
Costatoplateros Pic, 1949 was proposed as a subgenus of Plateros Bourgeois, 1879 for the single species Plateros fortecostatus Pic, 1949 on the basis of primary elytral costae more elevated than the secondary ones.This character was observed also in other Plateros like species formerly classified with Melampyrus and Ditoneces.Con sistent classification of all species with moderately stronger primary elytral costae in Costatoplateros would result in a polyphyletic assemblage.If we restrict the sub genus Costatoplateros to be monophyletic to the type species, the nominotypical subgenus Plateros would become paraphyletic having no apomorphic character dis tinguishing it from Costatoplateros, and therefore Costa toplateros is synonymized.
Ditoneces Waterhouse, 1879 was proposed for Plateros-like species having flabellate antennae in males and serrate antennae in females.This character occurs in many other Plateros species.Kleine (1926) already dis cussed the position of Ditoneces.He stated that its relat edness to Plateros is very close and females cannot be placed to this or that genus.He was not aware of any other distinguishing character except the pectinate antennae in males of Ditoneces while those of Plateros are filiform.Kleine (1926) even states: "Während es bei manchen Lycidengattungen moglich ist, durch Verglei chende Untersuchung des mannlichen Begattungsorganes die Gattungszugehorigkeit festzulegen, ist das sowohl bei Plateros wie bei Ditoneces nicht moglich, da in beiden Gattungen der Penis von grosser Vielgestaltigkeit ist." (i.e.While in many lycid genera it is possible to assign generic membership by comparison of male copulatory organs, this is possible neither in Plateros nor in Ditone ces, as the aedeagus is strongly polymorphic in both the genera).In spite of this fact Kleine preserved the genus Ditoneces.Unfortunately, the length of lamellae varies considerably, and closely related species often have both filiform and serrate to flabellate male antennae (e.g.Plateros cordatus group, Bocakova, 1997a).If all Plateros-like species with flabellate antennae were asso ciated in Ditoneces (Kleine, 1933) then the genus would be polyphyletic.If only the type species Ditoneces punctipennis Waterhouse, 1878 was included in Ditoneces, then the genus Plateros would become paraphyletic having no apomorphic character.Therefore, Ditoneces Waterhouse is synonymized.
Cautirodes Pic, 1921 was proposed as a subgenus of Ditoneces for the single species Ditoneces (Cautirodes) malaccanus Pic, 1921, which has slender male antennae with long lamellae.The length of lamellae is within the range of variation of other species formerly placed in Ditoneces, and no additional distinguishing characters have been observed.Owing to this and together with rea sons for the synonymization of Ditoneces given above, Cautirodes Pic, 1921 is synonymized.
Libnetomorphus Pic, 1921 was proposed as a genus close to Planeteros Gorham for 5 Plateros-like species, based on more or less elongate antennomere 3.Because antennomere 3 of the type-species Libnetomorphus cephalotes Pic, 1921 is 1/5 shorter than 4, it is in the range of length of antennomere 3 of Plateros.As no other dis tinguishing characters have been recognized, Libnetomor phus is synonymized.
Microplateros Pic, 1921 was proposed as a genus close to Planeteros Gorham for 7 Plateros-like species with antennomere 3 more or less subtrianguloid.The typespecies Microplateros diversithorax Pic, 1921 as well as other included species have serrate male antennae, and therefore the shape of antennomere 3 is trianguloid.Unfortunately, the shape of antennomere 3 considerably varies in Plateros, and closely related species often pos sess antennomere 3 of different shape (Bocakova, 1997a, b).If all Plateros-like species with trianguloid antenno mere 3 were associated in Microplateros then the genus would be polyphyletic.If only species closely related to the type species Microplateros diversithorax Pic, 1921 were included, then the genus Plateros would become paraphyletic having no apomorphic character, and there fore Microplateros is synonymized.
Melampyrus Waterhouse, 1879 was proposed for two species, both known only from females.Waterhouse (1879) stated that they differ from Plateros chiefly in being more pubescent and in having the antennae broader (serrate), and from Ditoneces in having alternate costae (i.e.primary costae) of the elytra more elevated.All the characters vary considerably in many Plateros species mostly classified with Ditoneces by previous authors.Females of these species have also serrate antennae and the extent of pubescence can hardly be explicitly defined.More elevated primary elytral costae can also be found in some southeast Asiatic Plateros species as well as in South American Costatoplateros.Closely related species often have both slightly elevated and non-elevated elytral costae.If Melampyrus associated all Plateros-like species with slightly elevated alternate elytral costae then the taxon would be polyphyletic.If only species closely related to the type species Melampyrus alternans Water house, 1878 were included then the genus Plateros would become paraphyletic having no apomorphic character.Therefore, Melampyrus Waterhouse, 1879 is synonymized.
Planeteros Gorham, 1883 was proposed for a single African species.Gorham (1883) stated that it was closely allied to the genus Plateros, and that "the key to the sepa ration of these difficult genera of small Lycidae will be found in the proportion of antennal joints".Gorham had only a few specimens at his disposal, and therefore it was easy for him to separate these genera.Examination of much more material indicates that this is an artifact and antennal length of Planeteros is within the range of varia tion in Plateros.Although Kasantsev (1997b) down graded Planeteros to subgeneric rank, it is here synonymized.
Tolianus Pic, 1921 was proposed for a single species as close to Melampyrus.It is based on a single female having 9 almost equal elytral costae and serrate antennae (antennae of the holotype are partly damaged and only antennomeres 1-4 are present).This specimen corre sponds with females of those males having flabellate antennae that were hitherto placed in Ditoneces.There fore similarly as Ditoneces, also Tolianus Pic, 1921 is synonymized.
As a result of the synonymization of these taxa, many secondary homonyms are created.However, assigning replacement names without revision would only add to a proliferation of names for taxa of unknown validity.
Comments on neotype designation.The neotype is designated with the express purpose of clarifying the taxonomic status of P. brasiliensis, the type species of Plateros.The original description did not include either description of male genitalia or their drawings which are essential for any decision on species identity in Plateros.
Lucas' collection is deposited in the Natural History Museum in Paris.My search for the holotype was not successful there, and therefore the type specimen is con sidered to be lost (Dr.Menier personal communication).
The neotype is the property of the National Museum in Prague and is consistent in all characters with the original description.Its redescription is given above.The neotype comes from Brazil, which is the original type locality of Plateros brasiliensis.
Comments.Type species of Mimolycinella and Sculp tocalleros should be deposited in M. Pic's collection (MNHN).Unfortunately, searching for them within MNHN has not been successful, although it is possible they can be found there in the future.As the original descriptions are entirely insufficient, these taxa are kept within Platerodini incertae sedis.

Distribution. Mexico.
Comments.Performance of the cladistic analysis of Platerodinae did not help too much to elucidate the rela tionships of Microlycus, although certain relatedness to Calolycus was indicated.In the resultant cladogram (Fig. 1b) both the genera had basal placement after Lyponia.One of the two most parsimonious trees obtained by com puter analysis (Fig. 4) included also the branch Micro lycus -Calolycus, and the same tree topology was obtained using successive weighting.Regardless, the group Microlycus -Calolycus is not adopted in the classi fication because male of Microlycus is not known, and the clade Microlycus -Calolycus was based on absence of paramerae.Moreover, these genera differ substantially in the morphology of female genitalia, the clade Micro lycus -Calolycus has split in the strict consensus tree, and has low bootstrap andjackboot values.Pic, 1922 (Figs 22, 35, 92, 128) Microlycus minutus Pic, 1922a: 22. Redescription. 2. Body dark brown, only pronotum (except basal median spot) yellow.Head small, partly concealed by pronotum, eyes small, interocular distance twice as long as eye span.Antennae filiform, reaching elytral midlength.Antennomere 1 stout, 2 very small, 3 almost twice longer than 2. Antennomere 4 is the longest of all, twice longer than 3, but 1.5 times longer than 5. Labrum transverse, mandibles short but stout, distally arcuate, their apices do not touch each other, maxillary palpi long with terminal palpomere weakly narrowed api cally, labial palpi short.Pronotum oblong, strongly trans verse, widest at basal margin, anterior angles rounded, posterior ones projected obliquely backwards.Pronotum flat, without areolae, sides slightly elevated.Scutellum apically emarginate.Elytra elongate, semicircularly wid ened in posterior half, about 2.8 times as long as humeral width.Each elytron with 9 costae, reticulate cells well expressed, primary costae 2 and 4 strongly elevated.Mesosternum trapezoidal, wing venation with AP3+4 absent.Spiculum gastrale short, shorter than half of female terminal sternum.Legs rather slender, compres sed, tibiae straight, their spurs small, negligible, tro chanters short.Female genitalia with long basally coalescent valvifers, coxites basally fused, styli half shorter than coxites.Body length 5.4 mm, humeral width: 1.7 mm.Male unknown.
Comment.The lectotype designation is made with the purpose of clarifying the application of the name Microlycus minutus Pic to the species figured in figs 22,35,92,128, and to distinguish it from the other syntype that rep resents the following species.
Comments.The reasons for leaving Calolycus incertae sedis within Platerodinae are the same as for Microlycus (see comments on Microlycus).
Comments.On the basis of the shape of male genitalia (Figs 117,118) and reduced mouthparts the genus is transferred to the subfamily Leptolycinae (see Bocak & Bocakova, 1990).Within this subfamily, Subdihammatus seems to be related to the recently described Skrivania Bocak et Bocakova, 1999 from Malaysia.Unfortunately, Kleine destroyed the basal portion of holotype aedeagus when he dissected it, and therefore the comparison of phallobases is impossible.
Comments.Blair (1928) overlooked fixation of the type species Melaneros acuticollis for the genus Melan eros made by Bourgeois (1891), and proposed a new genus Samoaneros for the same species.Therefore, Samoaneros Blair, 1928 is considered to be a younger objective synonym o f Melaneros Fairmaire, 1877.
Holotype of Melaneros acuticollis Fairmaire was deposited in the Zoological Museum in Hamburg, and was destroyed during the World War II.I have examined specimens identified by Blair deposited in BMNH that seem to correspond to Blair's (1928) redescription and drawings.Melaneros possesses many autapomorphies, but does not show the synapomorphies of Platerodinae.Examination of its female genitalia (Fig. 140) suggests some relatedness either to Metriorrhynchinae Kleine, 1926 or to tribe Calopterini of the subfamily Lycinae (see Bocak & Bocakova, 1990).Because its relationships are not sufficiently supported I leave the genus incertae sedis within the family Lycidae.
Comments.Although Duliticola Mjöberg, 1925 has always been classified with Lycidae, its close relativethe genus Platerodrilus was long considered to be a member of Drilidae.Crowson (1972) transferred Platerodrilus to Lycidae.Similarly Bocák & Bocáková (1990) regarded Platerodrilus to belong to Lycidae although they kept it incertae sedis within the family.The relationships of Platerodrilus and Duliticola need further study (Wong, 1996) which should result in the final deci sion whether to treat these close taxa as separate genera.Because the holotype of Macrolibnetis depressus bears all characters of the genus Platerodrilus as shape of head, long and stout mandibles, shape of pronotum as well as elytral structure, the genus Macrolibnetis Pic, 1938 is considered to be a junior synonym of Platerodrilus Pic, 1921.
overlooked this fixation, designated Melaneros atroviolaceus Fairmaire, 1877 as type species of Melaneros, and proposed name Samoaneros for the species Melaneros acuticollis.Bocak & Bocakova (1992) also overlooked Bourgeois's fixation of Melaneros acuticollis as type species of Melaneros, considered Blair's (1928) fixation of the type species Melaneros atroviolaceus for the genus Melaneros to be valid, and consequently they synonymized Plateros Bour geois, 1879 to Melaneros Fairmaire, 1877.The detection of the fact that Bourgeois (1891) had already fixed Melaneros acuticollis as type species of Melaneros com pels me to reinstate the validity of Plateros Bourgeois, 1879.