2 AMERICAN MUSEUM NOVITATES NO. 3328 adults rest by burying themselves in sand. SYSTEMATICS The early extant lineages of many groups of LEPTOCONOPS Skuse organisms are highly modified and this is also true of species of Leptoconops. Leptoconops Skuse, 1889: 288. Type-spe- Borkent (2000) recently published a de- cies, Leptoconops stygius Skuse, by mono- tailed description and analysis of 22 species typy. of Ceratopogonidae known from ancient, DIAGNOSIS AND DESCRIPTION OF ADULTS: 120–122 million-year-old Lebanese amber. As given by Borkent (1995), Clastrier and Although the fossil genus Lebanoculicoides Wirth (1978), Smee (1966), Szadziewski Szadziewski, which represents an even ear- (1988: 232), and Wirth and Atchley (1973). lier lineage of Ceratopogonidae than does In addition: females with 10–13 flagellom- Leptoconops (fig. 4), was present in the am- eres. ber, no specimens of Leptoconops were dis- Borkent(1995,2000)andWirthandAtch- covered among the 126 ceratopogonid spec- ley (1973) provided an overview of the dis- imens available for study. But as happens so tribution, biology and fossilrecordofthege- often to systematists, at the very final stages nus.BorkentandWirth(1997)cataloged127 of submission for publication, a few more extant and 6 extinct species of Leptoconops; specimens came to light and, as predictedby since then a further 7 extant and 3 extinct phylogenetic patterns, these were identified species (including those newly named here) as members of Leptoconops and of excep- have been recognized. tional interest. This paper describes two new Eight fossil species of Leptoconops are species of Leptoconops and discusses both now known from the Cretaceous and they the systematic and paleoecological implica- may be recognized using the followingkeys, tions of these discoveries. which are based in part on earlier keys by The oldest definite fossils of this genus Borkent (1996) and Szadziewski (1996). A werepreviouslyknownfromFrenchandTai- tentativelyidentifiedandunnamedmaleLep- myr amber at 93–99 ma (Szadziewski,1996; toconops known as a compression fossil Schlu¨ter, 1978). A poorly preserved com- from the Lower Cretaceous in Australia is pression fossil from Australia, dated at 115 not keyed. Neither is the unnamed female (cid:54) 6–118 (cid:54) 5 ma, was tentatively identified Leptoconops from Upper Cretaceous French as a male Leptoconops by Borkent (1997) amber briefly described by Schlu¨ter (1978). but its identity is somewhat uncertain. KEY TO CRETACEOUS MALE MATERIALS AND METHODS LEPTOCONOPS This study is based on 14 newly discov- The males of L. antiquus from Lebanese ered specimens of Ceratopogonidae in 4 amber and L. boreus Kalugina from Taimyr pieces of Lebanese amber. All pieces are in amber are unknown. small plastic boxes and are housed in the 1. Gonostylus with medial lobe at midlength and American Museum of Natural History. The subapical,pointedprojectionaboutone-third specimen numbers follow their cataloging from apex .......................... system. ..... sibiricus Szadziewski (Taimyr amber) Methods follow those described by Bor- – Gonostylus a single, undivided structure .... kent (1995, 2000). However, all specimens .................................. 2 were at least partially embedded in epoxy 2. Wingwithcostawell-developedbeyondapexof and polished using the method described by R; first tarsomere of hindleg with many 3 Nascimbene and Silverstein (2000). Inclu- (about 10 or more) thick spines (fig. 1B) ..... amplificatus, n. sp. (Lebanese amber) sions were examined with a Wild M3 dis- secting scope at 100(cid:51) and with a Carl Zeiss – Wing with costa terminating at apex of R3; first tarsomere ofhindleg withnomorethan Jenaval compound microscope by suspend- 6 thick spines ...................... 3 ing the amber in immersion oil on the un- 3. Apicolateral process on tergite 9 thick and derside of a coverslip as described by Bor- slightly enlarged apically ............. kent (2000). ........ clava Borkent (Hungarian amber) 2001 BORKENT: LEPTOCONOPS 3 – Apicolateral process on tergite 9 slender and form a monophyletic group and the new fos- elongate, tapering to apex ............ 4 sil taxa must therefore be placed in a new 4. Terminal flagellomere about 2–3 timesaslong subgenus. There is no synapomorphy group- as penultimate flagellomere ........... ing L. amplificatus and L. antiquus and it is .... copiosus Borkent (New Jersey amber) therefore uncertain whether L. (Palaeocon- – Terminal flagellomere about5–6timesaslong ops) is monophyletic. as penultimate flagellomere ........... 5 5. Gonocoxite relatively short and stout ..... DERIVATION OF SUBGENERIC EPITHET: The ... curvachelus Borkent (New Jersey amber) name Palaeoconops is from the Greek, pa- – Gonocoxite relatively elongate and slender .. laeo ((cid:53) ancient, old) and konops ((cid:53) gnat), ..... primaevus Borkent (Canadian amber) referring to the incredible age of these fos- sils. KEY TO CRETACEOUS FEMALE Leptoconops (Palaeoconops) amplificatus LEPTOCONOPS Borkent, new species ThefemaleofL.clavaBorkentfromHun- garian amber is unknown. Figures1A–J,2A 1. Antenna with 13 flagellomeres (figs. 1H, 2C) DIAGNOSIS: Male. The only Cretaceous ................................. 2 Leptoconopswith awell-developedcostabe- – Antenna with 12 flagellomeres .......... 3 yond R , a well-developed R , and strong 3 4(cid:49)5 2. First tarsomere of hindleg with more than 15 spines along the length of the first tarsomere thickspines(fig.1I);cercuslongandslender of the hindleg. Female. The only species of (fig. 2A) ........................... Ceratopogonidae with 13 flagellomeres and ..... amplificatus, n. sp. (Lebanese amber) an elongate, slender cercus. – First tarsomere of hindleg with a few (about DESCRIPTION: Male. Head: Most details 4–5)thickspines;cercusrelativelyshortand not visible. Antenna (fig. 1A) with well-de- well-developed anteroventrally (fig. 2G) ... veloped plume, basal foramen of pedicel not ........ antiquus, n. sp. (Lebanese amber) visible, 13 separate flagellomeres, antennal 3. Tarsal claws strongly curved basally, with thick, well-developed inner tooth ...... ratio (cid:53) 0.69, flagellomere 10/11 (cid:53) 0.77, fla- ... curvachelus Borkent (New Jersey am- gellomere 13 more elongate than preceding ber), sibiricus Szadziewski (Taimyr amber) flagellomeres(fig.1B,G).Palpuswith4seg- – Tarsal claws more or less evenly curved from ments, details not visible. Thorax: Most de- base, with at most a very slender, hairlike tails not visible. Wing: Length (cid:53) 0.64 mm. inner tooth ........................ 4 R well-developedtonearwingapex.Cos- 4. Flagellomeres 3–11 somewhat elongate ... 4(cid:49)5 ta well-developed to near wing apex, exact ..... primaevus Borkent (Canadian amber) costal ratio uncertain. Without macrotrichia, – Flagellomeres 3–11 spherical .. copiosus Bor- fine microtrichia on all membrane. Legs: kent (New Jersey amber), boreus Kalugina Femora, tibiae slender. Hindleg first tarso- (Taimyr amber) merewithnumerousstoutsetae(fig.1B).Se- tae on fore- and midleg trochanter not visi- Palaeoconops, new subgenus ble. Midleg tibia with apical spur. Hindleg DIAGNOSIS: Male and female. Only Lep- first tarsomere without thick basal spine or toconops with a well-developed costa ex- palisade setae. Claws simple. Genitalia (fig. tending to the wing apex. Also, female. Only 1C, D): Tergite 9 apparently tapering to sin- Leptoconops with an antenna with 13 flagel- gle apex. Gonocoxite moderately elongate. lomeres. Gonostylus thick basally, tapering to toothed TYPE SPECIES: Leptoconops amplificatus, apex; apical spine likely present but not n. sp. clearly visible. Paired structures (para- TAXONOMIC DISCUSSION: Cladistic analysis meres?) basally thick, tapering apically, di- of the two included species, L. amplificatus rected posterolaterally. Aedeagus not visible. and L. antiquus, of the new subgenus Pa- Female. Head: Eyes bare, broadly sepa- laeoconops providedbelow,showsthatthese rated dorsomedially. Vertex without single represent the earliest lineage within Lepto- dorsomedial seta. Frons, vertex areapossibly conops (fig. 4). All remaining Leptoconops with sutures (fig. 1F). Antenna with 13 sep- 4 AMERICAN MUSEUM NOVITATES NO. 3328 Fig. 1. Structures of Leptoconops amplificatus. A–D. Male (no. 79). E–J. Female. A. Antenna. B. Hindleg tarsomeres. C. Genitalia, oblique, dorsal view. D. Genitalia, oblique, lateral view. E. Wing Antenna (no. 101, holotype). F. Head capsule, anterior view (no. 101, paratype # 4). G. Apex of mouthparts (no. 101, holotype). H. Antenna (no. 101, holotype). I. First and second tarsomeres of hindleg (no. 101, holotype). J. Fifth tarsomere of midleg (no. 101, holotype). 2001 BORKENT: LEPTOCONOPS 5 arate flagellomeres, antennal ratio (cid:53) 0.75– likely due to the presence of either a verte- 0.87 (n (cid:53) 5), terminal flagellomere more brate blood meal or nectar. elongate than preceding flagellomere (fig. The holotype of L. amplificatus was as- 1H), penultimate flagellomere longer than sociated with eight paratypes of that species, preceding flagellomeres, first flagellomere in addition to a female Ceratopogonidae of sensilla not visible. Mouthparts moderately uncertain identity, one Mycetophilidae,three elongate (fig. 1F, G). Clypeus separated lat- Brachycera, one Chironomidae and one Cul- erally by membrane from head capsule. La- icomorpha. The allotype and one paratypeof brum with 4–6 short thick apical spines. L. amplificatus were in a single piece of am- Mandible with fine teeth. Lacinia with at ber with the holotype of L. antiquus and the least 15 large, retrorse teeth. Palpus with 4 holotype of a possible Lonchopteridae (Lon- segments, third segment somewhat ovoid, chopterites prisca Grimaldi and Cumming). capitate sensilla not visible (fig. 1G), palpus The paratype in piece 131 was associated segment 3/4 (cid:49) 5 (cid:53) 1.27–2.00 (n (cid:53) 5). Hy- with a female Rhagionidae identified by Gri- popharynx not visible. Thorax: Most details maldi and Cumming (1999) as ‘‘Genus C’’. The implicationsof theseassociationsarein- notvisible.Anteriorscutalapodemesnotvis- terpreted below. ible. Scutum with a few scattered elongate setae. Wing (fig. 1E): Length (cid:53) 0.47–0.65 TAXONOMIC DISCUSSION: The male was as- mm (n (cid:53) 6). R well-developed to near sociated with the female on the basis of the 4(cid:49)5 presence of numerous thick spines on the wing apex. Costa well-developed to near wing apex, apex of R /wing length (cid:53) 0.38– first tarsomere of the hindleg (not present in 0.40 (n (cid:53) 3). Without3macrotrichia, fine mi- females of L. antiquus). The holotype of L. amplificatus was in crotrichia present on all membrane. Alula moderately good condition, but was missing without macrotrichia. Single radial cell pre- the apex of the left antenna and the apicesof sent in at least some specimens, radial veins all legs other than the right hindleg. The al- compacted anteriorly. Base of M poorly de- lotype was also in moderately good condi- fined,bifurcationnotvisible,bothM andM 1 2 tion, but was embedded so deeply in theam- present. Legs: Femora, tibiae slender. Legs ber as to make some details difficult to dis- with thick spines on first tarsomere of each cern. The remaining paratypes were gener- leg, a few on fore and midleg, many on hin- ally in moderately good to excellent dleg (fig. 1I). Pair of thick setae on fore-and conditionbutsomeofthoseinpiece101(fig. midleg trochanter not visible. Midleg tibia 3) were buried so deeply in the amber— withapicalspur.Hindfirsttarsomerewithout which cannot be cut further because of other thick basal spine or palisade setae. Foreleg, inclusions—that visibility was rather limited midleg, hindleg claws of equal size, length, for these. more or less evenly curved, withoutwell-de- The holotype of L. amplificatus was the veloped basal tooth (fig. 1J). Genitalia (fig. middle specimen in a group of three closely 2A): Spermathecae not visible. Details of approximated females (fig. 3). The paratype sternite 8, 9, segment 10 not visible. Cercus ofL.amplificatusinpiece79wasthatfemale laterally compressed, elongate, slender. most close to the holotype of Lonchopterites BIONOMICS: Based on the elongate cerci prisca (amber piece figured in Grimaldi and (likely used to oviposit in sand) and the gen- Cumming, 1999: 83, showing the positionof eral habitat of extant members of the genus, the specimens). L.amplificatusprobablybredinsandy,saline One female paratype in piece no. 101 habitats (Borkent, 1995). The presence of (specimenno.4)appearedtohavesutureson finely serrate mandibles and laciniaewithre- the frons/vertex (fig.1F),butthesemayhave trorse teeth strongly suggests that females of been artifacts of preservation. If truly pre- thisspeciesfedonvertebrateblood(Borkent, sent,thesewouldbeuniquewithinthegenus. 1995, 1996). One female paratype of L. am- TYPES: Holotype, female adult in amberin plificatus in piece 101 (specimen No. 3, fig. plasticbox,labeled‘‘HOLOTYPELeptocon- 3) had a bloated abdomen and, considering ops amplificatus Borkent’’, ‘‘PARATYPES the evidence from the mouthparts, this is Leptoconops amplificatus Borkent: 8 fe- 6 AMERICAN MUSEUM NOVITATES NO. 3328 Fig. 2. A. Structures of female Leptoconops amplificatus. B–G. Structures of female Leptoconops antiquus. A. Cercus, lateral view (no. 101, holotype). B. Wing (no. 125, paratype). C. Antenna (no. 125, holotype). D. Apex of mouthparts (no. 125, holotype). E. Palpus (no. 125, paratype). F. Fifth tarsomere of foreleg (no. 125, paratype). G. Cerci, lateral view (no. 125, holotype). males’’, ‘‘Amber: N. Lebanon, Antoni Este- DERIVATION OF SPECIFIC EPITHET: The phan Coll. Bchare Mtn., 2300 m’’, ‘‘AM- name amplificatus (enlarged, extended) re- BER: Lebanon Lower Cretaceous (Neocom- fers to the presence of 13 flagellomeres in ian) No. 101, Amer. Mus. Nat. Hist., Inclu- the female antenna. sion(s): female Leptoconops’’ (AMNH); allotype, male adult, labeled ‘‘ALLOTYPE Leptoconops (Palaeoconops) antiquus Leptoconops amplificatus Borkent’’, ‘‘PAR- Borkent, new species ATYPE Leptoconops amplificatus Borkent’’, Figures2B–G ‘‘HOLOTYPE Leptoconops antiquus Bor- kent’’, ‘‘Amber: N. Lebanon, Antoni Este- DIAGNOSIS: Male. Unknown. Female. The phan Coll. Bchare Mtn., 2300 m’’, ‘‘AM- only species of Ceratopogonidae with 13 fla- BER: Lebanon Lower Cretaceous (Neocom- gellomeres and an elongate and basally ian) No. 79, Amer. Mus. Nat. Hist., Inclu- broadened (anteroventrally) cercus. sion(s): 3 Ceratopogonidae, 1 Empidoid’’ DESCRIPTION: Female. Head: Most details (AMNH); paratypes, 10 females: 8 in same not visible. Antenna with 13 separate flagel- piece with holotype; 1 in same piece as al- lomeres, antennal ratio (cid:53) 0.78–0.89 (n (cid:53) 2), lotype; 1 from Bchare Mtn., Lebanon, piece terminal flagellomere more elongate than No. 131 (AMNH). preceding flagellomeres (fig. 2C), first fla- 2001 BORKENT: LEPTOCONOPS 7 the genus, L. antiquus likely bred in sandy, saline habitats (Borkent, 1995). The holotype of L. antiquus had a number of associations in the same piece of amber: a male and female of L. amplificatus and the holotype of a possible Lonchopteridae (Lon- chopterites prisca). The presence of laciniae with retrorse teeth suggests that females of thisspeciesfedonvertebrateblood(Borkent, 1995, 1996). The significance of these asso- ciations is interpreted below in the discus- sion. The amber holding the paratype also included a small, poorly preserved arachnid (spider or possibly a mite?). TAXONOMICDISCUSSION:TheallotypeofL. antiquus is that specimen of Leptoconops in Fig.3. Amberpieceno.101showinglocation piece no. 79 most distant from the holotype of inclusions. Numbers refer to paratypes of Lep- of Lonchopterites prisca; the amber piece toconops amplificatus. Not all other non-cerato- was drawn in GrimaldiandCumming(1999: pogonid inclusions are shown. 83), illustrating the position of the speci- mens. gellomere sensilla not visible. Mouthparts The holotype of L. antiquus is quite de- moderatelyelongate.Clypeusnotclearlyvis- composed and part of the head and thorax ible. Labrum, mandible not visible. Lacinia are obscured by contaminants; the left wing with at least seven large, retrorse teeth. Pal- is missing. Much of the body of theparatype pus with 4 segments, third segment large, is collapsed and portions appeared distorted. somewhat ovoid, capitate sensilla not visible The paratype of L. antiquus differs from (fig. 2D, E), palpus segment 3/4 (cid:49) 5 (cid:53) 1.80 the holotype in some details:flagellomere12 (n (cid:53) 1). Thorax: Most details not visible. is more elongate, third palpal segment is Wing (fig. 2B): Length (cid:53) 0.59 mm (n (cid:53) 2). more spherical (fig. 2 D–E), and the cerci R well-developed to near wing apex. Cos- may be more elongate. These differences 4(cid:49)5 ta well-developed to nearwingapex,apexof mayindicatethepresenceoftwospecies,but R /wing length (cid:53) 0.38 (n (cid:53) 1). Without ma- may also be due to artifacts of preservation, 3 crotrichia, fine microtrichia present on all especially considering that both specimens membrane. Alula without macrotrichia. were in only moderate condition. Without radial cell, radial veins compacted TYPES: Holotype, female adult in amberin anteriorly. Base of M poorly defined, point plasticbox,labeled‘‘HOLOTYPELeptocon- of bifurcation not visible, both M and M ops antiquus Borkent’’, ‘‘ALLOTYPE Lep- 1 2 present. Legs: Femora, tibiae slender. Legs toconops amplificatus Borkent’’, ‘‘PARA- mostly lacking armaturebut withfourtofive TYPE Leptoconops amplificatus Borkent’’, thick spines on fifth tarsomere of hindleg. ‘‘Amber: N. Lebanon, Antoni EstephanColl. Pair of thick setae on fore- and midleg tro- Bchare Mtn., 2300 m’’, ‘‘AMBER: Lebanon chanter not visible. Midleg tibia with apical Lower Cretaceous (Neocomian) No. 79, spur. Hind first tarsomerewithoutthickbasal Amer. Mus. Nat. Hist., Inclusion(s): 3 Cera- spine or palisade setae. Foreleg, midleg,hin- topogonidae, 1 Empidoid’’ (AMNH); para- dlegclawsequal,moreorlessevenlycurved, type, 1 female, labeled ‘‘PARATYPE Lep- inner tooth not visible (fig. 2F). Genitalia toconops antiquus Borkent’’, ‘‘Amber: N. (fig. 2G): Spermathecae not visible. Details Lebanon, Antoni Estephan Coll. Bchare of sternite 8, sternite 9, segment 10 not vis- Mtn., 2300 m’’, ‘‘AMBER: Lebanon Lower ible. Cercus laterally compressed, elongate, Cretaceous (Neocomian) No. 125, Amer. broad basally (anteroventrally). Mus. Nat. Hist., Inclusion(s): female Lepto- BIONOMICS: Based on the elongate cerci conops?’’ (AMNH). and the general habitat of extant membersof DERIVATION OF SPECIFIC EPITHET: The 8 AMERICAN MUSEUM NOVITATES NO. 3328 name antiquus (ancient) refers to the old age of this Lebanese amber fossil species. CONCLUSIONS The discovery of the oldest members of Leptoconops in 120–122 million year old Lebaneseamberleadstothequestionoftheir relationship to otherspeciesofthegenusand what they tell us about the relationship of Leptoconops to other Ceratopogonidae. Two features suggest that the subgenus Palaeoconops, with the two new species, L. amplificatus and L. antiquus, formsthesister group to all other Leptoconops (fig. 4), as follows. 1. Female antenna with 13 flagellomeres (plesiomorphic); female antenna with 10– 12 flagellomeres (apomorphic). Leptoconops amplificatus and L. antiquus aretheonlymembersofLeptoconopswith13 Fig. 4. Phylogeny of the basal lineages of flagellomeres in the female antenna; all other CeratopogonidaebasedonBorkent(2000)andthe speciesinthegenushave10–12flagellomeres resultsofthispaper.Numbersinparenthesesrefer (fig. 4). Borkent (2000: 399) criticized Szad- to number of female flagellomeres. ziewski’s (1996) use of this feature as evi- dence of the monophyly of Leptoconops, be- cause outgroup comparisons with other Culi- developed and is about as thick as the other comorpha are inconclusive (Borkent et al., anterior wing veins. In Leptoconops, other 1987: 600). However, the recognition of the than L. amplificatus and L. antiquus, R is 4(cid:49)5 extinct genus Lebanoculicoides as the earliest present but very thin (often it can only be lineage of Ceratopogonidae (the sisterngroup seen with phase contrast). In L. amplificatus of all other Ceratopogonidae, including Lep- and L. antiquus, R is well- developed and 4(cid:49)5 toconops), with 13 flagellomeres, and the thickerthaninanyothermemberofthegenus presence of 13 flagellomeres in the sister and this is additional evidence that these two group of Leptoconops, suggests that Szad- species form the sister-group of all remaining ziewski (1996) was correct (fig. 4). Leptoconops. The recently described Jordanoconops A feature that is very likely related to the Szadziewski (Szadziewski, 2000) is placed as well-developed R in L. amplificatus and L. 4(cid:49)5 the sister group of Austroconops on the basis antiquus is the presence of a well-developed of the shared presence of r-m being more or costa extending right to the apex of R . In 4(cid:49)5 less parallel to R and R (see character 11 in other Leptoconops, the costa is not or only 1 3 Borkent, 2000). poorly developed beyond the apex of R. Be- 3 cause a well-developed costa extends to (or 2. Wing with well-developed R (plesio- 4(cid:49)5 beyond) the apex of R in Lebanoculicoides morphic); R thin and faint, very poorly 4(cid:49)5 4(cid:49)5 and members of all other families of Culico- defined or absent (apomorphic). morpha, it is likely that this feature is ple- Borkent (2000) used this feature as evi- siomorphicinL.amplificatusandL.antiquus. dence for the sister-group relationship be- The cladistically primitive position of L. tweentheextinctLebanoculicoidesandallre- amplificatus and L. antiquus indicates that maining Ceratopogonidae (fig. 4) and dis- they cannot be considered to belong to any cussed the character state distribution within of the previously recognized subgenera of the family. In Lebanoculicoides, R is well- Leptoconops and they are thereforeplacedin 4(cid:49)5 2001 BORKENT: LEPTOCONOPS 9 the new subgenus Palaeoconops. However,I which R may be present (but in a different 4(cid:49)5 cannot present any synapomorphy for the form). Further study is needed. new subgenus and there is no evidence that Female adults of Leptoconops and the ex- L. amplificatus and L. antiquus actuallyform tinctgeneraFossileptoconopsSzadziewskiand a monophyleticgroup.Itislogicallypossible Alautunmyia Borkent are the only Ceratopo- that one is the sister-group of the other plus gonidae that have a fused frons and vertex all remaining Leptoconops. without any sutures. Borkent (2000) pointed There are presently six other subgenera out that the character is presently uninterpret- (all extant) recognized within Leptoconops: able cladistically. One specimen of L. ampli- Leptoconops sensu stricto, Brachyconops ficatus may have some sutures on the frons/ Wirth and Atchley, Holoconops Kieffer, Me- vertexarea(thesemaybeartifactsduetoshriv- gaconops Wirth and Atchley, Proleptocon- eling and/or compression) and if truly present, ops Clastrier, and Styloconops Kieffer (Bor- these would be unique within the genus. kent and Wirth, 1997). There has been no Borkent (2000) presented a model for in- cladistic analysis of these taxa other than the terpreting fossil communities of Ceratopo- interpretation by Borkent (1995), who hy- gonidae based on information on their sex pothesized that Leptoconops sensu stricto, ratio and, in single pieces of amber, associ- Megaconops, and Proleptoconops form a ationsofconspecificindividuals,associations monophyletic group. The single synapomor- between species of Ceratopogonidae,andas- phy supporting this was the shared presence sociations with organisms other than Cera- of a highly modified,elongatefemalecercus. topogonidae. Because Lebanese amber cera- Members of Brachyconops and Styloconops topogonids had a sex ratio typical of thoseat haveshortcerci,theconditionpresentinoth- extant emergence sites (about 39% males), ergeneraofCeratopogonidaeandfamiliesof no intraspecific associations, a high number Culicomorpha. However, the presence of of interspecificassociations,andahighnum- elongate cerci in L. amplificatus and L. an- ber of associations with non-ceratopogonids, tiquus strongly suggests that the short cerci it was deduced that the Ceratopogonidae of Brachyconops and Styloconops are actu- were locally diverse but not abundant. This allysecondarilyreduced.Carefulmorpholog- pattern of high diversity and low abundance ical study of these two subgenera should be is similar to that found in most tropical hab- undertaken with this in mind. itats today. The 14 specimens of Leptocon- Borkent (2000) discussed three adult syn- ops in Lebanese amber do not alter this gen- apomorphies for the genus Leptoconops: eral conclusion. However, L. amplificatus is ‘‘wing lacking r-m’’, ‘‘female wing with ra- the first of 24 species of Ceratopogonidae in dius joining costa as a strong thickened stig- the amber (with a total of 140 specimens) to ma’’, and ‘‘posteromedial margin of female haveanyintraspecificassociationsinasingle sternite 8 with semicircular concavitybearing piece of amber: one piece had nine females 4 or more stout setae on its margin’’. Based and another held one male and one female. on the discussion above, the elongate cercus The striking presence of nine specimens of ofthefemaleislikelyafourthsynapomorphy. one species in one piece of amber in a fossil The presence of a clearly defined R in L. assemblage otherwise hypothesized to be di- 4(cid:49)5 amplificatus and L. antiquus may indicate that verse but with low species abundance, is Leptoconops is, other than Lebanoculicoides, likely explained by the habitat generally oc- thesister-groupofallremainingextantandex- cupied by species of Leptoconops. The im- tinct Ceratopogonidae. If so, R has been at matures of most extant members of this ge- 4(cid:49)5 least partially reduced at least twice: once nusarerestrictedtobeachhabitats,wherethe within Leptoconops (where in many species it larvae burrow in moist/wet sand. This is one can only be seen with phase contrast micros- of the few, more or less homogeneous habi- copy) and once within other Ceratopogonidae. tats of any size in lowland tropical regions In addition, as discussed by Borkent (2000), and it seems probable that the immatures of there are some problems of determining ho- the fossil L. amplificatus were restricted to a mologies with some Forcipomyiinae, some similar habitat. This likely explains the Dasyheleinae and some Ceratopogoninae in unique intraspecific associations within this 10 AMERICAN MUSEUM NOVITATES NO. 3328 species. The presence of nine L. amplificatus less of whether they were modified or not. in a single piece of amber can also be taken Features are arranged from anterior to pos- as evidence that the original resin-producing terior and from dorsal to ventral. Araucariaceae (Bandel and Vavra, 1981) likely grew close to the seashore, at least as Male antennal flagellomere 13 with subbasal part of its distribution. constriction (figs. 7f, 8d, 9b, 12f, 14c in Borkent (2000) provided a key to the gen- Borkent, 2000; fig. 1A): species of Lep- era of Ceratopogonidae in Lebanese amber. toconops, Archiaustroconops Szadziewski The discovery oftheLeptoconopsspecimens and Austroconops Wirth and Lee. described here requires modification to that Female antennal flagellomere 13 with apical key as follows. elongate projection (fig. 6f, g in Borkent, 2000): Protoculicoides schleei (Szad- 1. Wing with R4(cid:49)5 clearly present (fig. 1a, c in ziewski),P. punctusBorkent,P.succineus Borkent, 2000; figs. 1E, 2B) .......... 2 Szadziewski. – Wing without R visible (fig. 4a, c in Bor- 4(cid:49)5 Ommatidia broadly separated dorsomedially kent, 2000) ........................ 3 (Szadziewski, 1996: fig. 18b): Leptocon- 2. Wing with first and second radial cells clearly ops, Fossileptoconops lebanicus Szad- defined (fig. 1a, c in Borkent, 2000) ...... ........... Lebanoculicoides Szadziewski ziewski. – Wingwithfirstandsecondradialcellsreduced Female mouthparts moderately to very elon- and/or compacted (figs. 1E, 2B) ....... gate (fig. 6a in Borkent, 2000; Szadziews- .... Leptoconops (Palaeoconops), n. subg. ki, 1996: figs. 3b, 4b): Lebanoculicoides 3. Ommatidia broadly separated dorsomedially; mesozoicus Szadziewski, Protoculicoides cercus with elongate terminal setae (fig. 2a succineus, P. unus Borkent, P. punctus, P. in Borkent, 2000) (known only as female) schleei. .......... Fossileptoconops Szadziewski Mouthparts extremely short (fig. 10c in Bor- – Ommatidianarrowlyseparatedorabuttingdor- kent, 2000): Archiaustroconops bocapar- somedially (figs. 7a, 8j, 10c, 11a, 20b in vus Borkent, Archiaustroconops creta- Borkent, 2000); cercus without strikingly elongate setae (fig. 7j in Borkent, 2000) .. ceous (Szadziewski). .................................. 4 WingwithR present(fig.1a,cinBorkent, 4(cid:49)5 4. Wingwithr-mparalleltoR1(figs.11c,h,13a, 2000; figs. 1E, 2B): Lebanoculicoidesme- c, in Borkent, 2000) ................. sozoicus (thick vein); Leptoconops (Pa- ............ Austroconops Wirth and Lee laeoconops), n. subg. (thin vein). – Wing with r-m oblique to R (figs. 4a, 5a, 15e 1 Wing with r-m parallel to R (figs. 11c, h, in Borkent, 2000) .................. 5 1 12h,13a,c,14ainBorkent,2000):species 5. Wingwithpoorlydefinedradialcell(figs.15e, of Austroconops. 17a, 19f, 20a in Borkent, 2000); female CR (cid:44) 0.5 ............. Minyohelea Borkent Wing with reduced wing venation (radial cells not well-defined) (figs. 15e, 17a,19f, – Wing with radial cell(s) well-developed (figs. 4a, c, f, 5a in Borkent, 2000); female CR (cid:36) 20a in Borkent, 2000; figs. 1E, 2B): Lep- 0.8 .............................. 6 toconops (Palaeoconops), n. subg., Min- 6. Male flagellomere 13 with subbasal constric- yohelea Borkent. tion (figs. 7f, 8c, d, 9b in Borkent, 2000); Radialcellsingle,well-defined,elongate(fig. tarsalratio(Ta/Ta)offoreleg/hindleg(cid:36)1.4 4a, c, f in Borkent, 2000): Protoculicoides 1 2 ........ Archiaustroconops Szadziewski acraorum Borkent, P. schleei, P. unus, – Male flagellomere 13 without subbasal con- possibly F. lebanicus. striction (figs. 2e, 5c in Borkent, 2000); tar- Foreleg tibial spur very large (figs. 14f, 15a sal ratio of foreleg/hindleg (cid:35) 1.3 ...... in Borkent, 2000): Austroconops megas- ................ Protoculicoides Boesel pinus Borkent. The presence of Leptoconops in Lebanese First tarsomere of both fore and hindleg with amber also requires modification of the list thick basal spine (fig. 12b, d, i, k in Bor- of unique or distinctive character states use- kent, 2000): Austroconops gladius Borkent fulforidentifyinggeneraandspeciesofLeb- Each claw with well-developed inner tooth anese amber Ceratopogonidae provided by (figs. 9d, 11d, e, 12e, g in Borkent, 2000): Borkent (2000). All are given below, regard- Archiaustroconops hamus Borkent, Aus-