Novtates PUBLISHED by the AMERICAN MUSEUM of NATURAL HISTORY CENTRAL PARK WEST at 79TH STREET, NEW YORK, N.Y

AMERICAN MUSEUM Novtates PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024 Number 2880, pp. 1-23, figs. 1-56 June 10, 1987

Amber Fossil Drosophilidae (Diptera), with Particular Reference to the Hispaniolan Taxa

DAVID A. GRIMALDI'

ABSTRACT All the known fossils ofDrosophilidae occurring Miomyia io, new genus; Protochymomyza mio- in amber are treated. Neotanygastrella wheeleri, cena, new genus; Scaptomyza dominicana; and new species, is named, redescribed (cf. Wheeler, two Drosophilinae species (A and B) incertae se- 1963) with respect to certain features, and a ho- dis. Relative genealogical relationships for some lotype and paratype are designated for the two drosophiline genera and subgenera are briefly dis- specimens from Chiapas, Mexico. Seven new cussed. Illustrations and photographs of the spec- species are described and two others are recorded imens are included, as well as ideas on the signif- from amber of the Dominican Republic (approx- icance ofsome ofthe fossils for Caribbean historical imately early Miocene in origin): Chymomyza pri- biogeography. A key to amber-fossilized Dro- maeva; Drosophila (D.?) poinari; D. (Hirtodro- sophilidae is provided. sophila) paleothoracis; D. succini (incertae sedis);

INTRODUCTION Seven or eight fossilized specimens ofDro- will not be considered here. Two ofthe fossils sophilidae were previously known and all of are of a Neotanygastrella species from Chia- them occur in amber. Four (two males and pas, Mexico (late Oligocene to early Miocene two females) are Electrophortica succini Hen- in origin), which were described by Wheeler nig from the Baltic amber (Eocene to early (1963). Oligocene in age). Hennig (1965) gave very Cockerell (1923) described Drosophila ber- complete and detailed descriptions and draw- ryi from a specimen in amber found in Valle ings of E. succini, and Grimaldi (1987) dis- de Jesus, Santander, Colombia. He noted (p. cussed the placement of the species. This fly 331) that the "material, unfortunately of un-

' Assistant Curator, Department of Entomology, American Museum of Natural History.

certain age, is relatively soft, being easily cut are certainly the youngest in origin (Lambert with an ordinary knife." Langenheim (1969) et al., 1985). In general, though, an early Mio- mentioned two Colombian sites among the cene origin of most Dominican amber is a amber deposits known in the world, at Me- reasonable estimate (Brouwer and Brouwer, dellin and Giron (Antioquia Provincia), of 1982). This dating is based on the fact that unknown Tertiary age, and that Hymenea the amber deposits are intercalated between (Leguminosae) is the probable source. The marine microfossil deposits that were laid location of Cockerell's specimen is unknown down approximately 23 mya, so the date is to me. Softness of the material strongly sug- one of redeposition and is, therefore, a min- gests the fossilized resin to be very recent, imal age. and perhaps it is even copal (1000 or more Langenheim and Beckh (1968) first ana- years old). If found, the matrix of the speci- lyzed the resinous components of various men should be examined for its C'4 decay ambers and found that material from the Do- (Burleigh and Whalley, 1983). The original minican Republic closely matched Hymenea description is actually of sufficient detail to (Leguminosae), which is a genus of trees confirm the specimen's identity as a drosoph- widespread throughout Central and South ilid, but an identification beyond this is im- America and the Greater Antilles. Lambert possible at present. Loew (1850) mentioned, et al. (1985) found, however, three major but did not describe, a "Drosophila" from classes of resinous components among sam- Baltic amber. This was the fly later described ples from nine Dominican amber localities. by Willi Hennig, Electrophortica. Bachofen- They concluded that ifHymenea is the source Echt (1949) listed three known amber fossil of the amber then the fossil resin is more drosophilids, no doubt referring to Loew's variable than samples from extant Hymenea. specimen, to the other specimens later de- Also, relative dating of the amber was done scribed by Hennig, and to Cockerell's spec- based on the NMR intensity ofthe exomethy- imen. lene carbon in the sample. Other than the This study is an attempt to add some dates provided by Brouwer and Brouwer chronological data to emerging hypotheses (1982), the actual dates of origin for various on the phylogenetic relationships and histor- Dominican ambers are rather hypothetical. ical biogeography ofthe Drosophilidae. Gri- maldi (1987) hypothesized that some of the Antillean drosophilid lineages are relicts; that METHODS, MATERIALS, AND is, their endemicity is imposed by extinction. ACKNOWLEDGMENTS Specimens from the rich amber deposits of All amber pieces from the Dominican Re- the Dominican Republic (Schlee and Glock- public had been previously tumble-polished. ner, 1978; Baroni-Urbani and Saunders, For closer observation ofsome details, amber 1982) and elsewhere provide information on was ground down close to the specimen using minimal age and perhaps on extinctions (e.g., a lapidary wheel and various-size grinding Krishna and Emerson, 1983; Wilson, 1985) grits (Buhler 240-1000) with water; amber for some groups ofthe Antillean pomace flies. was polished with a white alumina powder The former aspect would be particularly use- (#2 Buhler-micropolish 2A) plus water. The ful for comparison to the postulated dates of fossils were examined using incident and re- geological origin and position of Caribbean flected light from incandescent and fiber op- land masses. tics lamps, generally at 25-50 x magnifica- Amber of the Dominican Republic origi- tion using a Zeiss SV-8 stereoscope with nates from at least 10 major mining sites lo- attached camera lucida. Specimens were po- cated throughout the eastern three-quarters sitioned by placing them on cotton that was ofthe country at various altitudes. As a result, molded to accommodate the shape of each geological processes such as sedimentation amber piece, generally in an oil medium with have had different effects upon each site, and a refractive index equal to that ofthe amber. the calculated ages of the amber deposits can Photography employed the use of an Olym- vary among sites. Virtually transparent pieces pus OM-2N camera with auto-bellows and a of amber, which come from mines at Cotui, Zuiko macrolens (f6, about 12 x magnifica- 1987 GRIMALDI: AMBER DROSOPHILIDS 3 tion) and a Zeiss Tessovar (1 sec exposure), thoracic leg with femur, tibia, and proximal and Kodak Pan-X film. The standard mea- 3/4 of basitarsus darkened; remaining seg- surements made, where possible, were total ments of leg light. Abdomen dark brown to length ofbody (TL), thorax length (ThL), wing black. Oviscape without peg ovisensilla (male length (WL), head width (HW), 4-vein index unknown). (4-V), and costal index ofwing (CI). Protocol DESCRPTrION: Head light yellow. Arista with for these measurements is given in Grimaldi 3 dorsal and 2 ventral branches and 5-6 short (1986). Measurements were made at 50x medial ones. Pedicel with 2 long setae and magnification with an ocular micrometer. several shorter ones. Ocellar setae extended Dr. David Lindberg ofthe Museum ofPa- about to facial margin. Face flat; without ca- leontology, University of California, Berke- mna. One pair ofvibrissae present; ends near- ley, kindly loaned the two hypotypes ofNeo- ly touch; subvibrissae tiny. Clypeus and palps tanygastrella described by Marshall Wheeler. retracted within oral cavity. Eyes light red, Mr. Jacob Brodzinsky allowed me to sort bare of interfacetal setulae; collapsed in ho- through his collection forthe purchase ofsome lotype. Inner vertical setae c. equal in length specimens. Dr. George 0. Poinar, Depart- to outer verticals and anteromedial to them. ment of Entomology and Parasitology, Uni- Proclinate orbital about midway between an- versity of California, Berkeley, generously terior and posterior reclinates and c. equal in loaned several specimens from his personal length or slightly shorter. collection. Dr. Wayne N. Mathis, Depart- Acrostichal setulae in 4 even rows. Dor- ment of Entomology, National Museum of socentrals thin and straight, parallel. Two hu- Natural History, sorted through the collec- meral setae present. Anterior scutellar setae tion at the Smithsonian, from which five am- slightly divergent; apical scutellars conver- ber pieces were borrowed. Drs. Gerhard gent. Postnotum prominent. Halter lighter Bachli, Ian Bock, Walter Hackman, Wayne than evenly colored, yellow thorax. Wing Mathis, Toyohi Okada, George Poinar, Nor- completely hyaline (wings are folded in ho- man Platnick, Randall Schuh, and Marshall lotype in way to prevent accurate measure- Wheeler provided comments on the manu- ments). Meso- and metathoracic legs mostly script, which Ms. Carol Ievolella kindly typed. yellow. Prothoracic leg with yellow coxa and Ms. Susan Klofak helped with the amber trochanter; tarsal segments 2-5 lighter, nearly preparations. white. Forefemur, tibia, and proximal 3/4 of Abbreviations for collections in which basitarsus black-brown. Forecoxa elongate, specimens are deposited are the following: length c. 2/3 that of femur. Abdomen slender, AMNH, American Museum of Natural His- laterally compressed in holotype. Tergites tory (New York); GOP, George 0. Poinar, black-brown, with few setae. Oviscape ex- Jr., personal collection, Berkeley, CA (to truded, small, subtriangular, with setiform eventually be deposited at AMNH); NMNH, and no peg ovisensilla. Cercus small, conic, National Museum ofNatural History, Wash- with fine setae about as long as cercus. ington, DC; UCMP, University ofCalifornia HOLOTYPE: Female; DOMINICAN RE- Museum of Paleontology, Berkeley. PUBLIC, specific locality unknown (GOP). ETYMOLOGY: Latin for "early," or "young," in reference to the plesiomorphic condition RESULTS AND DISCUSSION of the wing as stated below. CHYMOMYZA CZERNY COMMENTS: The specimen on which the description is based is shown in a photograph Chymomyza primaeva, new species (fig. 1 in Poinar, 1984). In that paper the Figures 1-4 specimen was incorrectly identified as Dro- DiAGNosIs: Arista with 3 dorsal and 2 ven- sophila. The specimen is of considerable in- tral branches. Anterior reclinate slightly terest because ofthe parasitic nematodes that shorter than posterior reclinate. Proclinate apparently emerged from it when it was about 1/2 the distance between the anterior trapped in resin, and because of a recent re- and posterior reclinates. Acrostichal setulae vision ofneotropical Chymomyza (Grimaldi, in 4 rows. Wing completely hyaline. Pro- 1986). The quadrate head shape, placement 4 AMERICAN MUSEUM NOVITATES NO. 2880

1~~~~~~~~~~~~~~~~~~~~~~~~1

Figs. 1-4. Chymomyza primaeva, holotype. 1. Head, posterodorsal view; 2. dorsal habitus; 3. ventral view of head and anterior portion of thorax, with legs; 4. female terminalia, lateral view. of the proclinate nearly midway between the drosophilids. Six species of Chymomyza ex- other two frontal-orbital setae, the long fore- ist in Central America, and three species, coxae, the dark forefemora + tibiae + basi- which belong to the aldrichii species group, tarsi, and the completely dark abdomen are occur on the Greater Antilles. derived characteristics which make it clear that this specimen is Chymomyza. Unfor- tunately, it is a female; so unless a male is DROSOPHILA FALLEN found the presence and/or state ofthe ventral Apart from the type subgenus, 14 subgen- rows of prothoracic femoral spines must re- era have been proposed in the genus Dro- main unknown. Also, as Okada (1976) first sophila. Approximately 1500 species have hypothesized, the presence ofhyaline (not in- been placed in the genus, which includes about fuscate) costal and subcostal wing cells is ple- halfofthe world's Drosophilidae. Clearly, the siomorphic; I concur with this conclusion taxonomic history ofDrosophila has been one based on outgroup comparison. Since the fos- of convenience: although synapomorphies sil has an entirely hyaline wing, it and four have been proposed for several subgenera, at other species in the costata species group are the generic level Drosophila is distinguished the only species of the genus known to lack by the absence of traits that are distinctive the synapomorphy. to other drosophiline genera. The monophyly Chymomyza is a cosmopolitan genus of49 of the genus is seriously questionable just on species; the adults are found at injured areas these grounds. Throckmorton (1975), in fact, on tree trunks, so the finding ofthis specimen viewed Drosophila and several ofits subgen- in fossilized resin is not unexpected for the era as being ancestral to other drosophiline 1 987 GRIMALDI: AMBER DROSOPHILIDS 5 genera; so at least in his view, the genus is Also, it is probably best not to propose not monophyletic. higher taxa to accommodate the amber fos- Because I believe that taxonomic names sils. At least Cretaceous to Oligocene amber should be genealogically meaningful, and also fossil taxa are often quite primitive compared for the purpose of being explicit, I will em- to extant relatives (i.e, they branch early in ploy here a definition of Drosophila that is a cladogram: for examples see Gagne, 1977; narrower than the one in common use. At Hennig, 1965; McAlpine and Martin, 1966; least for this paper, Drosophila sensu stricto Wilson et al., 1967; Woodley, 1986; but for includes species that have lost the prescutel- more recent fossils see Petrunkevitch et al., lar setae, and possess densely micropubes- 197 1). So, in order to cladistically classify the cent eyes and an ovipositor (oviscape) with fossils, higher-level and, at least for amber heavily sclerotized, peglike ovisensilla. Pos- flies, monotypic categories would often need session of a broad, flat facial carina may be to be proposed. Even the best fossils, like a synapomorphy with which to link several amber specimens, are by nature incomplete, Drosophila subgenera, including Drosophila so nomenclatorial chaos can accompany the and Sophophora. Figure 5 gives a preliminary discovery and reinterpretation ofsingle fossil cladogram of drosophiline relationships. Of specimens or characters. Other than the two pressing importance would be a scanning genera, no supraspecific taxa of Drosophili- electron microscope study of eye pilosity dae were named here. among drosophilids, since this trait appears Electrophortica succini is apparently the to be quite important in higher relationships only amber drosophilid that is not in the Dro- but also seems to have arisen three times in sophilinae. Unlike the Steganinae, the Dro- the scheme in figure 5. Detailed comparison sophilinae may be monophyletic, as based on of fine structure of the eye surface would the loss of the prescutellar setae (this is un- probably reveal whether the dense eye pilos- doubtedly a derived feature since prescutel- ity is convergent or synapomorphic. Also, I lars occur in sister families, the Camillidae, have tried to restrict the use of characters in Curtonotidae, Diastatidae, and Ephydridae). figure 5 to external features, such as those The other diagnostic feature, loss of one of applicable to the study ofamber fossils. More the usual two katepisteral setae, I have found detailed morphological analyses will be the to occur very sporadically in the subfamily: subject of future work. Also, because figure more commonly observed is a state where 5 is very incomplete in its treatment of most the katepistemals are unequal in size. The drosophilid taxa, statements on higher clas- presence of stout, heavily sclerotized, ovi- sification in the Drosophilidae are deferred, sensilla pegs on the oviscape is also a derived pending comprehensive comparisons. trait possessed by the Drosophilinae, but not I adhere to Patterson and Rosen's (1977) by all genera in the subfamily. arguments regarding the placement and naming of fossils into a classification composed mostly or wholly of extant taxa. Specifically, Drosophila (Drosophila?) succini, the fossils were treated cladistically; that is, new species where evidence allowed, they were placed as Figures 6-8, 12, 13 sister-groups to extant taxa. This is the most DIAGNOSIS: Anterior reclinate orbital di- general phylogenetic statement; should fur- rectly lateral to proclinate; vertex slightly ther evidence ever prove compelling, ances- raised; carina low; vibrissae in 2 pairs; eyes tral-descendant relationships may then be with dense micropubescence; ovisensilla pegs proposed. In my opinion, no taxonomic present (male unknown). problem exists with the placement and nam- DESCRIPTION: Integument light yellow to ing of ancestors, for they should be grouped tan, much of it obscured by a silvery coating with their descendants in the strict definition resulting from the embedding. Arista with 5 of "monophyletic." The Gordian knot is ac- dorsal and 2 ventral branches. Anterior sur- tually how and on what criteria ancestors face of pedicel with 2 stout setae. Flagello- should be recognized and, once these are pro- mere I with short, even setulae. Anterior rec- posed, predicting the level at which a taxon linate orbital seta c. 1/3 length of proclinate, (species or otherwise) is ancestral. lies just lateral to proclinate. Posterior recli- 6 AMERICAN MUSEUM NOVITATES NO. 2880

Fig. 5. Hypothesis of relative genealogical relationships for some major drosophiline genera and some subgenera of Drosophila. Numbers are synapomorphies (see table 1): those that are denoted by squares occur several times in the cladogram and may be convergent. Several fossil Drosophilidae are placed in the scheme. nate seta c. equal in length to proclinate; dis- tae. Two pairs of scutellars present, orienta- tances between setal bases c. l/2 length of an- tions distorted. One large notopleural pres- terior reclinate. Ocellars, inner and outer ent, directed posteriad; 1 small notopleural vertical setae well developed. Vertex raised (c. 1/3 length of other), erect. Ventral surface above dorsal margin ofeye. Carina low. Two of profemur with 3 setae, lengths c. equal to pairs of vibrissae present; each pair subtend- width of femur and located on distal 1/3 of ed by c. 4 smaller subvibrissae. Proboscis dis- segment. Middle tibia with ventroapical seta. tended, membranous; clypeus large and Wing hyaline; heavy costal-radial setulae end broad. Palp with 1 apical seta, 3 small ventral midway between R213 and R4+5. Cercus short, setulae. Mentum slightly sclerotized, pos- conic; oviscape broad in lateral view, later- sesses 2 elongate, erect, and fine setae. Setae ally flattened. Oviscape with 12-13 ovisen- apparently only on anterior surface of label- silla pegs on distal and ventral edges, plus 1 lum. Eyes with dense micropubescence. distomedially. One long ventrodistal, seti- Dorsocentral setae in 2 pairs; length of an- form ovisensilla also apparent. terior ones l/2 that ofposterior setae. Distance MEASUREMENTS: WL = 1.90 mm; CI = between anterior to posterior dorsocentral on 2.33; 4-V = 2.0. each side c. equal to 3/4 length of anterior se- HOLOTYPE: Female (NMNH, no. 8512). 1 987 GRIMALDI: AMBER DROSOPHILIDS 7

Origin ofthe amber from the Dominican Re- TABLE 1 public unknown. Character States on the Cladogram in Figure 5 ETYMOLOGY: From Latin, meaning "sun," (Circles on the cladogram are synapomorphies; in reference to the mythical source of amber. squares are derived features occurring several times COMMENTS: The specimen resides in a larg- that may be convergent.) er amber piece (fig. 13) in which there are 1. Loss of prescutellar setae. trapped numerous insects, including a dry- 2. Ovisensilla are stout, peglike, and heavily sclero- inine dryinid female (Hymenoptera), several tized. psychodids (Diptera), and three drosophilids. 3. Interfacetal (eye) setulae dense. Only the specimen described here is suffi- 4. Anterior reclinate orbital seta anterior to proclinate ciently visible (mostly in lateral view) to mer- orbital and lateral to it by varying degrees. it description. The large amber piece is rather 5. Head broad, flat, trapezoidal in dorsoventral view. dark, which indicates that the source is a rel- 6. Forefemora and tibiae black, contrast with tarsi and coxae. atively old amber deposit. 7. Abdominal tergites black, shiny. 8. Male fore- or midfemora with 2 ventral rows of Drosophila (Drosophila?) poinari, spines. new species 9. Facial carina bulbous. Figures 9-11, 14-16 10. Adults mycophilous; feed and rendezvous at mac- rofungal sporophores. DIAGNOSIS: Carina broad and flat; 1 pair 11. Facial carina, when present, narrow. of vibrissae present; eyes densely micropu- 12. Costal incision on wing deep, formed into lappet. bescent; stemites large, overlapping, sclero- 13. Costal lappet black. tized; oviscape with ovisensilla pegs (male 14. Acrostichal setulae in 2-4 rows between anterior unknown). dorsocentral setae. DESCRIPTION: Ground body color yellow to 15.? Two pairs of large (oral) vibrissae present (NB: light brown. Flagellomere I with short, even members ofDrosophila (D.) repleta and Drosophila setulae. Arista bears 4 dorsal, 2 ventral (Sophophora) obscura species groups retain the ple- flat, broadened at oral siomorphic state of I pair ofvibrissae; these groups branches. Carina low, may require a revised taxonomic placement). margin. Oral margin with raised "lip." One 16. Preapical setae lost on mid- and foretibiae. pair ofvibrissae, subtended by 5-6 finer, small 17. Facial carina prominent, broad, and flat. setulae. Clypeus shallow, broad in ventral 18. Egg with 2 anterior respiratory filaments. view. Palpus narrow; with 3 setae (1 subapi- 19. Male fore basitarsus with comb ofstout, sclerotized cal, 2 ventral), lengths equal c. to palp width. setae (also occurs in subgenus Lordiphosa). Labellum broad, fleshy. Eyes densely micro- 20. Egg with 4 anterior respiratory filaments. pubescent. Proclinate is longest frontal-or- 21. Carina lost (some Hirtodrosophila). bital seta; anterior reclinate is l/2 length of posterior reclinate orbital. Anterior reclinate lies slightly posterolateral to proclinate. Pos- terior reclinate slightly closer to proclinate lengths from 1/3 to 1 times width of femur. than to inner vertical. Inner verticals strongly Wings hyaline. Heavy costal-radial setae end convergent; outer verticals strongly diver- just before midline between R213 and R415. gent; both equal in length. Abdominal terga brown; median portion Thorax with 2 pairs of dorsocentrals. An- tergites 3-5 with diffuse yellow spot which teriors c. 1/2 length ofposterior ones; separated occupies entire length of tergite. Thorax and from posterior pair by distance c. equal to head are lighter brown, shiny. Sternites scler- length of anterior dorsocentrals. Apical scu- otized, large and overlapping; lighter than tellars cruciate. One humeral (postpronotal), tergites. Pleural membrane white. Ovipro- 1 notopleural, 1 anepimeral, 2 katepimeral vector everted. Oviscape laterally flattened; setae present. Supra-alar seta elongate. Kat- with row of small, ventral peg sensilla. Apex episternum with row of 7 setulae running be- of oviscape obscured. tween 2 large setae. Midtibia with ventroapi- MEASUREMENTS: ThL = 1.0 mm; HW = cal seta, 1 shorter preapical dorsal seta. Hind 0.65 mm; WL = 1.65 mm; CI = 2.6; 4-V = tibia with a preapical, dorsal, erect seta. Fore- 1.9. femur with 5-7 erect, ventrolateral setae; HOLOTYPE: Female; in light piece ofamber 8 AMERICAN MUSEUM NOVITATES NO. 2880

\-=' 6

vb-

11 Figs. 6-8. Drosophila (D.?) succini, holotype. 6. Lateral view of head; 7. wing; 8. lateral view of abdominal apex. Figs. 9-11. Drosophila (D.?) poinari, holotype. 9. Head, oblique frontal view; 10. head and thorax, oblique dorsal view; 11. abdomen, ventral view, with oviprovector everted through oviscape. Abbre- viations: arc, anterior reclinate seta; ce, cercus; cl, clypeus; cr, (facial) carina; iv, inner vertical seta; oc, ocellar seta; ov, outer vertical seta; prc, posterior reclinate seta; sV-VII, stemites V-VII; vb, vibrissae. 1 987 GRIMALDI: AMBER DROSOPHILIDS 9

h 13

'A 12 Figs. 12, 13. 12. Drosophila succini (NMNH 8512) close up. 13. NMNH amber piece no. 8512, showing position of holotype (h) and dryinine dryinid (d). containing 2 specimens, collected in Domin- johnstonae, and perhaps other species as well. ican Republic near Cotui (GOP). Paratype is The tergal color pattern suggests an affinity other specimen in same amber piece. with the Drosophila (D.) tripunctata species- ETYMOLOGY: Patronym, in honor ofGeorge group, except that members ofthe group have Poinar. Dr. Poinar has been very gracious in distinct, not diffuse, spots and small sternites. providing his personal collection for this study. Drosophila (Hirtodrosophila) paleothoracis, COMMENTs: Identity ofthe paratype is based new species mostly on setation since the coloration is ob- Figures 17-24 scured by a milky film over much ofthe specimen, including the wings. These are prob- DIAGNOSIS: Facial carina low, rounded; 1 ably the youngest of the fossil drosophilid pair vibrissae; pleura light brown, notum yel- specimens based on the clear color of the low; dorsalmost ovisensilla peg with gap be- amber in which they are preserved. They may tween it and closest pegs in posteroventral belong to subgenus Drosophila or Sopho- row (male unknown). phora (the latter defined, in the strict sense, DESCRIPTION: Head entirely yellow-or- on the possession of foretibial combs in the ange, as is notum. Frontal vittae golden, fine- male, and 2 egg filaments). The large sternites ly striate, shiny. Arista with 4-6 dorsal and are a derived feature that occurs in some 2-3 ventral branches. Flagellomere I with members of the Drosophila robusta (i.e., D. short setulae. Eyes lighter than head, densely colorata) and D. virilis species-groups, in D. setulose. Facial carina low, edge rounded (not 10 AMERICAN MUSEUM NOVITATES NO. 2880

type, cruciate in holotype. Two humeral, 1 notopleural, 2 supra-alar, 1 large and 1 small katepisternal setae present. No katepisternal setulae. Legs light, mostly yellow. Forefemur with 2-4 ventral setae, lengths < width of femur. Halter light. Wings hyaline. Tergites mostly brown, with light areas laterad and dark rim on posterior edge. Cercus, tergite VII, t VIII, and oviscape (s VIII) lighter. Oviscape not fully everted in holotype, but everted in some paratypes. Tergite VIII ex- 14 tended down to oviscape, meeting it with a broad base in uneverted terminalia, possesses c. 4 sensilla. Oviscape with c. 18-20 stout, ovisensilla pegs; dorsalmost one separated from others in the row by a gap of about 2 pegs and slightly lateral to other pegs. Ster- nites light brown; unmodified in size and shape from typical, separated, quadrate sternites. MEASUREMENTS: (AMNH paratype) WL = 1.55 mm; CI = 1.9; ThL = 0.85 mm; 4-V = 2.22 (holotype). CI = 1.85; 4-V = 2.36. HOLOTYPE: Female (AMNH) (fig. 22). Source of the amber in the Dominican Re- public is unknown, but is very likely not the easternmost mines around Cotui since color of the amber piece is dark. PARATYPES: 9; NMNH 10664; 29 in amber piece from Palo Alto or La Toca mines, San- tiago, Dominican Republic (AMNH); 29 in amber piece of unknown locality in Domin- 16 ican Republic (GOP) (figs. 23, 24). Figs. 14-16. Drosophila (D.?) poinari. 14. Ho- ETYMOLOGY: From Greek, for "old" tho- lotype; 15. paratype; 16. amber piece showing both racis, due to the relationships as discussed specimens. below. COMMENTS: The oviscape is the key feature flattened). Ocellars longest setae on head, then that allows identification of this species. A inner verticals. Anterior reclinate orbital gap between the dorsalmost ovisensilla peg much closer to proclinate than to posterior and the others, in the ventroapical row, as reclinate; slightly lateral to proclinate by c. 1 well as the pleural coloration, suggest that the seta width. One pair ofvibrissae present. In- species belongs to the thoracis species-group traocellar area light, same as head ground of Drosophila (Hirtodrosophila). Capture of color. Postocellar seta length equal c. to length a drosophilid in this group at tree resin seems to anterior reclinates. Palpus slender, yellow. unusual because members of its group are Clypeus and anterior margin of face brown; found almost exclusively at fungal fruiting clypeus broad. Labellum small, c. 11/2 times bodies. width of prementum. Notum lighter than katepisternum, an- NEW GENUS epimeron, and anepisternum, which are light MIOMYIA, brown. Acrostichal setulae in 6 rows. Ante- DIAGNOSIS: One katepisternal seta present; rior dorsocentrals c. ½/2 length of posterior no prescutellar setae. Ventral epandrial lobe ones. Apical scutellars convergent in para- (male) large, pendulous; oral cavity wide; la- l1987 GRIMALDI: AMBER DROSOPHILIDS

arc ov ar dsc

h sra ntl 17 18 s

Figs. 17-21. Drosophila (H.) paleothoracis. 17. Head and thorax, oblique dorsal view; 18. posterior end of abdomen, lateral view. 19-21: holotype. 19. Wing; 20. terminalia, oblique posterior view; 21. head, oblique dorsal view. Abbreviations: same as in figures 6-9, plus the following: ar, acrostichal setulae; dsc, dorsocentral setae (anterior and posterior); ks, katepisternal seta; ntl, notopleural seta; sc, scutellar setae (anterior and posterior); sra, supra-alar setae. bium sclerotized and stout. Extinct (female geological epoch of the amber's origin; and unknown). -myta (Greek), or fly. ETYMOLOGY: Feminine; from Miocene, the GENOTYPE: Miomyia io, new species. 12 AMERICAN MUSEUM NOVITATES NO. 2880

black-brown areas on katepimeron, medial portions ofnotum, and the entire portions of the abdominal tergites. Eyes silvery (not red, as in most drosophilids), glabrous. Flagello- mere I length 1.5 x that of pedicel. Arista with 5 dorsal, 2 ventral branches. One vibrissa, subtended by c. 7-8 shorter, black setulae. Palpus bears 1 large apical seta and several smaller subapical ones. Width oforal cavity c. 3 x width of labellum. Proboscis short; almost entirely retracted into oral cavity. Labium about as broad as labellum, 22 heavily sclerotized. Face flat (no carina present). Verticals, anterior reclinates, and ocellar setae about as long as arista. Inner verticals convergent, outers divergent. -AA Dorsocentral setae in 2 pairs; anterior pair qw, l/2 length posterior pair. No prescutellar setae 0 - a w l-.. 4.,;N present. Two pairs scutellar setae, lengths

#14 equal to dorsocentrals. Two notopleural, 1 setae notopleurals 0 1 supra-alar present (length -dik. ;;. l/2 that of supra-alar). One katepistemal seta 23 present, longer than posterior dorsocentrals. Forefemora without ventral spines, middle of forefemur with 1 upright seta on lateral surface. Foretibia with subapical dorsal seta. Hind tibia with dorsoapical seta. Five sternites visible, anterior one (s I + II) in 3 sec- tions. Epandrium with pendulous, small ventral lobes and several long setae lateral to pp each. Surstyli with prensisetae apparent; prensisetae setiform, not peglike; in even row on medial surface. Wing hyaline. 24 MEASUREMENTS: TL = 1.46 mm; ThL = 0.66 mm; WL = 1.31 mm; CI = 1.6; 4-V = Figs. 22-24. Drosophila (H.) palheothoracis, 2.16. nhotographs. 22. Holotype; 23. piece showing two HOLOTYPE: Male; in amber collected with- )aratypes; 24. close-up of paratype in figure 23. in a 20 m radius of El Valle, Dominican Re- public (AMNH). The specimen is in near per- Miomyia io, new species fect condition except for some orbitals; all Figures 29-32 setae are intact and even the color seems well DIAGNOSIS: See generic diagnosis above. preserved. DESCRIPrION: Ground body colo)r brown; ETYMOLOGY: Latin for "oh!," in reference

Figs. 25-31. Neotanygastrella and Miomyia. Neotanygastrella wheeleri, paratype (25) and holotype (26, 27). 25. Basal portion ofwing; 26. posterolateral view ofdorsal portion of thorax; 27. head, oblique frontal view; 28. Neotanygastrella sp. (Trinidad). Abbreviations: same as other figures, also: poc, postocellar seta. 29-31. Miomyia io, holotype. 29. Habitus, lateral view; 30. terminalia, oblique posterior view; ce, cercus; ep, epandrium; vepl, ventral epandrial lobe; sry, surstylus (bears prensisetae); 31. frontal view of head. 1987 GRIMALDI: AMBER DROSOPHILIDS 13

.1 < / 25

< =;.10

31 14 AMERICAN MUSEUM NOVITATES NO. 2880

These traits may be a result of convergence. Until the female of Miomyia is found, its position in the scheme in figure 5 must re- main tentative.

NEOTANYGASTRELLA DUDA As mentioned previously, Wheeler (1963) described the two Neotanygastrella specimens from Chiapas amber but did not name them. My examination also shows their placement to be in this genus. Unlike the Do- Fig. 32. Miomyia io, photograph of holotype minican specimens, these are crudely pre- (scale line in mm). served (only fragments are not obscured by fractures and debris). Several characters not mentioned by Wheeler that allow a finer anal- to the first find of a Dominican amber-fossilized drosophilid. ysis of relationships are given below. COMMENTS: Affinities apparently lie with Neotanygastrella wheeleri, new species several genera of Drosophilinae sensu stric- Figures 25-27 to. A well-developed ventral epandrial lobe is characteristic of Chymomyza, some Mi- DiAGNOSIS: See abbreviated description crodrosophila, and African Neotanygastrella. below. The fossil, however, retains the plesiomor- DESCRIPrION: Eyes dorsoventrally elon- phic arrangement of the orbital seta (unlike gate. Face elongate, vibrissa situated far an- Chymomyza and Neotanygastrella). Unlike terior (above line extended across ventral Microdrosophila, it does not possess the fol- margins of eyes). Carina present, bulbous lowing: a fringe ofheavy bristles on the costal portion raised slightly above vibrissae. Ocel- segment of the wing between R2+3 and R4+, lar area raised. Facial colorations not evident. a high (- 5.0) 4-V index, an apically lanceo- Inner vertical setae cruciate. Median portion late wing, dense eye micropilosity, a high nar- of notum and scutellum, and at least the an- row carina, thick pretarsi, or a deep costal terior portions of the supra-alar regions and incision on the wing. Shared with Microdro- the postpronotal lobes, are dark brown. sophila are a state of the surstylus (d) with- Ground body color light yellow. Postocellar out peglike prensisetae (as cited originally by area brown. Two very long supra-alar setae Malloch, 1921, and by Okada, 1985-but I present, lengths about equal to lengths ofan- believe this trait to be symplesiomorphic at terior dorsocentral setae. Anal vein of wing least at the level of the family), one pair of absent; basal-radial cell elongate and narrow, vibrissae with tiny subvibrissae, a wide face length at least 8 times the width. rnd oral cavity, and small size. The stout, HOLOTYPE: UCMP 12700/B5103. Collect- ,clerotized labium is a trait shared with some ed from Chiapas, Mexico (see Wheeler [1963] members of Drosophila (Hirtodrosophila); for details ofthe locality). Only the head and otherwise, the closest relationship appears to thorax are present. be with Microdrosophila. Microdrosophila is PARATYPE: UCMP 12864/B5 104. This a genus of 48 species, with most of the de- specimen is more complete, but many parts scribed taxa from Southeast Asia. It is a prim- are obscured. itive drosophiline genus by virtue ofthe ple- ETYMOLOGY: Patronym, in honor of Mar- siomorphic oviscape (no peg ovisensilla) and shall R. Wheeler. surstylus (lack of peglike prensisetae). How- COMMENTS: The single vibrissa on each side ever, Microdrosophila also shares several de- of the face, anterior reclinate setae that are rived traits with the recently derived genus shorter than and anterolateral to the procli- Mycodrosophila: deep costal incision on the nates, and the presence of a bulbous facial wing (synapomorphy 12; fig. 5), rounded and carina make it evident that the specimens short notum, and a very narrow facial carina. belong in Neotanygastrella. In addition, the 1987 GRIMALDI: AMBER DROSOPHILIDS thoracic coloration suggests a relationship to strongly convergent, ends touch; bases di- N. tricoloripes Duda (known from Costa Rica rectly anterior to outer verticals. Outer ver- to Brazil). A prominently raised ocellar tri- ticals equal in length to inner verticals, di- angle is found in females of some Neotany- vergent. Vibrissae strongly curved, ends gastrella, but the occurrence ofthe trait is not almost touch; 1 pair present. accurately known. Also, long supra-alar setae Notum uniformly golden, with 8 rows ac- are unusual in drosophilids; the trait has not rostichal setulae. No prescutellar setae pres- been surveyed well in the genus. ent. Anterior dorsocentrals c. 0.5 x length of A hypotype is essentially a voucher spec- posterior dorsocentrals. Two humeral, I no- imen for which a description has been pro- topleural, 1 lateronotal, 2 supra-alar (1 very vided but no name given. Since, in the opin- short), and 2 katepisternal setae present. ion ofthe description's author, the specimen Anterior scutellar setae divergent; apical scu- may or may not represent a new species, tellars cruciate for c. 1/3 their length. Wing naming the specimen is deferred. A name has completely hyaline. Halter light. Legs yellow- been provided here for the two hypotypes orange, including forefemora, tibiae, and tar- since I consider the specimens distinct from si. Forefemur c. 2 x length of forecoxa, with extant neotropical species ofNeotanygastrel- 3 long, erect setae on lateral surface. Ventral la. Coloration and setal patterns are often spines on fore- or midfemora not present. sufficient for distinguishing species in the ge- Tergites mostly golden, anterior ones each nus (fig. 28). Four species ofNeotanygastrella with a pair ofdark brown, paramedian areas; occur in Central America; one, N. antillea the markings gradually coalesced posteriad. Wheeler, occurs on Jamaica. Male genitalia: Ventral surface of epandrium c. 2 x longer than dorsal surface. Me- PROTOCHYMOMYZA, NEW GENUS dial surface of epandrium with 8-10 long, of DIAGNOSIS: Head trapezoidal in dorso- thin, and straight setae; lateral surface ventral view; anterior reclinate setae slightly epandrium bare. Ventral epandrial lobe long anterolateral to proclinate orbitals; eyes bare; and narrow: length c. equal to length of cer- anterior to outer cus. Ventral epandrial lobe with several long, inner vertical seta directly thin setae. Surstylus attached to epandrium vertical; postocellars elongate, nearly equal dorsomedial to where ventral epandrial lobe in length to verticals; face flat; tergites each lo- with a pair ofbrown, paramedian areas; pre- attached to epandrium. Surstylus narrow, scutellar setae lost. Extinct (e unknown). bate, bearing 5 or 6 small, sclerotized pren- ETYMOLOGY: Feminine; Greek "fore" Chy- sisetae that point mediad. and MEASUREMENTS: PARATYPE: TL= 1.50 momyza, in reference to the fossilized mm; ThL = 0.70 mm; HW = 0.60 mm; WL= phylogenetic status. 1.45 mm. GENOTYPE: Protochymomyza miocena, new HOLOTYPE: Male (GOP); DOMINICAN species. REPUBLIC. Specific locality within the country not available, but from mines in vi- Protochymomyza miocena, new species cinity of Santiago. Figures 33-37 PARATYPE: Male; same locality informa- DIAGNOSIS: See generic diagnosis. tion (GOP). This specimen has three large DESCRIPTION: Head entirely golden yellow, mites attached to the ventral surface of the except for light brown intraocellar area. Fron- abdomen (fig. 34). tal vitta with granular surface. Eyes glabrous, ETYMOLOGY: Feminine (Latin) derivative rosy-pink. Face flat, yellow. Pedicel, flagello- ofMiocene, the geological age ofthe amber's mere I yellow; arista with 4:2 dorsal/ventral origin. branches. Postocellar setae long, lengths c. COMMENTS: A new genus is erected for this equal to orbitals; cruciate for about 1/5 their fossil since the specimens apparently belong length; bases near outer corners ofocellar tri- to the sister group of Chymomyza (fig. 5). angle. Ocellar setae straight, extended to ped- The head shape, orbital setation, long coxae, icels. Anterior reclinate lateral and slightly small surstyli with stout prensisetae, plus anterior to proclinate. Inner verticals long, elongate and setose ventral epandrial lobes 16 AMERICAN MUSEUM NOVITATES NO. 2880

--:v

! av 1tIl35mb

Figs. 33-36. Protochymomyza miocena. 33. Dorsal habitus of holotype; 34-36. paratype terminalia, posterior and lateral views (showing mites on abdomen in fig. 34).

are derived features shared with Chymomy- ~~34 za. The flattened cerci are autapomorphic for this group offlies. The position ofthe anterior reclinate orbital, however, is not as far anterior to the proclinate as is found in Chy- momyza. Plesiomorphic, at least at the family level, are the unadorned femora, hyaline wings, six rows of acrostichals, and the abdominal coloration. SCAPTOMYZA HARDY Scaptomyza dominicana, new species Figures 38-41 Fig. 37. Protochymomyza miocena. Photo- DIAGNOSIS: Head quadrate, fiat; anterior graph of holotype. reclinate setae tiny; face nearly flat; 1 pair of 1 987 GRIMALDI: AMBER DROSOPHILIDS 17

39 40

Figs. 38-40. Scaptomyza dominicana, holotype. 38. Habitus, dorsolateral view; 39. head, dorso- frontal view showing proboscis with clypeus and long, narrow palps; 40. wing, portion. vibrissae present; acrostichals in 4 rows; no- Carina very weak; face almost flat. One pair tum with 4 grey vittae (male unknown). of vibrissae present; subvibrissae obscured. DESCRiwrION: This specimen is obscured Flagellomere I c. 112 the length of pedicel. in certain portions by fractures. The only Arista with 3 dorsal and 2 ventral branches, views where features are distinctly visible are 5-6 minute medial ones. Palpi elongate: with from fully dorsal and anterodorsal angles. 1 long terminal and several subapical setae. Most ofthe wing surface is closely adpressed Eyes glabrous; brick red. to the abdomen, and portions are also frac- tured, so some veins were not easily dis- cerned. Head width slightly greater than the width of the notum; head dorsoventrally compressed. Inner vertical setae longest ones on head; anteromedial to outer verticals. Outer verticals divergent, slightly shorter in length than inner verticals (inner vertical positions distorted). Orbital setae golden, not black. Anterior reclinates tiny, lateral, and slightly anterior to proclinates. Posterior reclinates about same length as inner verticals. Distance between proclinate and anterior reclinate is 1/3 distance between proclinate and inner ver- Fig. 41. Scaptomyza dominicana, photograph tical. Frontal region and face golden orange. of holotype, dorsal view. 18 AMERICAN MUSEUM NOVITATES NO. 2880

- /I -7T -17 77_..u '~~~~~~~ 4-"

43 Figs. 42-48. Drosophilinae sp. A. 42, 43. Hypotype, dorsolateral and ventral views; 44, 45. oblique posterior view of terminalia; 46, 47. detail of head, lateral and oblique ventral views; 48. wing.

Thorax ground color golden brown, with 4 ofposterior pair. One lateral prescutellar/side, silvery-grey vittae. The 2 light paramedian length c. 1/2 that of anterior dorsocentral. Ac- vittae diminished and ended at c. 3/4 the notal rostichals in 4 rows between anterior dor- length. Vitta pair lateral to these diminished socentrals, in 2 rows between posterior dor- just anterior to paramedians. Two long dorsocentrals, 6 rows anterior to dorsocentrals. socentrals present; anterior pair c. 1/2 length Pleural regions obscured. Scutellum unicol- 1 987 GRIMALDI: AMBER DROSOPHILIDS 19

48 Fig. 42-48. Continued. orous brown: apical scutellar setae conver- anterior reclinate is directly lateral or slightly gent, anterior scutellars c. 1/2 length ofapicals. anterolateral to the proclinate, is shared by Two subequal humeral, 2 large katepistemal, Chymomyza and Protochymomyza. A flat, 1 notopleural setae are present. Wings hya- widened head, like that found in the fossil, line. Legs elongate, unicolorous, orange- occurs in four species from indo-Pacific is- brown. Forecoxae relatively short: length lA-½/2 lands which belong to the subgenus Parascap- that of forefemur. tomyza. However, because there are two sub- Abdomen with black-brown terga, except equal humeral setae, the fossil species would for medial portions oft II, III, which are light otherwise belong to subgenus Scaptomyza, yellow and gradually faded into dark ground according to Hackman's (1959) definitions. color. Oviscape extruded, but details of ovi- Terminalia of female and especially of male sensilla obscured. Oviscape triangular in lat- Scaptomyza are replete with characters; eral view, apically pointed. without access to these characters, placement MEASUREMENTS: TL = 1.40 mm; ThL = of the fossil in Scaptomyza would be ap- 0.67 mm; HW = 0.53 mm; WL = 1.60 mm; proximate at best. Scaptomyza is a very large, CI = 1.2; 4-V = 2.2. cosmopolitan genus (226 species); the flies HOLOTYPE: Female (GOP); from mines in are very abundant in most parts ofthe world the vicinity of Santiago, Dominican Repub- except Australia. Eleven described species of lic. Scaptomyza occur in Central America, but ETYMOLOGY: From Latin, "Dominican," only three are on the Greater Antilles: S. pertaining to origin of the amber. (Mesoscaptomyza) conquilletti (circumcar- COMMENTS: The four rows of acrostichals, ibbean), S. (M.) paravittata (circumcarib- the grey-brown notal vittae, and long legs bean and southwestern U.S.), and S. (M.) vit- make it certain that this is a Scaptomyza tata (circumcaribbean). The fossil possesses species. However, the eyes are bare; in extant none of the diagnostic features of Mesoscap- species, the eyes have short, dense, interfa- tomyza, such as upturned apical scutellars, cetal setulae. The orbital setation, where the two rows ofacrostichals, and presutural dor- 20 AMERICAN MUSEUM NOVITATES NO. 2880

legs. Notum short, very curved, c. 2 times length of scutellum. Two pairs of dorsocentral setae present, posterior pair longer. Ac- rostichals in 6 even rows; no prescutellar setae. Two pairs of scutellars; apicals longer, perhaps cruciate. Pleural setae stout, black: 2 humerals, 1 notopleural, 2 anepimeral, 2 katepisternals (a long one sticks out from body, shorter one directed upward). Wings of holotype cut at proximal third. Halteres dark brown. Forefemur ofparatype with ventral row of 5 or 6 stiff, short setae. Mesotho- racic tibia with stout ventroapical bristle (length c. 2 times width of tibia). Metatho- racic tibia with dorsopreapical and erect seta, length c. equal to tibia width. Hind tibia and tarsomere 1 broadened slightly in full lateral view. Abdomen stout; width slightly greater than thorax width, length about equal to thorax length. Sternites heavily sclerotized, about as much as tergites. Sternite V divided slightly Fig. 49. Photograph of Drosophilinae sp. A, by median area of weak sclerotization. Ven- female. tral portion oft VIII connected by light sclerotization. Oviscape everted in both specimens; heavily sclerotized; with small peg socentrals. This fossil evidently represents ovisensilla and ventroapical setal sensilla. extinction of a primitive Scaptomyza lineage Tergite IX is a narrow saddle partially hidden which existed at least on Hispaniola. dorsally by t VIII. Cercus short, conical. MEAsuREMENTs: TL = 1.95 mm; ThL= DROSOPHILINAE INCERTAE SEDIS 0.82 mm; HW = 0.92 mm. Species A HYPOTYPE: Female (GOP); DOMINICAN Figures 42-49 REPUBLIC, specific locality not known. OTHER: Female (NMNH 8512); female DIAGNOSIS: Eyes densely micropubescent; (GOP) (fig. 49), specific localities in Domin- 1 pair ofvibrissae; carina low, rounded; body ican Republic also not given. completely black-brown, including halteres COMMENTS: Certainly a drosophiline; this and legs (male unknown). 4 species cannot be placed into any monophy- DESCRIPrION: Eyes with short, dense setu- letic group of the subfamily as the present lae; silvery, not red. Arista with 5-6 dorsal taxonomy allows. and 2-3 ventral branches, evenly spaced; 5 short, fine medial branches. One of vi- pair Species B brissae; subvibrissae much thinner and short- Figures 50-56 er. Width of oral cavity c. 1.7 times width of labellum. Labellum with 7 or 8 pseudotra- DIAGNOSIS: Eyes densely micropubescent; cheal lobes. Orbital setae stout, strongly anterior reclinate just lateral to proclinate; curved. Anterior reclinate short, c. 1/3 length face flat; vibrissae in 1 pair; femora and tibiae of posterior reclinate, much closer to procli- with light brown band on each; ovisensilla nate than to posterior reclinate. Proclinate pegs present (d unknown). and posterior reclinate about equal in length. DEscRIrrION: Coloration of head and tho- Mentum sclerotized. Carina low; surface not rax difficult to discern since a silvery and me- flattened but slightly rounded. tallic sheen covers the pattern. Postocellars Body completely black-brown, including convergent, lengths c. /4 that ofocellars. Inner 1987 GRIMALDI: AMBER DROSOPHILIDS 21

53 5, 54 55

Figs. 50-56. Drosophilinae sp. B, hypotype. 50. Head, oblique ventral view; 51. head and thorax, dorsolateral view; 52. wing; 53-55. pro-, meso-, and metathoracic legs, respectively; 56. abdomen, lateral view, showing color pattern and cercus plus oviscape. verticals convergent, outers divergent, lengths to inner vertical. Anterior reclinate orbital of each approximately equal and c. 2 times base laterally adjacent to proclinate. Anterior longer than postocellars. Posterior reclinate reclinate tiny: length c. 1/3 that ofother orbital orbital slightly closer to proclinate orbital than setae. Some interorbital setulae apparent. Eyes 22 AMERICAN MUSEUM NOVITATES NO. 2880 with dense interfacetal setulae. Arista with 5 primitive at the level of the genus: no facial dorsal, 2 ventral, and 4-5 minute medial carina, a row ofsetulae on the anepisternum, branches. Face flat. One pair of strong vi- and posterior reclinate setae closer to the brissae subtended by numerous shorter and proclinates than to the inner verticals. thinner setae. Palpus with 1 long, subapical seta and numerous smaller ones. Clypeus broad: width equal to width of oral cavity. KEY TO AMBER-FOSSILIZED Each labellar lobe with c. 6 pseudotracheal DROSOPHILIDAE rows. 1. Prescutellar setae present; arista pubescent; Thorax with 6 rows acrostichal setulae (be- Baltic amber .... . Electrophortica succini tween anterior dorsocentral setae). Anterior Prescutellars lost; arista plumose; Chiapas and dorsocentrals c. 1/2 length of posterior dor- Dominican amber ...... 2 socentrals. One small lateral prescutellar seta 2. Facial carina bulbous; head high; ocellar area on each side. Pair of anterior scutellar setae raised; wing with anal vein absent; Chiapas based on side ofscutellum. Pair ofapical scu- amber . .. Neotanygastrella wheeleri, n. sp. tellars; length c. twice that of anterior scu- If present, carina narrow or flat; head and tellars; cruciate. Three supra-alar setae per ocellar areas otherwise; anal vein well de- side: anterior one short (about 1/2 length of veloped; Dominican Republic amber .. 3 others). One large and 1 small humeral, 1 3. Interfacetal setulae sparse/absent (eye bare) notopleural, 1 lateral presutural seta present. [best seen with reflected light] ...... 4 Anepistemum with transverse row of5 short, Eyes densely micropubescent ...... 7 fine setulae. Katepisternum with 1 large seta 4. Head trapezoidal in dorsoventral view, flat ...... 5 projected posteriad, a smaller one projected Head spherical, not distinctly flattened; body anteriad, and 4 fine setulae. Femora and tibia entirely brown; oral cavity and face wide; with a light brown band on each. Wing hya- 1 katepisternal seta .... Miomyia, n. gen. line. Abdominal terga with dark medial bands 5. Postocellars elongate, equal in length to ver- that are gradually narrowed laterad. Lateral ticals; each tergite with pair ofbrown, para- portions terga III-IV with dark spot. Tergite median spots; 8 rows acrostichals; notum VIII with posterior fringe of c. 7 setae (each unicolorous .... Protochymomyza, n. gen. side). Oviscape apparently sclerotized, tri- Postocellars much shorter than verticals; ter- angular in lateral view, with 3-4 terminal gites mostly black-brown, with no distinct ovisensilla and about 10-11 smaller ones on markings; 4 rows acrostichals ...... 6 ventral side. 6. Forefemur and tibia dark, rest of leg light; notum unicolorous; anterior reclinate an- MEASUREMENTS: ThL 0.85 mm; WL terior to reclinate by distance .3 x width 1.70 mm; CI = 2.5; 4-V - 2.37. of setae ... Chymomyza primaeva, n. sp. HYPoTYPE: Female (GOP); in amber piece Foreleg unicolorous; notum with 4 grey vittae; collected from vicinity surrounding Santiago, anterior reclinate just lateral to or anterior Dominican Republic. to proclinate by 1 seta width ......

COMMENTS: The dense eye setulae and ovi- ...... Scaptomyza dominicana, n. sp. sensilla pegs make it appear that this species 7. Face flat; without a carina ...... is a Drosophila sensu stricto as previously ...... Drosophilinaesp. B defined. Coloration of the legs and lateral Face carinate ...... 8 portions of the abdomen, the apically nar- 8. Vibrissae present in 2 pairs; vertex slightly rowed oviscape, and anterior reclinate setae raised .... Drosophila (D.?) succini, n. sp. that are laterally adjacent to the proclinates, Vibrissae in 1 pair; vertex not noticeably raised in fact, suggest an affinity with the Drosophila .9...... 9 9. Body completely dark brown, even halteres repleta species-group. However, if the setal and legs ...... Drosophilinae sp. A spots, which are present on the notum ofmost Body mostly yellow, including halteres and species in the group, occur in the fossil, they legs ...... 10 are hidden by a golden sheen covering the 10. Carina broad, flat; sternites large, overlapping integument. At best the fossil is basal to the ...... Drosophila(D.?) poinari, n. sp. repleta species-group based on the few above Carina low, rounded; stemites separate synapomorphies; in fact, it has many features ...... Drosophila(H.) paleothoracis, n. sp. 1987 GRIMALDI: AMBER DROSOPHILIDS 23

LITERATURE CITED Loew, H. 1850. Uber den Bernstein und die Bernstein Bachofen-Echt, A. Fauna. Programm Realshule Meseritz, 1949. Der Bernstein und Seine Einschlusse. 44 pp. Wien: Springer-Verlag. Malloch, J. R. Baroni-Urbani, C., and J. B. Saunders 1921. Some notes on Drosophilidae (Diptera). 1982. The fauna of the Dominican Republic Entomol. News, 32:311-312. amber: the present status ofknowledge. (San- McAlpine, J. F., and J. E. H. Martin Trans. 9th Caribbean Geol. Conf. 1966. Systematics of Sciadoceridae and rela- to Domingo): 213-223. tives with descriptions of two new gen- Brouwer, S. B., and P. A. Brouwer era and species from Canadian amber 1982. Geologia de la region ambarifera ori- and erection of family Ironomyiidae ental de la Republica Dominicana. (Diptera: Phoroidea). Canadian Ento- Trans. 9th Caribbean Geol. Conf. (San- mol., 98:527-544. to Domingo): 303-322. Okada, T. Burleigh, R., and P. Whalley. 1985. A revision ofthe genus Microdrosophila 1983. On the relative geological ages ofamber with descriptions of ten new species and copal. J. Nat. Hist., 17:919-921. (Diptera; Drosophilidae). Int. J. Ento- Cockerell, T. D. A. mol., 27:310-326. 1923. Insects in amber from South America. Patterson, C., and D. E. Rosen Am. J. Sci., 5:331-333. 1977. Review of the Ichthyodectiform and Gagne, R. J. other Mesozoic teleost fishes and the 1977. Cecidomyiidae (Diptera) from Cana- theory and practice ofclassifying fossils. dian amber. Proc. Entomol. Soc. Wash., Bull. Am. Mus. Nat. Hist., 158:83-172. 79:57-62. Petrunkevitch, A., et al. Grimaldi, D. 1971. Studies of fossiliferous amber arthro- 1986. The Chymomyza aldrichii species-group pods of Chiapas, Mexico. Part II. U. (Diptera: Drosophilidae): relationships, Cal. Publ. Entomol., 63:1-106. new neotropical species, and the evo- New York Poinar, G. O., Jr. lution ofsome sexual traits. J. 1984. First fossil record of parasitism by in- Entomol. Soc., 94:342-371. sect parasitic Tylenchida (Allantone- 1987. Relicts in the Drosophilidae. In J. K. Nematoda). J. Parasitol., 70: Liebherr (ed.), Historical biogeography matidae: fauna, (in prep.). 306-308. ofthe Caribbean insect Schlee, D., and W. Gl6ckner Hackman, W. Stuttg. Beitr. Naturkd., ser. 1959. On the genus Scaptomyza Hardy (Dipt., 1978. Bernstein. Acta Zool. Fenn., 97:1- C, 8:1-72. Drosophilidae). Throckmorton, L. H. 71. 1975. The phylogeny, ecology, and geography Hennig, W. of Drosophila. In R. C. King (ed.), 1965. Die Acalyptratae Baltischen Bernsteins. Handbook of genetics, vol. 3, pp. 421- Stuttg. Beitr. Naturkd., 145:1-213. 469. New York: Plenum. Krishna, K., and A. E. Emerson 1983. A new fossil species of termite from Wheeler, M. R. Mexican amber, Mastotermes electro- 1963. A note on some fossil Drosophilidae mexicus (Isoptera, Mastotermitidae). (Diptera) from the amber of Chiapas, Am. Mus. Novitates, 2676:1-8. Mexico. J. Paleontol., 37:123-124. Lambert, J. B., J. S. Frye, and G. 0. Poinar, Jr. Wilson, E. 0. Republic: 1985. Invasion and extinction in the West In- 1985. Amber from the Dominican the Do- an analysis by nuclear magnetic reso- dian ant fauna: evidence from nance spectroscopy. Archaeometry, 27: minican amber. Science, 229:265-267. 43-51. Wilson, E. O., F. M. Carpenter, and W. L. Brown, Langenheim, J. H. Jr. 1969. Amber: a botanical inquiry. Science, 1967. The first Mesozoic ants. Science, 157: 163:1157-1169. 1038-1040. Langenheim, J. H., and C. W. Beckh Woodley, N. E. 1968. Catalogue of infrared spectra of fossil 1986. Parhadrestiinae, a new subfamily for resins (ambers). I. North and South Parhadrestia James and Cretaceogaster America. Bot. Mus. Leafl. Harv. Univ., Teskey (Diptera: Stratiomyidae). Syst. 22:65-120. Entomol., 11:377-387. Recent issues of the Novitates may be purchased from the Museum. Lists of back issues of the Novitates, Bulletin, and Anthropological Papers published during the last five years are available free of charge. Address orders to: American Museum of Natural History Library, Department D, Central Park West at 79th St., New York, New York 10024. i