sign in sign up
<<
Home , Avisauridae, Borogovia, Elopteryx, Heptasteornis, Paronychodon, Tarascosaurus

ORYCTOS,Vol. I : 87-104,Octobre 1998

SMALLTHEROPODS FROM THE LATE OFTHE HATEG BASIN OryESTERNROMANIA) . AN UNEXPECTED DIVERSITY AT THE TOP OF THE FOOD CHAIN

Zoltan CSIKI & Dan GRIGORESCU Bd. N. BalcescuNr. 1, 70111 Bucharest,Romania ; Laboratoryof , Faculty of Geology & Geophysics,University of Bucharest.

Abstract : The LateMaastrichtian deposits of the HategBasin have yielded numerousremains of herbivorousdinosaurs (tita- nosaurids,ornithopods, nodosaurids); but thoseof theropodsare scarce, represented mostly by teethand hindlimb elements. New materialallows the recognitionof an unexpecteddiversity of thesepredators in the Hategfauna. Teeth of variousmor- phologiesare reported here to representseveral distinct taxa of small theropods:a velociraptorinedromaeosaurid, a "troo- dontidlike" smalltheropod, cf . Euronychodonand perhaps a fourth,peculiar small theropod with sharp,but unserratedcari- naeon the teeth.Re-examination of previouslypublished theropod material also suggestssuch diversity. Femora, previous- ly refenedto ,probably belong to a derivedmaniraptoran. A distalend of a femur seemsto documenta smallcera- tosaur,while sometibiotarsi ( of BradycnemeandHeptasteornis) may representa non-maniraptorantetanuran the- ropod.Without diagnosticremains of smalltheropods, it is inappropriateto give the reportedmaterial generic names; conse- quentlyinformal useof the publishednames 'Elopteryx' and'Bradycneme' is recommended.Moreover, for mostpart of the isolatedtheropod remains from Hategthere are no reasonsto groupthem under the samename; one suchcase may be repre- sented,however, by someskull elementsand the velociraptorineteeth. The diversityof the small theropodsin the Hategfauna, together with the absenceof a large "top" theropod,represents the first suchcase reported for Late Cretaceousfaunas. This phenomenonis probablylinked to the restricted,insular habitat of the Hategfauna, which could not accommodateand supportany largersized predator.

Key words : Hateg Basin, I'ate , ceratosaurs,maniraptorans, dromaeosaurids,Euronychodon, diver- siry.insular habitat.

Les petits théropodesdu Crétacéterminal du Bassindu Hateg (Ouestde la Roumanie): une diversité inattendueau sommetde la pyramide trophique

Résumé : Les dépôtsd'âge Maastrichtien terminal du Bassinde Hateg ont fourni de nombreuxrestes de dinosauresherbi- vores(titanosauridés, ornithopodes, nodosauridés), mais ceux de théropodessont rares, représentés par desdents et desfrag- mentsde membrespostérieurs. Le nouveau matériela permisde mettreen évidenceune diversitéinattendue de cespréda- teurs dansla fauned'Hateg. Des dentsde morphologiesvariées attestent la présencede différents'l4ronsde petits théro- podes: un dromaeosauridé,un petit théropodeproche des troodontidés, cf. Euronychodonet peut-êtreun quatrièmepetit théropodecurieux avecdes dents à carènessaillantes mais sanscrénelures. Un nouvelexamen des restes des petits théro- podesde Hateg déjà publiésconfirme cette diversité. Les fémurs,rapportés à Elopteryx, appartiennentprobablement à un maniraptorien;I'extrémité distale de fémur sembledocumenter un petit cératosaure,tandis que des tibiotarses (holotypes de Bradycnemeet Heptasteornls)pounaient représenterun théropodetetanurane primitif. En l'absencede restesdiagnosiques des petits théropodes, il sembleprématuré de donnerdes noms génériques; nous recom- mandonspar conséquentI'utilisation informelle des noms publiés 'Elopteryx'et'Bradycneme'.De plus, il n'y a paslieu de regrouperdes restes isolés sous le mêmenom générique;le seulcas où I'on peut sepermettre de le faire concernequelques restesde crânespubliés auparavant et les dentsde dromaeosauridés. La diversitédes petits théropodesdans la fauned'Hateg, ainsi que l'absencede grandsthéropodes, est un casunique dans les faunesdu Crétacéterminal. Ce phénomèneest probablementlié à un habitatinsulaire qui n'a paspermis la colonisation par un grand prédateur.

Mots clés : Bassin du Hateg, Maastrichtien terminal, , , , Euronychodon,div ersité. habitat insulaire.

87 ORYCTOS,Vol. 1, 1998

INTRODUCTION PREVIOUS REPORTS ON THEROPOD The vertebrate fauna of the Hateg Basin (Late FROM TRANSYLVANIA Cretaceous : Latest Maastrichtian; Antonescu et al 1983) was discoveredin 1895 and first describediÏ l- hungaricus - a problematic and 1900 by Nopcsa. The fauna comes mostly from the misinterpreted theropod from Tbansylvania fluviatile, continental deposits of the Densus-Ciula Nopcsa (1901, cited by Le Læuff &, Buffetaut, and Sinpetru formations, cropping out in the north- 1991) erectedMegalosaurus hungaricu,son the basis western and central parts of the basin respectively of two (cf. Nopcsa, I9I5; contra Le Loeuff & (Grigorescu, 1992 : fig. 1). Isolated or associated Buffetaut, I99I) small, isolated teeth (MAFI Ob. remains of herbivorous dinosaurs (the titanosaurid 3106, not found in 1982; Weishampel, pers. comm.). sauropod dacus, the ornithopod of Megalosaurus hungaricus is currently regarded by uncertain affinities Rhabdodon robustus, the basal most authors (Le Læuff 8. Buffetaut, I99I; Le hadrosaurid Telmatosaurus transsylvanicus and the Loeuff, 1992) as an indeterminate theropod of (?) nodosaurid Struthiosaurus transilvanicus) are wides- Late Maastrichtian age. However, as pread, along with those of turtles and crocodilians. In Nopcsa ( 1915) noted, the teeth were found at contrast, the remains of the top predators (theropod Nagybâr1d (today Borod, Bihor County, Romania); dinosaurs) are very scarce, despite the intensive col- according to the label once accompanying the teeth, lecting effort conducted over two periods: the first they came from the "Gosau coal outcrops near two decades of the century by F. Nopcsa and in the Nagybfu6d locality". last two decades by teams from the Faculty of In the Borod basin, the Upper Cretaceous out- Geology and Geophysicsof the BucharestUniversity cropping rocks are represented by mostly detritic (FGGUB) led by D. Grigorescu and those of the deposits: sandstones and marls with interbedded Muzeul Civili zatiei Dacice si Romane, Deva microconglomerates and charophyte-bearing lacus- (MCDRD; formerly Deva County Museum); in the trine limestones, followed by a sandstone conglo- last four fieldwork was done jointly with D. merate unit with rudist bivalve-bearittg limestones Weishampel from the Johns Hopkins University in and rudist reefs, a thin, discontinuoussandstone-silts- Baltimore, USA. tone unit with tuffits and rhyolitic tuffs; the Upper Cretaceous sequence ends with a marly unit with Inoceramlts (Mutihac & Ionesi, I97 4). In the eastern part of the outcropping area, the lower paft of these deposits laterally grade into a typical coal-bearing "Gosaufacies"(Ianovici et al., 1976;Mutihac, 1990). The whole sequencecan be characterized as a trans- gressive unit overlaying older deposits and the crys- talline basement of the Plopis Mts. (to the north); the deposition reflects changing paleoenvironmental conditions from nearshore continental-lacustrine and litoral environments to inner shelf ones. The deposits arehighly fossiliferrous. The basal,terrigenous depo- sits contain a fauna with Corbula striatula, Cardium subdinense, Turritella dupiniana, Melanopsis dubia and are considered of a Coniacian-early Santonian age (Ianovici et al., 1976). The overlying calcareous deposits yielded a rich association dominated by FIGURE 1 - The geographicalsetting of the Hateg Basin and suffounding regions. Hippurites praecursor, Vaccinites gosaviensis, U a - outcrop areaof the Sinpetru Formation; oppeli santonianus in the lower part (indicating alate b - outcrop areaof the Densus-CiulaFormation Santonian age) and Vaccinites oppeli, V archiaci, V

88 CSIKI & GRIGORESCU- ROMANIAN THEROPODS inequicostatus in the upper part (indicating an early part of a left tibiotarsus (BMNH A.4358, formerly Campanian age; Ianovici et al., I9J6). under the same specimen number as the femur); The Borod Senonian sequenceis diachronous in referred material includes another proximal femur its terminal part. In the eastern part of the outcrop (BMNH A.1235) and two more distal tibiotarsi area, the uppermost limestone deposits yielded (BMNH A.1588, A.1528). All the referred material Lepidorbitoides mamillata, L. minor Orbitoides gen- comes from the Sibisel Valley (Weishampel, pers. sanicus, an assemblage of late Campanian comm.), but no further details are known about their Maastrichtian age (op.cit.), while in the northern part precise locality or possible co-occurrence. the overlying marls yielded clams (Inoceramus balti- Subsequently,Harrison and Walker (1975) sepa- cus) and a micropaleontological association with rated the tibiotarsi and erected for them the new stri- Globotruncana lapparenti lapparenti, G. lapparenti giform genera Bradycneme draculae (A. 1588) bulloides, G. elevata elevata, G. bulloides, indicating and andrewsi (A. 1528, A.4358). The an early Campanian age (Th. Neagu, pers. comm.). avian nature of all these remains were questioned by As M. hungaricus came from the "Gosau coal" numerous authors, who noted their reptilian (and (i.e.the lower, terrigenous part of the Senonian more exactly dinosaurian) affinity (ex. Elzanowski, sequence), although it may represent a (dwarf , cf . 1983; Grigorescu, I984a as "coelurosaurians"). Nopcsa, I9l5) theropod, it is more probably member Further suggestionswere made on their relationships of a stratigraphically older and also ecologically dis- by Noffnan (1985) and Osmolska(1987; Bradycneme tinct - yet still unknown - assemblage,different from and Heptasteornis were listed as possibly - that from the Hateg Basin. tids), Paul (1988; as the possible troodontids Bradycneme draculae and Troodon? andrewsi), 2- Theropod dinosaurs reported from the Hateg Osmolska &. Barsbold (1990; all three taxa as basin nomina dubia), Le Læuff (1992) and The first top predator identified from the Hateg Le Læuff and Buffetaut (I99I; as representing the Basin was called "Megalosaurus sp." by Nopcsa elopterygine dromaeosaurid Elopteryx nopcsai with (1915) who considered it as an indeterminate carno- Heptasteornis and Bradycneme as junior synonyms) saur, possibly a close relative of M. pannoniensis or and Howse and Milner (1993, âs the troodontid M. hungaricus from the Late Cretaceous of . Heptasteornis). No appropriate arguments were pre- Both of these later taxa, based solely on isolated sented to support these assertions,with the exception teeth, are listed as nomina dubia by of Le Læuff's papers. Molnar et al. (1990). Similarly early discovered small theropod Nopcsa's "carnosaur" material is very poor, remains come from the Densus-Ciula Formation at consisting of two caudal vertebral centra (Nopcsa, Valioara, from where two possible theropod teeth 1915: pl. III, figs. 6,7). These, which came from were mixed with and considered to be crocodilian Kadic's excavations at Valioara (Densus-Ciula teeth (in the depository of the Hungarian Geological Formation) , ffie very similar to titanosaurid mid-pos- Survey, catalogued as MAFI V. 12685 partim.). The terior caudal vertebrae and are here tentatively refer- nature of these teeth were recognized in 1994, during red to Magyarosaurus dacus, the only titanosaurid a short visit to the MAFI collection; their laterally taxon known from Hateg (Le Læuff, 1993). compressed,pointed, distally recurved shapeand ser- Paradoxically, misidentified theropod remains rated carinae suggest they may indeed belong to were unearthed and described before those of small maniraptoran theropods. Also from Valioara, " Megalosaurus sp.". Fragmentary limb bones from Kadic collected two paired frontals that Nopcsa the Sinpetru Formation deposits cropping out in the misidentified as Telmatosaurus (MAFI v. 1 3528); Sibisel Valley were describedby Andrews (1913) as recently, Jianu and Weishampel (1997) suggestedthat the pelecaniform bird Elopteryx nopcsai (the material they could in fact belong to an arctometatarsalianthe- is currently in the British Museum of Natural History, ropod of uncertain affinities. BMNH). The material is representedby a proxi- No new material assignedto small theropods was mal femoralftagment (BMNH A. 1234) and the distal found until the late seventies; at that time, several 89 ORYCTOS,Vol. 1, 1998 teeth of "coelurosaurian" dinosaurs were reported general shapes;cross-sections; positions of the ante- from different localities within the Sinpetru rior and posterior carinae; morphology of the den- Formation (Grigorescu, I984a,b; Grigorescu et al., ticles; maximum number of denticles per 1 mm; fore- 1985). Finally, the Sibisel Valley yielded in 1992 a aft basal lengths (FABL, with which all the other small, partial skull roof (Weishampel & Jianu, 1996) measurements could be compared; it is proportional representing a dromaeosaurid apparently with the total length), basal widths (BV/); lateral closely related to langstoni from compression indices (LCI = FABLÆV/), denticle the Late Cretaceous of North size difference indices (DSDI = ratio between the America. The same authors report on further cranial number of denticles on the anterior and distal carina and teeth material from the same locality to be publi- on lmm). (A11specimen numbers, where not speci- shed in the future. fied, are FGGUB). The excavations carried out after 1978 at several sites from both the Densus-Ciula and the L- Velociraptorine dromaeosaurids Sinpetru Formations led to the recovery of a large (figs. 3a, b, c, d; ab) number of small theropod remains (mostly teeth) that Material: R. 127I, R .I32I , R. 1322, R. 1428, unravel an unexpected diversity of these dinosaurs in R. 1430,, R. 1580, R. 1 582 (for measurementsand other the Hateg paleocommunity. details see table 1). The tooth crown is laterally compressed (LCI < 0.55), pointed and sharply recurved. Both the DESCRIPTION AND COMMENTS anterior and posterior edges are strongly curved so that the tip extends behind the base of the crown, Severalsmall theropodteeth were recoveredbet- giving it a strong backward arch. The labial and lin- ween 1979and 1996from different localities spread gual sides are convex; the degree of the lingual all over the outcroppingsurface of the Densus-Ciula convexity varies from slight (giving an asymmetrical (Valioara,Tustea) and Sinpetru(Sibisel Valley, Pui) cross-section;R. 132I, R. 1430, R. 1580) to highly formations(fig. 2).Thesewere found mostlyby wet- rounded (symmetricalcross-section, R. 1271,R. 1322, screeningtechniques and came from either channel R. 1 428, R. 1582). The anterior carina is serrated only lag-type,poly-taxic bonebed-typeor microvertebrate in its distal half (at most); below this serratedportion accumulation-typetaphocoenoses (Csiki , 1995).The it continues as a low, rounded although visible ridge. isolated,rootless crowns ate consideredto be shed The posterior carina is serrated from the tip to the teeth. Their state of preservationvaries from well- base. As a general rule, the posterior denticles ate preserved,complete, unworn crownsto enamel-less, wider at their base than the anterior ones (DSDI worn fragments,but their integratedstudy permits the 1.3); moreover, they are twice as tall. In the crowns recognitionof four major types of teeth on the basis with asymmetrical cross-sections (identified here as of their (Currie et al., 1990;Rauhut & Werner,1995) premaxillary, anteriormost maxillary or dentary teeth; Ostrom, 1969, Sues, 1977, Currie et àI., 1990) RUSCA IUO}--TA.)\A both the anterior and posterior denticles (5.5-6 den- BASI-'\i rL4.Tf G BASL\ ticles per mm and 4-6 denticles per Trrnl,respective- ly) are relatively larger than in the more posterior teeth (9.5-10.5 denticlesper mm andT denticlesper IrIm, respectively); the denticles also decreasein size at the ends of the serrated portions. The denticles (where well-preserved, unworn) ate straight, taller (Sibucl!'rlle') o r0 r! (. I - Sinpctru I - PuifBrrbrt Riycr 1'rllc1') than wide at the base, and pe{pendicular to the tooth J - \'rliorrr .l - Tustce6ggr Sitc) axis (sometimes slightly oblique near the tip, also noted in , Sues, L977, but considered FIGURE 2 - Fossiliferrouslocalities in the Hateg Basin subject of individual and/or ontogenetic variation in that yielded small theropod remains , Ostrom, 1969).

90 CSIKI & GRIGORESCU- ROMANIAN THEROPODS

Thxon Spec.nr. Localitvdata Length(L) AABLB\ryLruBLLCI ant.serrat. post.serrat. DSDI permm permm

velociraptorine FGGUBR.1428 Sinpetru,19957*(9.5) 4,6 2.1 2.06 0,46 10.5 1 1.5 "La Carare"

velociraptorine FGGUBR,IT]I Sinpetru,1994 1.3 3.2 9

velociraptorine FGGUBR.1321 Sinpetru,1983 11 5 2,5 2.2 0.5 5.5 4 1.39

velociraptorine FGGUBR,T322 Sinpetru,1983 6* 3.5 I,9 r.tI 0.54 9.5 l r.36

velociraptorine FGGUBR.1430 Sinpetru,19954.r* (6.5) 3.3 I,6 r.9l 0.48 9 "LaCarare"

velociraptorine FGGUBR.1580 Tustea,1994 5.2* 3.5 6 6 t?

velociraptorine FGGUBR.1582 Valioara,1995 1.9 0.9 0,4 2 0,44

troodontid-like FGGUBR.1318 Sinpetru,1982 12,5 8 5.3 1.56 0.66 5 5 1

troodontid-like(?) FGGUBR.1319 Sinpetru,1981 11 I 5 1.38 0.63

troodontid-like FGGUBR.1320 Sinpetru,1992 5.3 2,9 1182

troodontid-like MAFIv.12685a Valioara,? II,2 5.8 4.1 r,g 0.1r 5 5 1 (Kadiccoll.)

troodontid-like MAFIv.12685b Valioara,? 8 3,9 2,3 2.05 0.6 5 5-6 0.91 (Kadicc011.)

cf, EuronychodonFGGUBR.1431 Sinpetru,1995 5.1 1.8 I 2.83 0.55 "LaCarare"

indet.theropod (?) FGGUBR.1583 Valioara,1995 1.5* 1 2 Fantanele

Table 1. Measurementsof theropod teeth from the Hateg Basin. All measurementsare given in mm Abbreviations: FABL - fore-aft basal length (Currie et al.,l99O): BW - basal width (Cunie et al.,1990); LCI - lateral compressionindex (Grigorescu, 1984b); DSDI - denticle size difference idex (Rauhut & Werner, 1995). * - incomplete tooth, with (...) approximated length.

9r ORYCTOS,Vol. 1, 1998

FIGURE 3. Small theropodteeth from the HategBasin. Velociraptorine dromaeosaurid : a - FGGUB R.1428;b - FGGUB R.1430;c - FGGUBR.I322 andd - FGGUBR.1580 (all in labialview). "Troodontid-like"small theropod: e - FGGUBR.1320 (?anteriordentarytooth)[email protected])in:f-labial;g-lingualview.Scalebar=3mm.

Distally they ate rounded and slightly apically dromaeosaurid teeth appear to be laterally compres- pointed. The labial and lingual sides are convex or sed, acutely tapered, recurved, seffated both fore and present a slightly depressed median longitudinal aft ... with disparity between the mesial (anterior) and groove bordered by more inflated anteior and poste- distal (posterior) serrations"; the same features are rior parts; this depression becomes shallow upward present also in the teeth from the Hateg Basin. and ends before reaching the tip. The teeth may pre- Characters such as strongly laterally compressed sent small wear facets, developed especially near the teeth bearing both anterior and posterior serrations tip and extending sometimes onto ihe anterior edge. with obvious size differences between them (the As noted by Ostrom (1969 : p. 157), the "tooth anterior ones being half as large as the posterior ones) form among theropods is perhaps even less reliable are listed by Ostrom (1990) as derived features than is dental formula, as a taxonomic criteria. The (synapomorphies) of Dromaeosauridae. Among dromaeosaurids,however, may be the exception. All them, velociraptorines (including Deinonychus, 92 CSIKI & GRIGORESCU- ROMANIAN THEROPODS

Saurornitholestes and Velociraptor) are diagnosed by teeth swallowed during feeding; its general morpho- more elongated, more pointed denticles, slightly api- logy supports the assignmentof R. 1582 as a veloci- cally hooked at their tips (Currie et al., 1990). The rapforine. following features seen in the teeth from the Hateg Basin are shared with velociraptorines (or dromaeo- saurids): strong lateral compression (LCI 2- "Thoodontid-like" small theropods (figs. 3e, 4a) Carpenter, 1982; Buffetaut et al., 1986; Rauhut &. Material: R.1318, (?)R.I3I9, R.1320, MAFI Zinke, 1995; Rauhut &. Werner, 1995); sharp distal v.12685a,b(see table 1.). curvature so that the tip extends well behind the base The tooth crown is conical, Iaterully compressed (see Ostrom,, 1969 on Deinonychus ; Osborn, L924 (but less so than the teeth here referred to as "veloci- and Sues, I9l7 on Velociraptor ; Currie et al., 1990 raptorine" or "euronychodont"; LCI > 0.6), and only on Saurornitholestes); anterior carina that does not slightly recurved distally (the tip only reaches the twist onto the lingual side and bears (usually smaller) level of the posterior end of the tooth base). Some serrations only on its distal paft (Ostrom, 1969 on teeth also show a moderate lingual curvature at the tip Deinonychus ; Osborn, 1924 on Velociraptor; Currie (R. 1318, MAFI v.I2685a). The anterior edge is et a1.,,1990 on Saurornitholestes; Carpenter, 1982; slightly curved, while the posterior one is almost Buffetaut et al., 1986); a posterior carina that is ser- straight, very slightly concave only near the tip; the rated over its entire length with denticles at least resulting general shapeis that of an almost isosceles twice as large as those from the anterior carina triangle. Both labial and lingual sides afe convex and (Ostrom, 1969 on Deinonychus ; Osborn, 1924 on the anterior and posterior carinae lay in the midline; Velociraptor ; Sues, 1978, Currie et al., 1990 on these carinae are strong and keel-like (especially so Saurornitholestes ; Carpenter, 1979, 1982; Buffetaut in the larger specimens). Both carinae show serra- et 4I., 1986; Rauhut &. Zinke, 1995; Rauhut & tions; when well preserved,they extend from the base Werner, 1995); a shallow sulcus on both sides that to the tip anteriorly as well as posteriorly. does not extend to the tip (Carpenter, 1982), straight, The denticles on the carinae are relatively wider distally slightly hooked denticles (Currie et al., 1990; at their bases than in the pu{ported velociraptorine Rauhut &, Werner, 1995) and wear facets usually teeth and are sub-equal on the anterior and posterior developed only near the tip (Ostrom, 1969 on carinae (- 5 denticles per mm; DSDI = 0.91-1). Deinonychus ; Carpenter, 1982). However, the height of each denticle is usually less It is worth mentioning that R. 132I and R. 1322 than their basal width (it equals 50Vo of the basal were previously identified as velociraptorines, possi- width on the posterior, but only 35Voon the anterior bly related to Saurornitholestes(Currie, I99I, pers. carina in R. 1318 and 90Vo posteriorly in MAFI comm.). v.I2685a,b). The denticles decreasein size near the The teeth differ from those of troodontids by tip or the base; they are rectangular, perpendicular to having differently shaped and relatively smaller den- the tooth axis and distally very slightly rounded, ticles and by the absenceof the "blood pits" (Currie, almost straight. The blood groove between denticles 1987) and a basal constriction. They also differ from tends to expand into a "blood pit" without attaining those of by their non-lanceolate the development seen in troodontids (Currie, 1987). shape and apically pointed, relatively larger denticles The sides of the teeth are smoothly convex; only and from those of "paronychodons-euronychodons" some teeth (ex. R. 1319) are slightly folded at their in their biconvex cross-section and serrated carinae. base, the folds continuing for a very short distance up As already mentioned, both premaxillary lanterior the crown. There is a constriction at the base of the maxillarylanterior dentary teeth and more posterior crown, effiphasized by the strong carinae. R. 1318 ("cheek") teeth were identified, basedmainly on their shows a large wear facet on the antero-terminal part cross-section.Even a so-called "digested tooth" was of the (?)labial side, extending down the crown for recognized, too (R. 1582); this unseffated, chalky almost half of its length (Fig . 4a); this deep, oval, gray-colored tooth with enamel-less surface is simi- obliquely anteriorly dipping facet extends onto the lar to those describedby Currie et al. ( 1990) as shed anterior carina as well. 93 ORYCTOS,Vol. 1, 1998

FIGURE 4. SEM photographsof small theropod teeth from the Hateg Basin : a - "troodontid-like" small theropod tooth (FGGUB R.1218); b - velociraptorine dromaeosaurid tooth (FGGUB R.1321). Enlargementsof the boxed areasshow the morphological distinctnessof the denticles in the two taxa.

When compared to the dentition seen in other In all these respects, the above described teeth Cretaceous small theropod taxa (dromaeosaurids, differ from those of either dromaeosaurids, troodontids, Richardoestesia,, "paronychodons", Richardoestesia or "euronychodons"; they differ, "euronychodons") the above-describedteeth compa- however, from those of the troodontids and "parony- re closer to those of the troodontids. They share the chodons" as well in the relatively smaller denticles (5 following characters: constriction at the base of the denticlesper mm vs. 1.5-3 denticlesper mm) which crown (considered as a troodontid synapomorphy by are of different shape (straight, low, rounded versus Currie, 1987); relatively large, subequaldenticles on strongly apically hooked). The teeth also show clear- the carinae; blood pits developed at the base of the cut distinction from any other small theropod teeth denticles (although at a lesser degree than in known from Hateg and they conceivably represent Troodontidae); these character association is also another taxon, here informally called "troodontid- shared with the "paronychodons" (sensu Zinke 8. like" and probably more closely related to troodon- Rauhut, 1994). Moreover, the only slightly-recurved tids (?and "paronychodons") than to other small the- shape of the tooth and the shape and position of the ropods (but, as Zinke &. Rauhut, 1994, cautioned, wear facet compare closely to that seen in Troodon there are severedifficulties in assigning isolated teeth formosus (see for ex. R.1318, fig . 4a comparativeto to certain taxa in less well-known faunas, such as Currie et al., 1990 : fig. 8.3.c a maxillary tooth; from Hateg, and one cannot consider isolated teeth as Currie, L987 : figs . 4e, 5k, 5r, 5u premaxillary, "troodontids" based solely on characterson the basis maxillary, anterior and posterior dentary teeth, res- of a "constricted base" - although this may well be a pectively). Among troodontids, the teeth from Hateg tempting assumption). are more similar to those of Troodon formosus (Currie, 1987) than to any of the Asian taxa with 3- cf. Euronychodon (fig. 3f, g) known dentitions (unnamed troodontid from the Material: R. 143I (seetable 1). Early Cretaceousof Mongolia, Barsbold et al., 1987; R. 143I is a small, elongated, strongly recurved mongoliensis, Osborn, 1924; S. and pointed tooth. The crown is strongly convex on junior, Barsbold, I97 4; youngi,, the labial side, and flat on the lingual side. Both the Russell 8. Dong, 1993) in that the anterior carina is anterior and the posterior carinae are positioned on serrated to the tip. It is worth mentioning that Currie the lingual side and are unselrated; they origin ate at identified one of these teeth (R. 1318) as troodontid the tip (although some uncertainty remains as the (1991,pers. comm.). tooth is rather worn), but end before reaching the

94 CSIKI & GRIGORESCU- ROMANIAN THEROPODS base. The same lingual side shows two longitudinal Maastrichtian beds and conceivably may represent a grooves just near the carinae, separatedby a median distinct taxon of theropods" (ibid., p. II7). Sigé et al. ridge; the anterior groove is naffower and deeper, (1997) consider Euronychodon morphologically while the posterior one is wide but shallow. identical to , taxon showing then a Teeth with the same character complex are wide range of variability; however, their figure and known from several localities ranging from the Early the tentative comparison with Richardoestesia (a Cretaceous ( of Ufla, Spain; Rauhut &. North American taxon with teeth of distinct morpho- Zinke, 1995) to the Late Cretaceous (Campanian of logy including serratedanterior and posterior carinae, Champ-Garimond, , Sigé et al., 1997; Early clearly different from R. 143I; Currie et al., 1990) Maastrichtian of Taveiro, Portugal, Antunes 8. cast doubts on the coffectnessof their conclusion. Sigogneau-Russell, I99I; Lafro, Spain, Le Læuff, Rauhut &. Zinke (1995) also consider their cf. 1992 and Quintanilla del Coco, Spain, Pol et al., Euronychodon as a valid taxon, possibly related to 1992, respectively ; Maastrichtian of North America, the primitive ornithomimosaur Pelecanimimus poly - Currie et al., 1990). The North American teeth, odon from the Barremian of Las Hoyas, Spain previously referred by Cope (1867, cf.Currie et al., (Percz-Moreno et al., 1994). 1990) to Paronychodonlacustris, a taxon widely dis- The occurrence of cf. Euronychodon teeth in the tributed in the Campanian-Maastrichtian of North Late Cretaceous of Hateg along with those in America, were recently discussed by Currie et al. Portugal and Spain, may provide evidence of the (1990); theseauthors restrict the name Paronychodon existence of a peculiar, small theropod, widely distri- only to the teeth with unseffated carinae coming buted in the south European archipelago. mostly from Maastrichtian beds. Defined as such, P lacustris differs from R. 143I, however, in having a 4- indet larger number of longitudinal ridges (but it should be Material: R. 1583. (Note added in coffection: noted that no author figured any "true" Paronychodon more, uncatalogued referred specimens were recent- teeth, so that no direct comparisons can be made ly recovered from the same site the Fantanele here). microverteb rate site at Valioara). Closely comparable teeth from the Cretaceousof Only the tip and distal part of the crown are pre- Europe were described as Euronychodon portucalen- served. The crown is laterally compressed,pointed, sis (Antunes &. Sigogneau-Russell, I99l) and cf. curved distally and slightly lingually. The labial side Euronychodon (Rauhut 8. Zinke, 1995); these ate is largely rounded transversely,but with a small, lon- more similar (indeed, almost identical; compare fig. gitudinally depressed area near the anterior edge, an 3f,,g to Rauhut&.Zinke, 1995:fig. lb) to R.143I. areathat ends at some distance from the tip. The lin- However, it should be noted that the teeth reported as gual side is more flattened (but still convex) with a cf. Paronychodon by Pol et al. (1992 : fig. 5a) and similar, wider depressionpresenting two low longi- Sigé et al. (1997 : fig. 13) differ from R. 1 43I in tudinal ridges. The most peculiar feature of the having a convex (largely convex in the Quantanilla crown is represented by the two non-serrated, but del Coco specimens)lingual side, being serrateddis- sharp, thin carinae on the anterior and posterior tally (Quantanilla del Coco) or missing anterior and edges.As the specimenis well preservedand unworn posterior carinae at all (Champ-Garimond). at its tip, clearly its edges would have also been Norman (1990) and Le Læuff (1992) listed smooth in life. Euronychodon as nomina dubia and This laterally compressed, pointed tooth most Theropoda nomina dubia, respectively.Le Læuff (op. probably belongs to a small theropod dinosaur. cit.) considered the referred teeth as "growth anoma- However, the unseffated condition of both the ante- lies", following Currie et al. (1990), although even rior and posterior carinaeis uncommon among thero- these authors did not reject the name Paronychodon pods; it is present in the pu{ported primitive mani- on the same ground, noting that: "the name raptoran Lisboasaurus estesi from the Kimmeridgian Paronychodon lacustris should be restricted to non- of Guimarota, Portugal (but see Buscalioni et al., serrated forms. These tend to be more common in 1996, who regard L. estesi as a crocodylomorph), in 95 ORYCTOS.Vol. 1, 1998

Spinosaurus from the early Late Cretaceous of nor- by Harrison & Walker, I9l 5), caudal vertebrae thern Africa (Elzanowski & Wellnhofer, 1992), in the (FGGUB R.70, 7I; Grigorescu, 1984b), proximal primitive ornithomimosaur Pelecanimimus (Perez- ulna and distal tibia (FGGUB R.72 and 73, respecti- Moreno et al., 1994) and in the basal metornithine vely; Grigorescu, I984b), paired frontals (MAFI olecranus (Perle et al., 1993). v. 13 528; Jianu & Weishampel, 1997) and a partial Among these taxa the only one that might be skull roof (associated frontal, MCDRD 454 and positively compared to R. 1583 is Pelecanimimus, parietals MCDRD 254, described as dromaeosaurid whose teeth are more blade-like, but still with unser- closely related to Saurornitholestes by Weishampel rated carinae distally in the jaws (Mononykus has and Jianu, 1996). These remains will not be described leaf-shaped, spatulate, straight teeth, here (for detailed descriptions, seethe references)but teeth are conical, while those of Lisboasaurus are lin- their taxonomic assignments will be discussed gually flat and unrecurved) In this case, and if (except for the MCDRD and MAFI material in whose Rauhut and Zinke (1995) correctly pointed out a pos- referral we agree with Weishampel and Jianu, 1996 sible Euronychodon Pelecanimimus relationship, and Jianu and Weishampel, 1997, respectively). then R.1583 could belong to the same taxon as the Vfe prefer to ffeat the remains as separate items tooth referred to above as cf. Euronychodon. It (contra Le Loeuff, 1992, who assignedall the Hateg should be noted, however, that no skeletal material small theropod material to the dromaeosaurid referable to any primitive ornithomimosaur has been Elopteryx nopcsai). reported until now from Hatog. The proximal femoral fragments ('Elopteryx Alternatively, the tooth may represent juvenile nopcsai',,figs. 5a-e) show several features that point (neonate) velociraptorine dinosaurs as those descri- to their affinity with maniraptorans. First of all, bed by Norell et al. (1994).The outline of R. 1583 (as 'Elopteryx' is member of the ,sharing with well as that of the newly recovered, still uncatalogued them the followitrg synapomorphies : specimens) confoffns to that of the teeth figured by . mediodorsally inclined femoral head (Perez- these authors (Norell et al., 1994, fig. 2.), and their Moreno et al., 1993); generally small dimensions (tooth height rarely o lesser trochanter laterally displaced, adjacent and exceeds 1,5 rrrnl, and usually falls in the 0.5- 1 mm cranial to the greatertrochanter and projecting above interval) ate also suggestive of their juvenile origin. the proximal margin of femoral head (Rowe &. They seem, however, to be more labio-lingually com- Gauthier, 1990; Perez-Moreno et al., 1993); pressed than the velociraptorine teeth from Ukhaa . absenceof a trochanteric shelf (Percz-Moreno et al., Tolgod. As a more remote possibility, the teeth might 1993). come from another, non-dinosaurian archosaur (cro- It is also member of Coelurosauria, sharing the codilian) taxon. following characters : . tip of lesser trochanter level with the greater tro- chanter proximally; REASSESSMENTS OF PREVIOUSLY . fourth trochanter weak or absent (Perez-Moreno et PUBLISHED SMALL THEROPOD REMAINS al., 1993; Serenoet al., 1996). Among coelurosaurs, 'Elopteryx' can be charac- As previously mentioned, other described small terized by : theropod material from the Hateg Basin includes o confluence of the lesser trochanter with the greater proximal left femoral fragments (BMNH A. 1234, trochanter (Benton, 1990), condition shared with the 1235; in I9I3 Andrews described them as Elopteryx troodontids Saurornithoides mongoliensis (Russell, nopcsai), a distal left femoral fragment (FGGUB 1969, Currie &. Peng, 1993) and ? Sinornithoides R.351, describedas Elopteryx nopcsai by Grigorescu youngi (Russell &. Dong, 1993), the dromaeosaurid & Kessler, 1981; we do not agtee with Le Læuff, Variraptor mechinorum (Le Læuff & Buffetaut, this 1992 who listed R.351 as a distal tibiotarsus), distal volume ; seealso Le Loeuff et al., 1992) and the basal tibiotarsi (BMNH A. 1528, 1588, 4359, describedas metornithine Mononykus olecranon (Perle et al., Bradycneme draculae and as Heptasteornis andrewsi 1993). However, the continuous plate of bone formed

96 CSIKI & GRIGORESCU- ROMANIAN THEROPODS

I

FIGURE 5. Small theropod hindlimb fragments from the Hateg Basin. 'Elopteryx nopcsai': proximal femur (BMNH R.1234) ina-cranial;b-caudal;c-mediocaudalandd-proximalview.Proximalfemur(BMNHR.1235)ine-caudalview. 'Bradycnemedraculae': left distal tibiotarsus (BMNH R.4359) in f - cranial and g - caudal view; right distal tibiotarsus (BMNH R.1588) in h - cranial and i - caudal view. Scale bar = 1 cm. (After Le Loeuff, 1992)

by the two trochantersare oblique as in dromaeosau- In ,however, the posterior trochanter rids (seefig. 5d) ratherthan pe{pendicular to the long is differentlyshaped (op. cit.: fig. 6). As the absence axis of the femoral head as in troodontids(Currie & of the posterior trochanter is a synapomorphyof Peng, 1993); (Chiappe & Calvo, 1994),the aviannatu- . well-developedposterior trochanter;shared with re of 'Elopteryx' is seriouslydoubted. dromaeosaurids (Ostrom, 1990), troodontids (Troodonformosbts, Le Læuff, L992;Saurornithoides . very reduced or absent fourth trochanter; shared mongoliensis,Currie &. Peng, 1993;Sinornithoides with troodontids (Sinornithoides,Saurornithoides), youngi, Russell& Dong, 1993), portento- dromaeosaurids(Deinonychus, Variraptor), elmisau- sus (Norm&î, 1990), Archaeopteryxlithographica rids (Currie, 1990), celer (Normàî, and enantiornithinebirds (Chiappe& Calvo, 1994). 1990),Mononykus (PerLe et al., 1994).

97 ORYCTOS,Vol. 1, 1998

scc

a b c d

FIGURE 6. Small theropod (?abelisaurid):distal femur (FGGUB R.351) in a - medial; b - lateral; c - caudal and d - cranial view. Abbreviations: tfg - tibiofibular groove; scc - supracondylarcrest. Scale bar = 10 mm.

Moreover, as 'Elopteryx' lacks certain synapo- If not avialian, R.351 shows featuresthat could indi- morphies of or higherbirds (absence cate a close relationship to ceratosaurs: of posterior trochanter; Chiappe &. Calvo , 1994), it . absence of a well-marked cranial intercondylar hardly can be positioned among the Neornithes as it groove on the distal femur. The presence of a cranial was proposedby Andrews (1913). intercondylar groove is considered a synapomorphy Finally, although we cannot (and would not) of Tetanurae by Perez-Moreno et al. (1993) and definitively rule out the conclusion reached by Le Novas (1992). Loeuff et al.(I992), it is still a possibility that This character is also shared with the derived mani- 'Elopteryx' is a troodontid or some other derived raptoran Avimimus (Norm &f, 1990), dromaeosaurids maniraptoran instead of a dromaeosaurid. (Ostrom, 1976), troodontids (Cunie &. Peng, 1993), The distal femur (FGGUB R.3 5I, fig. 6) referred ornithomimosaurs (with the possible exception of to Elopteryx by Grigorescu & Kessler (1981) shows , Molnar et al., 1990) and oviraptoro- some peculi ar features worth of noting. Its suggested saurs. avian nature can be questioned because it seems to . presence of a distinctively deep groove at the base lack several characteristic avian features (Chiappe & of the crista tibiofibularis (Rowe &. Gauthier, 1990; Calvo, 1994),as : listed as a synapomorphy of Ceratosauria by Percz- o femoral popliteal area bounded distally by a trans- Moreno et al., 1993; HoItz, 1994). verse bridge (synapomorphy of Metornithes); o presenceof a non-elliptical muscle scar on the cta- o presence of a prominent patellar groove (synapo- niodistal region of the femur. morphy of Enantiornithes + Ornithurae). This character is shared with neoceratosaursi.e. the However, it still can belong to an avialian theropod, common ancestor of and sharing the following synapomorphy with and all its descendants,and some non-tetanuran avi- Enantiornithes + Ornithurae : pods (sensu Novas, 1992) as Pianitzlcysaurus, ' tibiofibular crest on the lateral condyle of the distal (Perez-Moreno et al., 1993). femur.

98 CSIKI & GRIGORESCU- ROMANIAN THEROPODS

. presence of a craniomedial crest originating above Arctometatarsalial clade (under DERRAN option of the medial (tibial) condyle, called the supracondylar PAUP). BMNH 4.1588 and 1528 (at least) show this crest. This crest has been described in Carnotaurus transverse groove (comp are Harrison 8L Walker, sastrei, a middle Cretaceous abelisaurid from I97 5 : figs. lb, c, also fig. 59, this paper, with the Argentina and also reported to be present in the condition seen in , Currie &. Zhao, 1993: European Maastrichtian abelisaurids Betasuchus and figs.23 D, J). "Laflo theropod" (cf. Tarascosaurus) (Le Læuff &. Among Tetanurae 'Bradycneme' can be closely Buffetaut, 1991). compared to troodontids, Avimimus and derived A prominent craniolateral tubercle (called "apo- ; all of thesetaxa show a more or less well fused physis for the external gastrocnemial muscle" by tibiotarsus. The broad, cranio-caudally flattened Grigorescu 8. Kessler, 1981) is also present in shape of the distal tibiotarsi is similar to that seen in Avimimus portentosus (Normâr, 1990). troodontids (see Russell, 1969 : fig. 11 Troodon; Autapomorphies of R.351 include the triangle- Barsbold, 1974 : fig. 4 - Saurornithoides;Osmolska, shaped, depressed popliteal area bordered by the 1987 : fig. 1 - ), showing well pronounced, converging ascending ridges of the distal articular cranially projected condyles separatedcranially and condyles and a cranially extended medial condyle. distally by a deep intercondylar groove that becomes The relationshipsof R.351 are not easy to establish; very naffow caudally; a calcaneum that is either as shown, it sharesmore characterswith neocerato- fused to the astragalus (Russell 8. Dong, 1993) or saurs(the clade that also includes abelisaurids,other- missing (Currie & Peng, 1993); and an external fossa wise known from the Late Cretaceous of France and at the base of the ascendingprocess. Spain; Buffetaut &Le Læuff ,, L991) than with either This later character is listed as character I23 by dromaeosaurids,troodontids or other tetanuranthero- Holtz (1994) as being present in some members of pods. Neither can it belong to 'Elopteryx nopcsai' Troodontidae, and Ornithomi- which represents a derived maniraptoran theropod, mosauria; from these taxa, Tyrannosauridae and unless '8. nopcsai' is a peculiar, unknown type of have broader distal fibular ends theropod. However, more and better preserved mate- (than can be presumed for 'Bradycneme') that arti- rial is neededto solve this question. culate with a distinct calcaneum and condyles that are The tibiotarsi are here referred to as separated by broad, shallow depressions rather by 'Bradycneme draculae' (fig. 5f-i). We agree with Le grooves (Molnar et al., 1990; Barsbold & Osmolska, Loeuff et al. (1992) who synonymize Harrison &. 1990). Moreover,' Bradycneme'also compareswith Walker's (I97 5) Bradycneme and Heptasteornis troodontids in the following features: Iateral condyle (contra Paul, 1988), but disagreein that the tibiotarsi extends more distally than the medial one; cranially necessarily belong to 'Eloptery'; in fact it will be slightly divergent condyles; lateral condyle triangle- shown that they probably do not belong to a taxon of shapedin distal view with a craniolaterally projecting derived maniraptoran affinities. "lip" (compareRussell, 1969 : fig. 11c with Harrison 'Bradrvcneme'has the following derived charac- & V/alker, I97 5 : pl. 65, fig. 5) and a narrow latero- ters in common with Tetanurae (Sereno et al., 1996): cranial groove for articulation with the fibula. It dif- . astragalar articular condyles oriented craniodistally ; fers, however, in the unusually asymmetrical, "bro- . astragalar ascending process plate-like, relatively ken" cranial view, the presence of a transversegroo- well developed; ve on the cranial face of the articular condyles (see o astragalat condyles with cranial transverse groove. above) and a relatively lower, triangular rather than The characteris listed by Sereno et al. (1996; cha- tall, parallel-sided astragalar ascending process (if racter 19) as a tetanuran synapomorphy, reversed in correctly interpreted from the figures of Harrison 8L coelurosaurians more derived than , Walker, I97 5). and by Holtz (1994; character 23) as a tetanuran Le Læuff (1992) suggested that these tibiotarsi synapomorphy reversed in either Coelurosauria are similar to that of Deinonychus; however, we are (under ACCTRAN option of PAUP) or the unable to identify any specific similarity with [Dromaeosauridae+Aves] + [ + Deinonychus(see Ostrom,1969 :fig. 68) which has a 99 ORYCTOS,Vol. 1, 1998 shallower intercondylar groove cranially (i.e. cranial- with certain primitive ornithomimosaurs (such as ly less protruding condyles), no fossa at the base of Pelecanimimus). the ascending process and no transverse groove on 'Elopteryx nopcsai' is restricted to the proximal the articular condyles. In conclusion 'Bradycneme femora of a derived maniraptoran, close to either dro- draculae' is considered here as another small thero- maeosaurids, troodontids (or even Avimimus), or pod taxon from Hateg that shares most characters representing a new taxon within this clade ("elopte- with troodontids, but also has a synapomorphy that rygines" of Le Loeuff et aL.,1992). seems to exclude it from the derived coelurosaurs By contrast, the distal femur previously referred (see above, Holtz's character 23). Thus, the systema- to Elopteryx nopcsaiby Grigorescu& Kessler(1981) tic position of '8. draculae'cannot be more precise- shows synapomorphies of neoceratosaurian thero- ly assessedat present.This is why we prefer to use pods and may represent a small abelisaurid. Small- the name' Bradycneme' only informally. sized abelisaurids are known from the Late The other remains from Hateg that were pre- Cretaceousof both South America (Noasaurus leali; viously assigned to small theropods do not warrant Novas, 1992) and Europe (Tarascosaurussalluvicus, special discussion. The vertebrae, fragmentary tibia Betasuchus bredai; Le Læuff &, Buffetaut, I99I, and ulna figured by Grigorescu (1984b) do not exhi- 1995), but, owing their fragmentary nature, no direct bit any certain theropod features and are here regar- comparisonscan be made. ded as indeterminate dinosaurs.A small, slender tibia 'Bradycneme draculae' is informally retained to from Tustea locality (FGGUB R.252), labeled as denominate the small theropod taxon representedby "theropod" does not belong to a theropod dinosaur as the distal tibiotarsi. Although the possibility, expres- it lacks a crest for the articulation of the fibula (fibu- sed by Le Læuff et al. (1992), that they do belong to lar crest; Benton, 1990). the same taxon as the proximal femora, cannot be ruled out, 'Bradycneme' shows at least one clear synapomorphy (cranial transversegroove on astraga- CONCLUSIONS lar condyles) that seemsto exclude it from the mani- raptoran (even coelurosaurian) Tetanurae, a clade Reevaluation of the remains belonging to rare where 'Elopteryx' was shown to probably belong. small theropod dinosaurs recovered from the Hateg Without more complete, associatedmaterial, it is Basin, reco gnized only in the last decade, reveal a impossible to recognize whether such disarticulated previously unexpected diversity (see also Jianu &. material like that from the Hateg Basin belongs to the Weishampel, 1997). As no associated remains have same taxon; moreover, as their cuffently recognized so far been recovered, the isolated teeth, femoral synapomorphies show, the hindlimb elements seem fragments and distal tibiotarsi were treated separate- to represent taxa from different theropod clades. The ly in the present study. This approach presumably only case of recognizable conspecificity may be gives a greater apparent than real diversity; further represented by the velociraptorine teeth and the skull discoveries of better preserved, associated material roof fragments described by Weishampel &. Jianu may in fact reveal the presence of fewer new small (1996); but it must be pointed out that even if theropod taxa with unusual character associations, 'Elopteryx' should prove to represent a dromaeosau- not yet known from elsewhere. rid, it seems to exhibit such primitive characters that The study of the isolated small theropod teeth led exclude its conspecificity with the theropod of to the identification of three (perhaps four) different Weishampel and Jianu. Thus, the name 'Elopteryx' taxa: a velociraptorine dromaeosaurid, a "troodontid- should not be used for this latter taxon. The conspe- like" small theropod, and a Euronychodon - like the- cificity of the other small theropod remains is even ropod (in of decreasing abundance). Another harder to demonstrate for the present. Norman taxon with distally recurved, carinated but unseffated (1990 : p. 280) stated, that: " ... in ideal circum- teeth may be indicated by a tooth fragment. stances, taxa should be only erected and given bino- Conversely, this tooth may belong to the same taxon as mial status on the basis of diagnostic characters cf .Euronychodon, taxon with possibly close affinities (synapomorphies). If material is discovered which is 100 CSIKI & GRIGORESCU- ROMANIAN THEROPODS

of considerational scientific interest by virtue of its (Weishampel et al., 1991) which, in turn, is certainly stratigraphic or geographic position, or because it related to the insular, small-areahabitat of this fauna. represents new evidence of hitherto unappreciated V/e are not aware of any dinosaur fauna that lacks a faunal association, then it would be perfectly legiti- large top predator (usually represented by "megalo- mate to define the new material on general (family, saur-grade"theropods, allosauroids or tyrannosaurids ordinal status) terms rather than more specifically, in the northern continents, and abelisaurids in thus drawing attention to important material and Gondwana); even in the Late CretaceousAsian fau- ideally provoking greater research." ; therefore, we nas, with their diverse small theropod faunal compo- choose to report on but not to name any of the Hateg nent, the top predators afe represented by tyranno- small theropods until better preserved, more comple- saurids (though of smaller size; Holtz, 1994). Even in te material is found. The already proposed names the case of the "Djadochtan" (Jerzykiewicz 8L (Elopteryx nopcsai, Bradycneme draculae with Russell, I99I) faunal assemblages from the Bayn Heptasteornis andrewsi as junior ) are retai- Dzak and correlative strata, where small theropods ned only for informal usage. (VeI o c i rap t o r, Ov irap t o r; Sauro rnitho i d es, Av imimu s) At least three different taxa of small theropods are outstandingly abundant and tyrannosaurid arerecognized from the Hateg Basin: a neoceratosaur remains are rate, these remains were continuously (probably an abelisaurid), a (?non-coelurosaurian) found. The case of the Hateg fauna is singular and tetanuran ('Bradycneme draculae') and a derived surely can be explained by the restrictive, insular maniraptoran ('Elopteryx nopcsai'); the presenceof a environment these dinosaurs lived in. velociraptorine dromaeosaurid is further substantia- ted. How the described teeth may relate to these dif- ferent taxa is unclear for the present, but the Euronychodon-type teeth point to the presence of a ACKNOWLEDGMENTS peculiar kind of theropod known from the European Cretaceousonly by its teeth (it is noteworthy that nei- We would like to thank Dr. Kordos Laszlo at the ther neoceratosaurs nor known tetanurans seem to MAFI for his kind help during our stays in Budapest have had such teeth). However, we caution again that and for his permission to study and publish on the further discoveries of better preserved specimens specimens (MAFf v. I2658a,b) from the Hateg Basin may show that some of these remains with different in his care. Special thanks are due to Dr. Dave affinities may in fact represent a single taxon of new, Weishampel (Baltimore) for providing references,for peculiar small theropods. the arrangements made at the Johns Hopkins A final remark is worth making about the paleoe- University, Baltimore, USA for the SEM photos cological implications of the data presented here. reproduced in figure 4., and also for the discussions ContemporaneousEuropean faunas seem to have had regarding the Hateg Basin fauna and its theropods. larger theropods at the top of the food chain: abeli- Figures 3, 5 and 6 were skillfuly drawn by our col- saurids such as Tarascosaurus from the Campanian legues Mihai Tomescu, Cezar Gherasim and Adi and Early Maastrichtian of France and Early Cormos, to whom the authors thank this *ây, too. Maastrichtian of Spain; Betasuchus from the Late The authors also wish to acknowledge all the people Maastrichtian of Netherlands (Le Læuff & Buffetaut, who took part at the Hateg excavations and were of IggI, 1995) or "megalosaurid"-gradetheropods from great help in the recovering of the small theropod the Late Maastrichtian of Portugal (Antunes &. material; listing the names of all of them would take Sigogneau-Russell, 1996), along with small thero- to much space here, but neverthelesswe are grateful pods (dromaeosaurids, euronychodons, etc.). The to them. Research in the Hateg Basin was partially Hateg fauna, in turn, seems to have a more "diffuse" funded by the National Geographic Society and The top of the food pyramid, with a larger number of Dinosaur Society. small theropods replacing the top predator. This phe- nomenon seems to be related to the general body dwarfiSffi, already noted for other Hateg dinosaurs

101 ORYCTOS,Vol. 1, 1998

NOTE ADDED IN CORRECTION taxon of new, peculiar small theropods") may turn out to be coffect. Consequently, rather than represen- After submitting the manuscript, we became ting a wide assemblageof small theropods,the hind- aware of the description of a new theropod dinosaur limb elementsfrom the Hateg Basin, discussedin the taxon from the Late Cretaceous of Mongolia paper, might as well belong to a new, unusual thero- (Osm6lska, 1996), Bagaraatan osborni. The holoty- pod taxon. May it be like this, the new theropod taxon pe specimen of Bagaraatan includes among others from Hateg might be closely related to Bagaraatan the almost complete femur and tibiotarsus, evidently osborni (and possibly to the Iren Nor avimimid, as found in association. well), thus presenting evidence of the Asian affinities Osmdlska ( 1996) was rather confused when of the Hateg fauna (already suggested by Csiki, assessitrg the relationships of the new theropod 1gg5). taxon. She noted the presence of unusual character associationsin Bagaraatan, such as the confluence of REFERENCES ANDREWS, C. W. I9I3. On some bird remains from the Upper the lesser and greater trochanters and the presenceof Cretaceousof Transylvania.Geological Magazine, 5 : 193-196. proxi- a well-developed posterior trochanter on the ANTONESCU, E.; LUPU, D. & LUPU, M. 1983.Corrélation paly- mal, respectively the absenceof well-marked cranial nologique du Crétacéterminal du sud-estdes Monts Metaliferi et intercondylar groove and the presenceof a distincti- des Dépressionsde Hateg et de Rusca Morrtana.Annales de vely deep groove at the base of the crista tibiofibula- I'Institute de Géologie et Géophysique,59 : I I-11 . ris on the distal femur (all these features are also ANTUNES, M. T. & SIGOGNEAU-RU:SSELL,D. 199I. Nouvelles données sur les Dinosauresdu Crétacésupérieur du Portugal. found in an indetermined avimimid frclm the Late Comptes rendus Académie des Sciences de Paris, II, 313 : Cretaceousof Iren Nor; Osmdlska, 1996 : p. 31). On 113-1 19. the crus there are also worth noting the fusion bet- 1996. Le Crétacé ween the tibia and the proximal tarsals and the pre- terminal portugais et son apport au problème de I' des senceof a horizontal groove on the cranial face of the dinosaures.Bulletin du Muséum national d'Hisloire naturelle. astragalus. Bagaraatan also shares with the Hateg Paris,4, 18C (a): 595-606. BARSBOLD, R. 1974. Saurornithoididae,a new family of small theropod(s?) the presence of a latenl tubercle (ecte- theropod dinosaurs fiom Central Asia and North America. picondyle of Osmdlska, 1996, fig. 11) on the lateral PalaeontologiaPolonica, 30 : 5-22. articular condyle of the femur (also present in the -OSMÔLSKA, H. & KURZANOV S.M. T987.ON Iren Nor avimimid), anarrow popliteal areaand head a new troodontid (Dinosauria, Theropoda) from the Early of femur raised slightly above the level of the proxi- Cretaceous of Mongolia. Acta Palaeontologica Polonica, mal end of the greater and lesser trochanters (at the 32 (r-2) : 121-132. 1990. Ornithomimosauria pp samelevel in Deinonychus). Finally, the author notes ; 225-244. In WEISHAMPEL, D.B.; DODSON, P. & the features shared by Bagaraatan with the cerato- OSMÔLSKA, H.(eds.) The Dinosauria. University of California saurs(including those of the hindlimb; also presentin Press,Berkeley and Los Angeles. R. 351 or 'Bradycneme'), although considers them BENTON, M.J. 1990. Origin and interrelationshipsof dinosaurs ; homoplasic rather than suggestingreal phylogenetic pp. 11-30. In WEISHAMPEL, D.B.; DODSON, P. & relationships. Consequently, in spite of the great OSMÔLSKA, H. (eds.) The Dinosauria. University of California Press,Berkeley and Los Angeles. number of conflicting characters, Osmôlska places BUFFETAUT, E.; MARANDAT, B. & SIGÉ, B. 1986. Découverte the new taxon from the among the de dents des Deinonychosaures(, Theropoda) dans le Avetheropo da, a position that is concordant with that Crétacé supérieur du Sud de la France. Comptes-rendus de suggested by us for 'Elopteryx' and partially with I'Académiedes Sciences de Paris,II, 303 (15) : 1393-1396. that for 'Bradycneme' BUSCALIONI, A.D.; ORTEGA, F.; PEREZ-MORENO, B. &. Bagaraatan osborni thus presents some of the EVANS, S.E. 1996. The Upper Jurassicmaniraptoran theropod Lisboasaurusestesi (Guimarota, Portugal) reinterpreted as a cro- peculi ar featuresnoted by us in the isolated theropod codylomorph. Journal of Vertebrate Paleontology, 16 (2): hindlimb material from Hateg Basin, showing that 356-362. our cautioning ("further discoveries of better preser- CARPENTER, K. 1979. Vertebratefauna of the ved specimens may show that some of these remains (Maastrichtian), Weld County, Colorado. Contributions to with different affinities may in fact represent a single Geology, University of Wyoming,lT (1) : 37-49. r02 CSIKI & GRIGORESCU- ROMANIAN THEROPODS

1982. Baby dinosaurs from the Late Cretaceous HARTENBERGER, J..L. ; RADULESCU, C.; Lance and Hell Creek formations and a description of a new SAMSON: & SUDRE, J. 1985. Découverte de Mammifères et of theropod. Contributions 1o Geology, University of Dinosaures dans le Crétacé supérieur de Pui (Roumanie). Wyoming,20(2) : 123-134. Comptesrendus Académie des Sciencesde Paris, II, 301, (19) : CHIAPPE, L. & CALVO, J.O. 1994. Neuquenornisvolans, a new 1365-1368. Late Cretaceous bird (Enantiornithes: ) from HARRISON, C.J.O. & V/ALKER, C.A. I9l5. The Bradycnemidae, Patagonia, Argentina. Journal of Vertebrate Paleontology, a new family of from the Upper Cretaceous of Romania. 14 (2) :230-246. Palaeontology,18 (3): 563-510. CSIKI, Z. 1995. Studiul paleontologic si tafonomic al faunai HOI:IZ, T.R. 1994. The phylogenetic position of the Maastrichtian superioare din depozitele continentale ale Tyrannosauridae:implications for theropod systematics.Journal Bazinului Hateg (Formatiunile de Densus-Ciulasi de Sinpetru) : of Paleontology,68 (5): 1100- nn . implicatii paleoecologice. Unpubl. thesis, University of HOWSE, S.C.B. & MILNER, A.R. 1993.Ornithodesmu^s - a mani- Bucharest: I-91 . raptoran theropod dinosaur from the Lower Cretaceous of the CURRIE, P.J. 1987. Bird-like characteristicsof the jaws and teeth of Isle of Wight, England.Palaeontology,36 (2): 425-437. troodontid theropods (Dinosauria, Saurischia). Journal of IANOVICI, V.; BORCOS, M.; BLEAHU, M.; PATRULIUS, D.; VertebratePaleontology, 7 : 72-8I . LUPU, M.; DIMITRESCU, R. &. SAVU, H. 1976. Geologia - 1990.Elmisauridae ;pp. 245-248.IaWEISHAMPEL, Muntilor Apuseni. Ed. Academiei Romane,Bucharest : l-631. D.B.; DODSON, P. & OSMÔLSKA, H. (eds.)The Dinosquria. JERZYKIEV/ICZ,, T. 8. RUSSELL, D.A. 1991. Late University of California Press, Berkeley and Los Angeles. stratigraphy and vertebrates of the Gobi Basin. Cretaceous - RIGBX f. K. & SLOAN, R.E. 1990.Theropod teeth Research,12 : 345-317. from the Judith River Formation of southern Alberta, Canada: JIANU, C.-M. & WEISHAMPEL, D.B .1991 .A new theropoddino- pp.10l-125. In CARPENTER, K. &. CURRIE, P . J. (eds.) saur from the Hateg Basin, Western Romania, in the Hungarian Dinosaur Systematics: Approache s and Perspective s. Cambridge Geological Survey collection. Sargetia, ser. Sci. Nat., 17 : Universitv Press. '- 239-246. & PENG, J.-H. 1gg3. A juvenile specimen of LE LOEUFF, J. 1992. Les vertébrés continentaux du Crétacé Saurornithoides mongoliensis from the Upper Cretaceous of supérieur d'Europe : Paléoecologie, Biostratigraphie et northern China. Canadian Journal of Earth Sciences, 30 : Paléobiogeographie.Ph.D thesis, Mémoires des Sciencesde la 2224-2230. Terre UniversitéP. et M. Curie, Paris : l-21I. - & ZHAO, X.-J. L993. A new carnosaur(Dinosauria, 1993. European Titanosaurids. Revue de Theropoda)from the Jurassicof Xinjiang, People'sRepublic of Paléobiologie, vol. spéc. 7 : 105-lll . China. Canadian Journal of Earth Sciences,30 : 2037-208I. --& BUFFETAUT, E. I99I. Tarascosaurus DINCA, A.; TOCORJESCU,M. & STILLA, A. 1912.Despre virsta salluvicus nov.gen, nov.sp., dinosaure théropode du Crétacé depozitelor continentale cu dinosaurieni din Bazinul Hateg si supérieurdu sud de la France.Géobios,25 (5) : 585-594. Rusca . Ddri de Seamd, Institutul de Geologie si 1995.The evolution of Geofizica,Bucharest, 58 (4) : 83-94. Late Cretaceous non-marine vertebrate fauna in Europe ; pp. ELZANOWSKI, A. 1983. Birds in Cretaceousecosvstems. Acta 181-184.ftc SUN, A. & WANG, Y. (eds.)Sixth Symposiumon PalaeontologicaPolonica, 28 : I 5-92. Mesozoic Terrestrial Ecosystemsand Biota, Short Papers, China & WELLNHOFER: 1992.A new link between OceanPress, Beijing. theropods and birds from the Cretaceousof Mongolia. Nature, 1998.A new dromaeo- 359 : 821-823 saurid (Dinosauria, Theropoda) from the Upper Cretaceousof GRIGORESCU, D. 1984a.New tetrapodgroups in the Ma.astrichtian southernFrance. Oryctos, L. 105-112. of the Hateg Basin : Coelurosauriansand multituberculates;pp. MECHIN, P. 8. 99-104. In REIF, W-8. &. WESTPHAL, F. (eds.) Third MECHIN-SALESSX A. 1992.The first record of dromaeosaurid Symposium on MezozoicTerrestrial,Ecosystems, Short Papers, dinosaurs. (Saurischia, Theropoda) in the Maastrichtian of Tùbingen. southernEurope : palaeobiogeographicalimplications. Bulletin 1984b.New paleontologicaldata on the dino- de la Societégéologique de France,163 (3) : 37-343. saur beds from the Hateg Basin. Special Volume, 75 Years of the MOLNAR, R.E.; KURZANOV S.M. 8. DONG, Z. 1990. Laboratory of Paleontology, University of Bucharest ; 11 1- 1 18. Carnosauria; pp. 169-209.IaWEISHAMPEL, D.B.; DODSON, 1992. Nonmarine CretaceousFormations of P. &. OSMÔLSKA, H. (eds.) The Dinosauria. University of Romania ; pp. 142- 164. In}I4'NITER,N.J. & PEI- JI, Ch. (eds.) California Press,Berkeley and Los Angeles. Aspectsof Nonmarine CretaceousGeology. China Ocean Press, MUTIHAC, V. 1990. Structura geologicd a teritoriului României. Beijing. Editura Tehnicâ.Bucharest : l-419. &. KESSLER. E. 1981.A new specimen of -& IONESI, L. 1g74. Geologia României. Editura Elopteryx nopcsai from the dinosaurian beds of Hateg Basin. Tehnica.Bucharest : I-646. Revue Roumaine de Géologie, Géophysique et Géographie, NOPCSA, F. 1915.Erdély dinoszauruszai.Magyar kirdlyi Fôldtani Géologie,24 : I7I-175. IntézetÉvkdnyve, 23, I: 3-24. 103 ORYCTOS,Vol. 1, 1998

-1923. On the geological importance of the primitive Coco (Burgos Province, Spain). Neues Jahrbuch fu, Geologie reptilian fauna in the Uppermost Cretaceous of Hungary; with und PalciontologieAbhandlungen, 184 (3): 279-314. the description of a new tortoise. Quarterly Journal of the RAUHUT, O.W.M. & ZINKE, J. 1995. A description of the GeologicalSociety, 79, I Q14): 100-116. Barremian dinosaur fauna from Ufla with a comparison to that of NORELL, M.A.; CLARK, J.M.; DASHZEYEG, D.; BARSBOLD, Las Hoyas. II International Symposium on Lithographic R.; CHIAPPE, L.M.; DAVIDSON, A.R.; MCKENNA, M.C.; Limestones.Extended Abstracts : I23-I27 . PERLE, A. &. NOVACEK, M.J. 1994. A Theropod Dinosaur 8{. WERNER, C. 1995. First record of the Embryo and the Affinities of the Flamming Cliffs Dinosaur family Dromaeosauridae (Dinosauria: Theropoda) in the Eggs.Science,266 : 779-182. Cretaceous of Gondwana (Wadi Milk Formation, northern NORMAN, D.B. 1985. The illustrated encyclopediaof dinosaurs. Sudan.Palciontologische Zeitschrift, 69 (314): 475-489. SalamanderBooks. London: 1-208. ROWE, T. & GAUTHIER, J. 1990.Ceratosauria;pp. 151-168.In 1990.Problematic theropoda:"Coelurosaurs" ; pp. WEISHAMPEL, D.B.; DODSON, P. & OSMÔLSKA, H. (eds.) 280-305. In V/EISHAMPEL, D.B.; DODSON, P. &. The Dinosauria. University of California Press, Berkeley and OSMÔLSKA, H. (eds.) The Dinosauria. University of Los Angeles. California Press,Berkeley and Los Angeles. RUSSELL, D.A. 1969. A new specimenof Stenonychosaurzsfrom NOVAS, F.E. 1992.La evolucion de los dinosaurioscarnivoros ; pp. the (Cretaceous) of Alberta. Canadian 126-163.In SANZ, J.L. &. BUSCALIONI, A.D. (eds.) Los Journal of Earth Sciences,6 : 595-612. Dinosqurios y suo entorno biotico. Inst. "Juan de Valdes". -8. DONG, Z. 1993. A nearly complete skeleton of OSBORN, H.F. 1924. Three new theropoda,Protoceratops zone, a new troodontid dinosaur from the of the central Mongolia. American Museum Novitates,144 : L-I2. Ordos Basin, Inner Mongolia, People's Republic of China. OSMÔLSKA, H. 1987.Borogovia gracilicrus gen et sp.n., a new Canadian Journal of Earth Sciences,30 :2163-2113. troodontid dinosaur from the Late Cretaceousof Mongolia. Acta SERENO, P.C.; DUTHEIL, D.B.; IÀROCHENE, M.; LARSSON, PalaeontologicaPolonica, 32 (I-Z) : 133-150. H.C.; LYON, G.H.; MAGWENE, P.M.; SIDOR, C.A.; -1996. An unusual theropod dinosaur from the VARRICCHIO, D.J. & WILSON, J.A. 1996. Predatory Late Cretaceous Nemegt Formation of Mongolia. Acta Dinosaurs from the Sahara and Late Cretaceous Faunal PalaeontologicaPolonica, 41, I : 1-38. Differentiation. Science,272 : 986-991. & BARSBOLD, R. 1990.Troodontidae ; pp. 259- SIGÉ, B. ; BUSCALIONI, A.D. ; DUFFAUD, S. ; GAYET M. ; 268. In WEISHAMPEL, D.B.; DODSON, P. & OSMOLSKA, ORTH, B. ; RAGE, J.-C. & SANZ, I.L. l99l . Etat des données H.(eds.) The Dinosauria, University of California Press, sur le gisement Crétacé supérieur continental de Champ- Berkeley and Los Angeles. Garimond (Gard, Sud de la France). Miinchner OSTROM, J.H. 1969. Osteology of Deinonychusantirrhopus, an GeowissenschaftenAbhandlung, A 34 : 11 1- 1 30. unusual theropod from the Lower Cretaceous of Montana. SUES, H.-D. I9l7 . The skull of Velociraptor mongoliensis,a small Bulletin of the PeabodyMuseum of Natural History, 30 : 1-65. Cretaceoustheropod dinosaur from Mongolia. Palciontologische 1916.On a new specimenof the Lower Cretaceous zeitschrift, 5l (3/4) : 113-184. theropod dinosaur Deinonychusantirrhopus. Breviora, 439 : I-21. -1978.A new small theropod dinosaur from the Judith 1990. Dromaeosauridae ; pp. 269-279. In River Formation (Campanian) of Alberta, Canada. Zoological WEISHAMPEL, D.B.; DODSON, P. & OSMÔLSKA, H. (eds.) Journal of the LinneanSocieQ,62: 381-400. The Dinosquria. University of California Press, Berkeley and V/EISHAMPEL, D.B.; NORMAN, D.B. &. GRIGORESCU, D. Los Angeles. I99L The Dinosaurs of Transylvania: island biogeography in the PAUL, G. 1988. Predatory dinosaurs of the world. Simon & Late Cretaceous.National Geographic Research & Exploration, Schuster,New York : I-464. 7 : 68-81. pÉgEz-MoRENo, B.; sANZ, J.L.; SUDRE, J. & sIGÉ, B. 1993.A &. JIANU, C.M. 1996. New Theropod theropod dinosaur from the Lower Cretaceous of southern Dinosaur Material from the Hateg Basin (Late Cretaceous, France.Revue de Paléobiologie,vol. spéc.7 : 173-188. Western Romania. Neues Jahrbuch fii, Geologie und

BUSCALIONI./ v. \^t A.D.: MORA- PalciontologieAbhandlltngen, 200 (3): 387 -404 ^g^v^ ^r^v^v TALLA, J.J.;ORTEGA, F. & RASSKIN-GUTMAN, D. 1994.A ZINKE , J. &. RAUHUT, O.W.M. 1994.Small theropods(Dinosauria, unique multitoothed ornithomimosaur from the Lower Saurischia)from the Upper Jurassicand Lower Cretaceousof the Cretaceousof Spain.Nature,370 : 363-367. Iberian Peninsula. Berliner GeowissenschaftenAbhandlung, PERLE, A.; CHIAPPE, L.M.; BARSBOLD, R.; CLARK, J.A. & E13 : 163-177. NORELL, M.A. 1994. Skeletal Morphology of Mononykus olecranus (Theropoda: ) from the Late Cretaceous of Mongolia. American Museum Novitates,3105 : I-29. POL, C. ; BUSCALIONI, A.D. ; CARBALLEIRA, J. ; FRANCÉS, Note reçue le 01-07-1998 acceptéeaprès révision le l5-03-1998. V. ; MARTINEZ, N.L. ; MARANDAT, B. ; MORATALLA, J.J.;SANZ, J.L. ;SIGÉ, B. &VILLATTE, J.1992.Repriles and mammals from the Late Cretaceousnew locality Quantanilla del r04