<<

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/232388110

A systematic reassessment of Early multituberculates from Galve (, )

Article in Cretaceous Research · February 2011 DOI: 10.1016/j.cretres.2010.10.003

CITATIONS READS 12 34

3 authors:

Ainara Badiola José Ignacio Canudo Universidad del País Vasco / Euskal Herriko… University of Zaragoza

15 PUBLICATIONS 109 CITATIONS 430 PUBLICATIONS 3,138 CITATIONS

SEE PROFILE SEE PROFILE

Gloria Cuenca-Bescós University of Zaragoza

300 PUBLICATIONS 4,166 CITATIONS

SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate, Available from: José Ignacio Canudo letting you access and read them immediately. Retrieved on: 29 July 2016 Author's personal copy

Cretaceous Research 32 (2011) 45e57

Contents lists available at ScienceDirect

Cretaceous Research

journal homepage: www.elsevier.com/locate/CretRes

A systematic reassessment of multituberculates from Galve (Teruel, Spain)

Ainara Badiola*, José Ignacio Canudo, Gloria Cuenca-Bescós

Grupo -IUCA,1 Paleontología, Facultad de Ciencias, Universidad de Zaragoza, Pedro Cerbuna 12, E-50009 Zaragoza, Spain article info abstract

Article history: This paper includes a systematic reassessment of the Early Cretaceous (late Hauterivianeearly Barre- Received 19 April 2010 mian) multituberculate of Galve (Teruel, Spain), previously studied by Crusafont-Pairó and Adr- Accepted in revised form 1 October 2010 over, and Crusafont-Pairó and Gibert in 1966 and 1976, respectively, as well the study of other Available online 4 November 2010 unpublished specimens found in the revised collection of Institut Català de Paleontologia (ICP). We here include for the first time the emended descriptions and comparisons as well as the SEM photographs of Keywords: all the specimens found in the collection and update the biostratigraphic data that they have provided. Early Cretaceous The multituberculate assemblage from Galve comprises at least four taxa: the paulchoffatiid Gal- Galve veodon nannothus, the eobaatarid Eobaatar hispanicus, another eobaatarid or a possible plagiaulacid, Iberian Peninsula hahni gen. et sp. nov., and the pinheirodontid Lavocatia alfambrensis. The Parendotherium Systematic Review herreroi has been removed from the multituberculate faunal list of Galve. Another two different taxa, Iberica hahni gen. et sp. nov provisionally classified as indet. and or Eobaataridae indet. could be added to the list if their validity can be assessed once more specimen are found. The presence of more than four different “plagiaulacidan” multituberculate taxa in the late Hauterivianeearly of the Galve area suggests a high biodiversity of these faunas in the Early Cretaceous age of the Iberian Peninsula. Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction “plagiaulacidan” multituberculates constitute the most abundant and diverse remains in the mammal fossil assemblages, although other The village of Galve (Central Iberian Range), taxa have also been found, such as “eupantotherian” (dryolestoid, (northeastern Spain), and its surroundings are becoming one of the peramurid) and “symmetrodontan” (spalacotheriid) most important areas of the Iberian Peninsula in the study of Early (Krebs, 1993; Canudo and Cuenca, 1996; Hahn and Hahn, 2002; Cretaceous vertebrate faunas. Many researchers have carried out their Cuenca-Bescós et al., in press). In this paper, we present a system- investigations in this area over the past 50 years (see Ruiz-Omeñaca atic reassessment of Early Cretaceous multituberculate fossils from et al., 2004 and references therein). The uniqueness of Galve is Galve, previously studied by Crusafont-Pairó and Adrover (1966) and based upon its numerous and well-correlated fossiliferous bone- Crusafont-Pairó and Gibert (1976), as well as a study of other beds, in long stratigraphic sequences. Along with the Wealden Group unpublished specimens found in the revised collection of Institut de of the Isle of Wight (United Kingdom), it is one of the two areas where Paleontologia Miquel Crusafont de Sabadell (IPS), Barcelona, now the greatest vertebrate biodiversity has been described in the Early renamed as Institut Català de Paleontologia (ICP). The fossils were Cretaceous of Europe (Martill and Naish, 2001; Ruiz-Omeñaca et al., collected during the 1960s and 1970s in the course of the palae- 2004). The first Spanish mammal fossil was described at ontological survey and sediment-screen-washing campaigns carried Galve (Crusafont-Pairó and Adrover,1966), and the holotypes of many out by German and Spanish teams led by Walter Kühne of the Freie other Early Cretaceous vertebrate fossils have been described as well Universität Berlin and Miquel Crusafont-Pairó of the Universidad de (see next section). As regards mammals, eight Early Cretaceous fossil- Barcelona in the late Hauterivianeearly Barremian and early Barre- bearing beds are today recorded around Galve. Isolated teeth of mian sites of Colladico Blanco and Herrero, respectively. There is no mention in Crusafont-Pairó and Gibert (1976) whethereach specimen comes from the Colladico Blanco or Herrero site. Several synonymies * Corresponding author. Tel.: þ34 976761000x2248; fax: þ34 976761106. have been proposed for the teeth that appear figured in the afore- E-mail addresses: [email protected] (A. Badiola), [email protected] (J.I. Canudo), [email protected] (G. Cuenca-Bescós). mentioned papers (see Hahn and Hahn, 2006 for a review), but their 1 www.aragosaurus.com. descriptions have not been updated, nor have the figurations of all the

0195-6671/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.cretres.2010.10.003 Author's personal copy

46 A. Badiola et al. / Cretaceous Research 32 (2011) 45e57 specimens been published. We here include the emended descrip- (2005); the oospecies turolensis Amo tions and comparisons of these fossils and update the biostratigraphic Sanjuán et al. (2000); and six mammals, including the multi- data they have provided. tuberculates Parendotherium herreroi Crusafont-Pairó and Adrover (1966), Galveodon nannothus Hahn and Hahn (1992), Eobaatar his- 1.1. Abbreviations panicus Hahn and Hahn (1992), and Lavocatia alfambrensis Canudo and Cuenca (1996), the spalacotheriid henkeli Krebs Institutional. ICP, Institut Català de Paleontologia, Sabadell, (1985), and the peramurid Pocamus pepelui Canudo and Cuenca Spain; MPZ, Museo de Paleontología de la Universidad de Zaragoza, (1996). The “eupantotherian” dryolestoid Crusafontia cuencana Zaragoza, Spain; FCPT, Fundación Conjunto Paleontológico de Ter- Henkel and Krebs (1969) is also reported in Galve (Krebs, 1985, uel-Dinópolis, Teruel, Spain. 1993; Martin, 1998). Its holotype was described in the late Barre- Localities. CAN, La Cantalera; CB, Colladico Blanco; CR, Cerrada mian site of Uña (Southwestern Iberian Range, see Buscalioni et al. Roya-Mina; P2, Pelejón-2; PO, Poca; and YH, Herrero. 2008 for a review). In a recent revision of the dryolestoid mammals Dentition. We use capital letters (I, C, P, M) for upper dentition from Galve we allocate the material from Galve to a new Crusafontia and lower-case letters (i, c, p, m) for lower dentition. Cusps are species and relegate C. cuencana to Uña (Cuenca-Bescós et al., in labelled following Kielan-Jaworowska et al. (2004: fig. 8.28, press). The mammal fossil-bearing beds of Galve are included modified from Hahn and Hahn, 1998: fig. 1a): cusps of the labial (b, within the and formations, and the multi- B) and lingual (l, L) rows are numbered mesiodistally, indicating tuberculate fossils come from Pejelón-2 and Colladico Blanco their corresponding number and letter; a capital letter (L, B) in the (upper part of the El Castellar Fm.), and Herrero, Cerrada Roya-Mina upper cheek tooth row and a lower-case letter (l, b) in the lower and Poca (Camarillas Fm.) (Fig. 1). cheek tooth row (e.g., B3: third labial cusp of an upper tooth; l2: The discordantly overlies the Villar de second lingual cusp of a lower tooth). The cusp formula is that Arzobispo Formation; here the majority of the authors had inter- proposed by Kielan-Jaworowska et al. (2004: 279), which is indi- preted a stratigraphic gap from the middle? up to the cated by the number of cusps in consecutive rows given from labial upper (Díaz Molina and Yébenes, 1987; Soria de to lingual, separated by a colon. We also use the corresponding Miguel, 1997; Liesa et al., 2006; Meléndez et al., 2009). However, letter for each side for an easier reading (e.g., 3B:4L). in the lower and middle part of the Castellar Formation there are Measurements. L, Length, measured parallel with the mesiodistal charophytes and pollen that are ValanginianeHauterivian in age axis; W, maximum width, measured between the labial and lingual (Díez et al., 1995; Martín-Closas, 1989; Riveline et al., 1996); margins of the crown. Both are given in mm. therefore, the stratigraphic gap is shorter in time, and it spans only from the Middle Berriasian? to the lower . The El Cas- 2. Geological and palaeontological backgrounds tellar Fm. represents the first syn-rift unit included in the Wealden facies of the Galve Subbasin and has a thickness of roughly 100 m, Roughly one hundred vertebrate fossil-bearing beds have been consisting of a lower part with lutites, , conglomerates found around the village of Galve (see palaeontological sites and and (80 m), which represent alluvial, palustrine and bibliography in http://www.aragosaurus.com). This village, situ- lacustrine subenvironments, and an upper part comprising alter- ated about 60 km from the city of Teruel and 120 km from the city nating marls and limestones (20 m) typical of a lacustrine system in of Zaragoza in , northeastern Spain, is geologically located in phases of expansion and retraction (see Meléndez et al., 2009 for the Galve Subbasin. The latter is one of the seven subbasins of the a review). The charophyte association observed in the carbonated larger Mesozoic basin of Maestrazgo (Fig. 1). The Maestrazgo Basin upper part of the El Castellar Formation, which includes oogonia was structured into smaller sedimentary basins due to the intra- attributed to Atopochara trivolvis triquetra (Grambast, 1968), dates continental extensional tectonic deformation of the Iberian rifting, the top of this formation as upper Hauterivianelowermost Barre- which is related to the spreading of the Tethys and Central Atlantic mian (Martín-Closas, 1989; Schudack, 1989; Riveline et al., 1996). that took place during the Late to Early Cretaceous (see The basal part of the rests conformably on Salas et al., 2001; and Liesa et al., 2006 and references therein). the uppermost part of the El Castellar Formation in the Galve The Galve Subbasin is a NNWeSSE elongated basin, 40 km long Subbasin, whereas in other areas the Camarillas Fm. unconformably and 20 km wide. There is a wide sedimentary succession of Upper overlies the El Castellar Fm. (see Soria de Miguel, 1997 for a review). Jurassic and Lower Cretaceous marine, transitional and continental The Camarillas Fm. comprises lutitic, silty and sandy facies depos- deposits, which range from the Kimmeridgian/ transition ited in a low-sinuosity, multi-channel fluvial system. The palae- up to the lower (e.g., Soria et al., 1995; Canudo et al., 1997; ochannels, with mainly sandy facies, mostly exhibit vertical Liesa et al., 2006). The vertebrate fossil-bearing beds are included aggradation. The lutites, massive silts and sandstones correspond to in four formations: Higueruelas (Kimmeridgian/Tithonian), Villar the floodplains and bank deposits (crevasses) of these fluvial de Arzobispo (upper Tithonian-middle? Berriasian), El Castellar systems. The latter pass upward into a near-shore deltaic plain (Valanginian?-lowermost Barremian) and Camarillas (lower Bar- characterized by lutitic facies with continental as well as marine remian) (See Ruiz-Omeñaca et al., 2004 for a review; Díez et al., fossils (charophytes, pollen, gastropods, ostreids, arenaceous fora- 1995 for the age of the lower part of Castellar Formation). minifera, vertebrates as dinosaur and shark teeth, etc.) and sandy The vertebrate fossils of Galve consist mostly of isolated bones palaeochannels that show sedimentary structures related to the and teeth, although palaeoichnological and palaeoological remains tidal action in the fluvial palaeocurrents (Díaz Molina and Yébenes, are also present. The fossil assemblages comprise sharks, bony 1987). Revision of the charophyte and pollen-spore fossil assem- fishes, amphibians (), squamates, crocodiles, , blages confirms the lower Barremian stratigraphic position of the and mammals (e.g., Estes and Sanchiz, 1982; Díaz Molina and Camarillas Fm. in Galve (Schudack, 1989; Martín-Closas, 1989). Yébenes, 1987; Sanz et al., 1987; Ruiz-Omeñaca et al., 2004; The first Spanish Mesozoic mammal fossil was described in Canudo et al., 2006; Badiola et al., 2009a; Gasca et al., 2009). Galve, in the Herrero site (Crusafont-Pairó and Adrover, 1966: 30), Galve is the type locality of eleven taxa: the shark Lonchidion and corresponds, as mentioned in the previous section, to the microselachos Estes and Sanchiz (1982); the amphibian Galverpeton multituberculate P. herreroi. Later, Crusafont-Pairó and Gibert ibericum Estes and Sanchiz (1982); the dinosaurs Aragosaurus (1976) described further Early Cretaceous multituberculate fossils ischiaticus Sanz et al. (1987), and Galvesaurus herreroi Barco et al. from the Galve area; these correspond to those we revised herein. Author's personal copy

A. Badiola et al. / Cretaceous Research 32 (2011) 45e57 47

Fig. 1. Geographical and geological location of Galve (Teruel, Spain). A simplified geological map of the Iberian Peninsula and the structural map of the Maestrazgo Basin during the Upper JurassiceLower Cretaceous rifting process, with the main depositional area in grey. This rifting structured the Maestrazgo Basin into seven subbasins (, Galve, Las Parras, Morella, Perelló, Salzedella and Peñagolosa). The sites which have yielded multituberculate fossils are indicated in bold type. Modified from Liesa et al. (2006).

Subsequent investigations into Mesozoic mammals of Spain were discoveries (Badiola et al., 2009c). The late Hauterivianeearly Bar- continued by German palaeontologists, who described some Bar- remian fossils from the site of La Cantalera (Oliete Subbasin, Fig. 1) remian mammals. As for multituberculates, a series of publications have already been published (Badiola et al., 2008), whereas those from the 1990s and 2000s by Gerhard and Renate Hahn document from the late Barremian site of Vallipón (Morella Subbasin, Fig. 1) the Early Cretaceous (Barremian) “plagiaulacidans” from the will be published in a paper currently in preparation. Iberian Ranges (see Hahn and Hahn, 2006 and references therein), including those from the Galve Subbasin, the paulchoffatiid G. 3. Systematic palaeontology nannothus and the eobaatarid E. hispanicus (Hahn and Hahn, 1992: 150e158, 2002: 258). The “plagiaulacidan” multituberculate fossil collection of Galve At the beginning of the 1990s the Aragosaurus research team we revise here comprises isolated teeth previously studied by from the Universidad de Zaragoza (Spain) took up the study of the Crusafont-Pairó and Adrover (1966) and Crusafont-Pairó and Gibert Iberian Early Cretaceous mammals, together with other Mesozoic (1976) but also some unpublished specimens found in this collec- vertebrate faunas, in the Central Iberian Range. As regards multi- tion as stated earlier. The fossils come from the upper part of the El tuberculates, the pinheirodontid multituberculate L. alfambrensis Castellar Formation (the Colladico Blanco site) and the base of the was added to the list (Canudo and Cuenca, 1996:218e222). Since Camarillas Formation (the Herrero site) (Fig. 1). The specific origin then, new multituberculate finds have been reported in the abstracts of each isolated tooth, i.e. which of these sites it is from, is not of a number of congresses and meetings (e.g., Badiola et al., 2009a,b; mentioned in the previous study (Crusafont-Pairó and Gibert, Cuenca-Bescós and Canudo, 2004), including an overview of these 1976). Five teeth, referred to as Forma 1 (p4), Forma 5 (P4/5) and Author's personal copy

48 A. Badiola et al. / Cretaceous Research 32 (2011) 45e57

Forma 6 (M2 figured in pl. 3, fig. 1), as well as Forma 2 (P5) and three, with B2 situated more or less at the mid-length of the Forma 3 (I2, figured in pl. 4, fig. 2) in Crusafont-Pairó and Gibert (op. longitudinal labial margin, between its lingual counterparts L1eL2, cit.), and assigned to E. hispanicus and P. herreroi, respectively, by rather than near the mesiolabial corner, and in having a talonid-like Hahn and Hahn (2006), are missing in the revised collection. In this “heel” structure on the distal margin, which is absent in E.? pajar- paper, the rest of the specimens (seventeen teeth) have been onensis. In the latter species a thick distal cingulum is present, but it assigned to at least four different “plagiaulacidan” taxa. does not protrude notably as in Iberica gen. nov. The other “pla- Order: Multituberculata Cope, 1884 giaulacidans” exhibit mesiodistally short and rounded anterior Suborder: Plagiaulacida McKenna, 1971 [pro Plagiaulacoidea upper premolars with a semicircular or subquadrangular crown Ameghino, 1889] outline in occlusal view. Family: Plagiaulacidae Gill, 1872 or Eobaataridae Kielan-Jawor- Included Species. Only the type species. owska, Dashzeveg and Trofimov, 1987 Distribution. Central Iberian Range, northeastern Iberia, the : Iberica gen. nov. Mesozoic Maestrazgo Basin (Oliete and Galve Subbasins). Early Figs. 2 and 3 Cretaceous (late Hauterivianeearly Barremian). 1976 Conjunto II of Forma 4 Crusafont-Pairó and Gibert, p. 59, Iberica hahni sp. nov. pl. 2, fig. 1. Figs. 2 and 3 1992 Parendotherium herreroi Crusafont-Pairó and Adrover, Derivation of name. Dedicated to Gerhard and Renate Hahn for 1966; Hahn and Hahn, p. 154. their contribution to the study of the Early Cretaceous multi- 2006 Eobaatar hispanicus Hahn and Hahn, 1992; Hahn and tuberculates of the Iberian Peninsula. Hahn, pp. 240e243, fig. 233. Holotype. FCPT (CAN 1/936), left P1/3 from the La Cantalera site 2008 Plagiaulacidae or Eobaataridae gen. et sp. indet. Badiola, (Fig. 2: 1). Canudo and Cuenca-Bescós, p. 1462, pl. 3, figs. AeE. Type Locality and Horizon. La Cantalera (Teruel, Spain), in the Remarks. Five P1/3s previously described as conjunto II of Forma 4 Oliete Subbasin, Formation, late Hauterivianeearly Barremian by Crusafont-Pairó and Gibert (1976:61e63, pl. 2, fig. 1), together (Badiola et al., 2008). with a further unpublished P1/3 fragment found in the collection, Other referred specimens. Six P1/3s from the revised Galve and another P1/3 from the La Cantalera site (Iberian Ranges, Oliete collection, coming from the Colladico Blanco site (El Castellar Subbasin, Fig. 1), have been assigned to a new genus and species Formation; late Hauterivianeearly Barremian) and/or the Herrero (Fig. 2). The tentatively referred teeth, a p4 fragment from La Can- site (Camarillas Formation; early Barremian) at Galve (Teruel, talera, which was classified together with the P1/3 from the same Spain), located in the Galve Subbasin: IPS-49920, left P1/3; and IPS- locality as Plagiaulacidae or Eobaataridae gen. et sp. indet. in Badiola 49922, 49923, and 49921, 49924, and 49931, right P1/3s. et al. (2008: pl. 3), and two M2s from the revised Galve collection Measurements. See Table 1. (Fig. 3), seem belong to Plagiaulacidae rather than Eobaataridae but Distribution. As for genus. we cannot state it with absolute certainty. The studied material, Diagnosis. As for genus, monotypic. then, have not been assigned to any of these two families. Description. P1/3 (emended from Crusafont-Pairó and Gibert, Derivation of name. After the name of the Iberian Peninsula. 1976; Badiola et al., 2008): the seven specimens are well Type species. Iberica hahni sp. nov. preserved, though some of them show the cusp B2 partially broken Diagnosis. A new “plagiaulacidan” multituberculate described on (Fig. 2:5e7). By contrast, the roots are not completely preserved in the basis of the P1/3s, characterized by having a rectangular crown in any specimen. The crown is roughly rectangular in shape, some- occlusal view, with labial and distal bulges, and four cusps not what elongated mesiodistally, in occlusal view, with labial and arranged in clear rows (2B:2L). Small B1 displaced towards the distal bulges. All teeth show four cusps (2B:2L), which are not mesiolabial corner of the crown and a thicker B2 confined to the arranged in clear rows. B1 is the smallest cusp, which is confined to longitudinal midpoint of the labial margin, between its lingual the mesiolabial corner of the crown and is not transversally paired counterparts L1 and L2. Rounded mesial margin, with isolated small to its lingual counterpart L1; in most of the specimens, B1 is situ- cuspules present on it and/or a short, transverse ridge encircling it, ated in a more mesial position than L1. B2, by contrast, is similar in and a narrower distal margin, with a talonid-like “heel” structure, size to the lingual cusps and is confined to the longitudinal quite elongated distally, with or without cuspules. Cusps covered by midpoint of the labial margin, between its lingual counterparts L1 few faint enamel radiating ridges. and L2. On the distal margin a narrow and distally elongated talo- Differential diagnosis. Iberica resembles Parabolodon elongatus nid-like “heel” structure is present, which sometimes exhibits (Simpson, 1928) and Eobaatar? pajaronensis Hahn and Hahn, 2001, a large cuspule (Fig. 2: 6). The mesial margin is rounded, with one in its possession of P1/3s with a mesiodistally lengthened crown (a or two isolated small cuspules present on it and/or a short, trans- tooth longer than it is broad), but differs from these species in verse ridge encircling it. All cusps are ornamented with faint, fine having four cusps rather than three, with cuspules on the mesial and short enamel radiating ridges. and distal margins, which are absent in the other two taxa, and in Comparison. The upper anterior premolars in most Early Creta- that both the cusp arrangement and morphology of the distal ceous “plagiaulacidan” multituberculates are tri- to tetra-cuspid margin are clearly different. Iberica gen. nov. resembles P. elongatus teeth, with or without additional small cuspules, and they are in its possession of P1/3s with a talonid-like “heel” structure on the usually short and rounded, with a semicircular or subquadrangular distal margin, but this structure is distally much more protruding in crown outline in occlusal view (e.g., Hahn and Hahn, 2004: 124, fig. Iberica gen. nov., and in the P1/3s of the latter taxon L1 is confined 5) instead of being rectangular and elongated mesiodistally like the to the mesiolingual corner and B2 is situated between its lingual P1/3s under study and the P1/3s of E.? pajaronensis and P. elongatus counterparts L1 and L2. By contrast, in P. elongatus L1 is situated described in the late Barremian site of Pie Pajarón, in the province almost at (P1) or at (P2eP3) the transversal midpoint of the mesial of Cuenca, Spain (Southwestern Iberian Range, see Buscalioni et al., margin and B2 corresponds in transverse position to its lingual 2008 for a review), and in the Berriasian of Dorset, south England counterpart L2 in P2eP3. Moreover, P. elongatus is a much larger (Hahn and Hahn, 2001: figs. 1 and 2; and Hahn and Hahn, 2004: fig. species than Iberica gen. nov. The size of E.? pajaronensis is similar 5b), respectively. The studied P1/3s, however, differ from the P1/3s to that of Iberica gen. nov., but the latter taxon differs from E.? of the afore-mentioned two species (see differential diagnosis). The pajaronensis in having P1/3s, apart from four-cusped instead of cusp number and their disposition along the occlusal surface in the Author's personal copy

A. Badiola et al. / Cretaceous Research 32 (2011) 45e57 49

Fig. 2. SEM photographs of P1/3s of Iberica hahni gen. et sp. nov. (Plagiaulacidae or Eobaataridae) from the Early Cretaceous of La Cantalera (1) and Galve (2e7), Central Iberian Range, in Teruel, Spain. 1e2, left P1/3s [1: FCPT (CAN 1/936); 2: IPS-49920]. 3e7, right P1/3s (3: IPS-49922; 4: IPS-49923, 5: IPS-49921; 6: IPS-49924; 7: IPS-49931). a, occlusal; b, mesial; c, distal; d, lingual; and e, labial views. studied premolars have not yet been described in any other “pla- (Ejemplar 1 of Forma 6, Crusafont-Pairó and Gibert, 1976:63e64, giaulacidan” multituberculates (see diagnosis), and they have been pl. 3, fig. 2); IPS-49927, right M2 (Fig. 3, Table 1). assigned to a new taxon. p4: as at least a quarter part of the mesial margin is broken off, Tentatively referred material. FCPT (CAN 1/937), fragment of the original length of the tooth as well as the number of cusps and a right p4 from the La Cantalera site, which corresponds in size to serrations on the cutting edge are unknown (Fig. 3: 1). The seven P1/3 from the same locality (Badiola et al., 2008: pl. 3), and two preserved serrations are roughly of the same length, are equally other M2s found in the revised Galve collection: ISP-49925, left M2 spaced, and bent similarly downwards. The crown seems to be Author's personal copy

50 A. Badiola et al. / Cretaceous Research 32 (2011) 45e57

Fig. 3. SEM photographs of the teeth tentatively assigned to Iberica hahni gen. et sp. nov. (Plagiaulacidae or Eobaataridae) from the Early Cretaceous of La Cantalera (1) and Galve (2e3), Central Iberian Range, in Teruel, Spain. 1, fragment of a right p4 [FCPT (CAN 1/937)]. 2, left M2 (IPS-49925). 3, right M2 (IPS-49927). a, occlusal; b, mesiolabial; c, distal; d, lingual; e, labial; and f, mesial views. higher than long, or as high as it is long, but the original length is The presence of more than four cusps, but fewer than 15, on the not preserved to confirm this character. The cutting edge, despite cutting edge and the fact that the presence of long serrations of the fragmentary nature of the tooth, seems not to be highly arched. similar length and equally spaced, which cover nearly the upper half The absence of horizontal erosion on the cutting edge indicates that of the crown, allow us to assign it to Eobaataridae or Plagiaulacidae the premolar could have been used for cutting, instead of for as opposed to Paulchoffatiidae, Pinheirodontidae, Allodontidae, shearing. A prominent and dorsally worn shelf-like concave Zofiabaataridae, Glirodon Engelmann and Callison, 1999, or Argin- structure (sensu Kielan-Jaworowska et al., 1987, p. 6) is located near baataridae (Badiola et al., 2008). The morphology of the p4 fragment the base of the distolabial corner, protruding to a notable degree seems to be closer to a p4 of the Plagiaulacidae than of the Eobaa- labially over the surface of the tooth. This worn structure possibly taridae. The eobaatarid p4s tend to be more highly arched than in the occupies roughly a quarter of the distolabial corner of the crown. Plagiaulacidae and the worn, shelf-like concave structure present at Author's personal copy

A. Badiola et al. / Cretaceous Research 32 (2011) 45e57 51

Table 1 Dental measurements (in mm) of the studied “plagiaulacidan” multituberculates from the Early Cretaceous (late Hauterivianeearly Barremian) sites of Galve and La Cantalera (Teruel, Spain). Teeth abbreviated to CAN come from the La Cantalera site and the rest of the specimens with IPS abbreviation correspond to the fossil collection of Galve we here revise, previously studied by Crusafont-Pairó and Gibert (1976).

TAXON MATERIAL L W Iberica hahni gen. et sp. nov. FCTP (CAN 1/936) P1/3 1.06 0.68 IPS-49920 P1/3 0.92 0.61 IPS-49921 P1/3 1.1 0.78 IPS-49922 P1/3 1.02 0.76 IPS-49923 P1/3 1.06 0.77 IPS-49924 P1/3 1.26 0.8 IPS-49931 P1/3 c. 1.06 e Tentatively assigned to I. hahni gen. et sp. nov. FCTP (CAN 1/937) p4 >2.26 1.24 IPS-49925 M2 1.87 1.86 IPS-49927 M2 1.88 1.6

Eobaatar hispanicus IPS-49918 P1/3 1.5 1.33 IPS-49919 P1/3 1.39 1.13 Tentatively assigned to E. hispanicus IPS-49929 ?P4 >1.48 >1.07 IPS-49930 ?P4 >1.34 >0.89

Plagiaulacidae or Eobaataridae indet. IPS-49917 P1/3 1.31 0.84 Paulchoffatiidae indet. IPS-49916 I2 1.39 0.99 IPS-49928 I3 c. 1.31 c. 1.13

“Plagiaulacida” indet. IPS-49915 I2 >>1.06 >>0.75 IPS-49926 M2 >1.98 >1.47

the base of the distolabial corner in the p4 from La Cantalera is first lingual cusp has caused the appearance of two tips that are horizontally longer than that of eobaatarids in which this worn slightly developed apically, which do not appear separated from structure is known. Nevertheless, in the Plagiaulacidae in which p4 each other by a groove in lingual view. Ornamentation is only is known, the shelf-like structure is either much more extended present on the labial edge of the mesiolabial shelf and seems to horizontally (Plagiaulax Falconer, 1857, and Plioprion Cope, 1884), consist of a few small grooves (Fig. 3: 2). The other specimen (Fig. 3. occupying two-thirds of the labial wall, or this worn structure is 3) is quite abraded, but corresponds in size to the other molar and much shorter and less protruding labially (Bolodon Owen, 1871). A seems to show similar morphological features. shelf-like structure that protrudes more markedly labially than in The crown outline, the cusp formula and the development of the the afore-mentioned plagiaulacids is also present in a p4 fragment mesiolabial shelf or wing of the M2s under study are closer to the found in the Berriasian deposits of Dorset, classified in an open Eobaataridae or Plagiaulacidae than to other “plagiaulacidans” (e.g., nomenclature as Plagiaulacidae indet. (Kielan-Jaworowska and Hahn and Hahn, 2004: 132, fig. 10). The M2 of the Paulchoffatiidae Ensom, 1994, pl. 2, figs. 6 and 7). However, in the British p4 this is characterized by having a high number of cusps (2e4:3e7), structure is horizontally longer than in the p4 from La Cantalera. The mainly in the lingual row, which is considerably longer than the British p4 also seem to have a crown that is much longer than high labial row. The mesiolabial shelf is absent in Allodontidae, in the with a straighter dorsal margin on the cutting edge. Other main pinheirodontid Iberodon Hahn and Hahn, 1999, and in the argin- feature that is used to differentiate the p4s of eobaatarids and pla- baatarid Ameribaatar Eaton and Cifelli, 2001, and it is much less giaulacids is the number of basal cusps, with more than one in the developed in other representatives of Arginbaataridae and Pin- Plagiaulacidae and just one in the Eobaataridae. In the p4 in ques- heirodontidae. In Glirodon it is present, but a clear third row of fully tion, however, the presence of only one or more than one cusp developed cusps appears in this region, which is absent in the Galve cannot be certainly assessed. With the available data, then, the p4 specimens. The cusp formula of the M2s under study (2B:3L) is could belong to either Eobaataridae or Plagiaulacidae. different from that of eobaatarids. In most M2s of the Eobaataridae M2: the M2s under study are roughly trapezoidal in shape in there are usually 3 cusps, rather than two, present on the labial row occlusal view, with the mesial margin wider than the distal (Fig. 3: (Eobaatar Kielan-Jaworowska, Dashzeveg and Trofimov, 1987, and 2e3). The mesial margin is slightly sigmoid, not straight, and the Sinobaatar Hu and Wang, 2002) and/or a strongly marked ridge mesiolabial portion of the crown is broadened, forming a fully without (Eobaatar) or with an additional row of cusps (row bb) developed shelf-like extension, external to the first labial cusp (the present in the mesiolabial shelf (Sinobaatar, Monobaatar Kielan- antero-buccal shelf sensu Hahn and Hahn, 2004: p. 130; or antero- Jaworowska, Dashzeveg and Trofimov, 1987). The M2 of the lateral wing sensu Kielan-Jaworowska et al., 1987: p. 13). Along the eobaatarid Loxaulax Simpson, 1928, shares with the M2s under mesial margin a narrow rim is present, which at the midpoint of its study the presence of two labial cusps and the absence of a sharp trajectory exhibits a mesio-distal widening and passes onto the first ridge or an additional row bb on the mesiolabial shelf, but differs labial cusp reaching its tip, and continues laterally, surrounding the from the Galve molars in having four lingual cusps rather than mesiolabial shelf structure. The cusp formula is 2B:3L in both teeth, three, a sharp sigmoid mesial margin, and a more ornamented with the labial row somewhat shorter than the lingual one. The two mesiolabial shelf structure, with irregular comma-shaped grooves rows are separated by a deep longitudinal groove. The first labial and pits. The studied M2s resemble the only M2 known in Pla- cusp forms a longitudinal ridge, whereas the second labial cusp giaulacidae, Bolodon osborni Simpson, 1928, in possessing the same exhibits a more bulbous shape in occlusal view. The two cusps are cusp formula and crown outline, with a slightly sigmoid mesial separated by a short, deep transverse groove and coalesce at their margin, but differ from the British species in their greater size and bases. In the lingual row, the cusps are mesiodistally more elon- in having more crescentic and mesiodistally elongated lingual gated than those of the labial row, and they are separated by weak cusps, which are not separated by deep transverse grooves. In B. transverse grooves in lingual view (Fig. 3: 2d, 3d). The wear of the osborni, the lingual cusps exhibit a more bulbous morphology and Author's personal copy

52 A. Badiola et al. / Cretaceous Research 32 (2011) 45e57

Fig. 4. SEM photographs of the dentition of Eobaatar hispanicus and Plagiaulacidae or Eobaataridae indet. from the Early Cretaceous of Galve, Central Iberian Range, in Teruel, Spain. 1e2, Eobaatar hispanicus [1: right P1/3 (IPS-49919); 2: left P1/3 (IPS-49918)]. 3e4, two?P4s tentatively assigned to E. hispanicus [3: left?P4 (IPS-49929); 4:?right?P4 (IPS-49930)]. 5, Plagiaulacidae or Eobaataridae indet. (IPS-49917, right P1/3). a, occlusal; b, mesial; c, distal; d, lingual; and e, labial views. they are separated from each other by deep transverse grooves Fig. 4:1e4 which coalesce at their bases (e.g., Hahn and Hahn, 2004: fig. 10c; 1976 Conjunto Ib of Forma 4 Crusafont-Pairó and Gibert, pp. Kielan-Jaworowska and Ensom, 1992: pl. 3, figs. 7e8). The M2s 61e63. under study then seem belong to Plagiaulacidae rather than Remarks. Two P1/3s described as conjunto Ib of Forma 4 by Eobaataridae. Nevertheless, due to scantiness of M2s in the Pla- Crusafont-Pairó and Gibert (1976:61e63) correspond roughly in giaulacidae (only known in B. osborni) the interspecific variation of size to a P1/3 described by the same authors as conjunto Ia of Forma these two groups using M2s cannot be properly described. 4 (op. cit.: pl. 1, fig. 2). However, like Crusafont-Pairó and Gibert The referred teeth (P1/3s) to I. hahni gen. et sp. nov. as well as (1976) we also differentiate the two groups because of their those tentatively assigned (p4 and two M2s) to this taxon suggest different crown outline and cusp arrangement; the P1/3s of con- that we cannot state its assignation to either Plagiaulacidae or junto Ib have here been assigned to E. hispanicus (Fig. 4:1e4), Eobaataridae with absolute certainty. The studied material then has whereas that of conjunto Ia (Fig. 4: 5) has been classified in an open not been included into any of these two families. nomenclature as Plagiaulacidae or Eobaataridae indet. (see below). Family: Eobaataridae Kielan-Jaworowska, Dashzeveg and Trofi- Other two unpublished upper posterior premolars (probably P4) mov, 1987 are also here tentatively assigned to E. hispanicus. Genus: Eobaatar Kielan-Jaworowska, Dashzeveg and Trofimov, Referred specimens. IPS-49918, left P1/3; and IPS-49919, right 1987 P1/3. Eobaatar hispanicus Hahn and Hahn, 1992 Measurements. See Table 1. Author's personal copy

A. Badiola et al. / Cretaceous Research 32 (2011) 45e57 53

Description. P1/3 (emended from Crusafont-Pairó and Gibert, the three afore-mentioned P1/3s from Galve. B2 in the P1/2s from 1976: 61): three-cusped teeth (1B:2L), which are subequal in size. Uña is situated between its lingual counterparts L1eL2, rather than B1 is confined to near the mesiolabial corner, whereas L2 is situated near the mesiolabial corner, and due to the presence of much almost obliquely, midway along the distal margin of the crown. A thinner cingula on the distal and labial margins, the teeth exhibit wide and quite protruding distal cingulum, with a flat shelve shape, a subtriangular crown outline in occlusal view instead of a sub- is present, which continues along the labial margin, quite enlarged quadrangular outline like in the Galve specimens. It is difficult to laterally. Because of the enlargement of the distal and labial cingula, assess whether these differences are intraspecific or interspecific the teeth show a subquadrangular crown outline in occlusal view. given the variation in the isolated specimens available from Galve One of the specimens, IPS-49918 (Fig. 4: 2), is quite abraded. In the and Uña. Nevertheless, in many eobaatarids in which the P1eP3 better-preserved tooth, IPS-49919 (Fig. 4: 1), thin enamel radiating series is preserved in the maxilla, such as Monobaatar mimicus, ridges are present on the cusps and the tips of the cusps are very Hakusanobaatar matsuoi Kusuhashi, 2008, and Sinobaatar (Kielan- slightly worn. There are no cuspules present in either of the studied Jaworowska et al., 1987:16;Kusuhashi, 2008: fig. 3; Kusuhashi teeth. The two roots are partially preserved; they seem quite long et al., 2009: figs. 4, 7, 17), the three-cusped anterior upper premo- and robust. IPS-49918 is larger in size than IPS-49919. The larger lars exhibit the same cusp arrangement and the cingula are more or tooth would be a more anterior premolar (P1 or P2), as occurs in the less similarly developed. In E. magnus, by contrast, the isolated P1eP3 series of many “plagiaulacidan” multituberculates (e.g., anterior premolars with a subtriangular outline in occlusal view are Bolodon, Parabolodon Hahn and Hahn, 2004, Hakusanobaatar interpreted as corresponding to P1 and/or P2, whereas those with Kusuhashi, 2008, and Sinobaatar; see Simpson, 1928:45e48; and a more subquadrangular crown, due to greater enlargement of the Kusuhashi et al., 2009: table 2). distal and labial cingula, are considered to be P3. Nevertheless, Comparison. The premolars under study resemble the P1/3s of there is a less difference in the cingula development among these the Plagiaulacidae and Eobaataridae in having mesiodistally short, P1/3s of E. magnus and the disposition of cusps is similarly arranged three-cusped teeth, with a triangular to subquadrangular outline in in all the P1/3s of this species (Kielan-Jaworowska et al.,1987: pls. 3, occlusal view, and share with the P1/3s of B. osborni, Eobaatar 12). The assignation of the isolated P1/2s from the late Barremian magnus, E. hispanicus and various species of Sinobaatar the pres- site of Uña to E. hispanicus could thus be questionable. ence of distinct crescentic cingula surrounding the labial and distal Tentatively referred specimens. IPS-49929, fragment of a left?P4; margins (e.g., Kielan-Jaworowska et al., 1987: pls. 3, 12; Hahn and and IPS-49930, fragment of a?right?P4. Hahn, 2002: fig. 1; Kusuhashi et al., 2009: i.e. figs. 7, 14). The two ?P4: neither of the teeth is completely preserved, and they are P1/3s of the studied collection, however, differ from the P1/3s of B. both quite abraded. The roots are partially preserved in one of osborni and E. magnus in lacking a minute cuspule, situated at them, and these seem robust (Fig. 4: 4). In the better-preserved roughly mid-length along the mesial margin in the first species, tooth (Fig. 4: 3) two rows of cusps seem to exist, with a 3B:4L cusp respectively in front of B1 in the second species, and in having formula. The first two labial cusps are subequal in size, whereas the a wider and distally more protruding distal cingulum and the labial third one is much smaller than B1eB2 and is confined to the dis- cusp displaced into a more mesial position (e.g., Kielan-Jaworowska tolabial corner of the crown. The lingual cusps increase in size et al., 1987: pls. 3, 12). In B. osborni and Bolodon crassidens Owen, distally. A tiny cuspule is situated at the mid-length of the mesial 1871, B1 is situated near or at the distolabial corner, level with its margin, and another two seem to exist at the distal end of each of lingual counterpart L2. Moreover, the cusps in B. crassidens are the lingual and labial cusps row. A rounded mesiolabial wing is ornamented by more conspicuous and thicker enamel radiating present lateral to B1 and B2, quite extended labially. The lingual ridges than in the P1/3s under study. The all premolars in the wall is incomplete in the better-preserved tooth, but none of the P1eP3 series in the species of Sinobaatar exhibits a similar trian- lingual cusps seems to be partially or completely worn (Fig. 4: 3). gular shape rather than a subquadrangular shape in the last ones in Few but conspicuous enamel radiating ridges are present on the occlusal view due to the presence of only the distal cingulum, less abraded cusps. which is much less distally projected than in the Galve specimens. The absence of a third cusp row bb on the studied premolars The two P1/3s under study resemble the P1/3 fragment from allows us to discard their assignation to the Paulchoffatiidae. In another late Hauterivianeearly Barremian locality at Galve many other “plagiaulacidans” whose P4 is known, such as Pla- described by Hahn and Hahn (2002: fig.1) as E. hispanicus, in having giaulacidae, Allodontidae, Pinheirodontidae and Eobaataridae, the a similar cusp arrangement, with the labial cusp confined to near same cusp formula (3B:4L) is present. The P4 of the Plagiaulacidae the mesiolabial corner and L2 situated almost at the midpoint of the differs from the specimens under study in having an eroded lingual distal margin of the crown, and in having a wide and quite wall that slopes down deeply and all the cusps partially or protruding distal cingulum that continues along the labial margin, completely eroded. Moreover, the occlusal surface on the P4 of the quite enlarged laterally. The latter features cannot be confirmed in plagiaulacid Bolodon is much narrower between the longitudinal the specimen described by Hahn and Hahn (op. cit.), because it is midpoint and the distal portion of the labial margin (Kielan- partially broken labially, but a subquadrangular outline of the tooth Jaworowska et al., 1987: pl. 12) than in the premolars under in occlusal view can be assessed. Accordingly, we assign the two P1/ study, and in B. crassidens the third labial cusp is replaced by three 3s of the Galve collection under study, then, to E. hispanicus. The small cuspules, unlike in the Galve premolars. In the allodontid P4 size of the P1/3 fragment studied by the afore-mentioned authors is the mesolabial wing is not present, while in the Pinheirodontidae, probably smaller than the other two P1/3s from Galve. The size the third cusp row bb is absent as in the studied teeth, but the cusp difference, however, can be related to the tooth position in the formula is 2:4 rather than 3:4, or B3 is only incipiently developed. P1eP3 series, as mentioned previously; the bigger premolars The P5 and P4 in the Eobaataridae exhibit a similar crown outline in would be positioned in an anterior position, while the smaller one occlusal view, although they can be differentiated from each other likely belongs to P2 and/or P3. by the number of labial cusps and the shape of the distal margin: Two P1/2s from the late Barremian site of Uña, province of both P4 and P5 teeth exhibit four lingual cusps, but three labial Cuenca, Spain (Southwestern Iberian Range; see Buscalioni et al., cusps are present on P4 and only two on P5, and the distal portion 2008 for a review) were also assigned to E. hispanicus by Hahn of the crown in P5 is distinctly narrowed, consisting of only the and Hahn (1992: 156, figs. 7e8), but these teeth show both lingual cusp row. The premolar fragments from Galve, then, prob- a clearly different cusp arrangement and cingula development from ably belong to P4. Author's personal copy

54 A. Badiola et al. / Cretaceous Research 32 (2011) 45e57

Fig. 5. SEM photographs of the “plagiaulacidan” multituberculates dentition from the Early Cretaceous of Galve, Central Iberian Range, in Teruel, Spain, classified in an open nomenclature as Paulchoffatiidae indet. (1e2) and “Plagiaulacida” indet. (3e4). 1, right I2 (IPS-49916). 2, fragment of a left I3 (IPS-49928). 3, highly abraded I2 (IPS-49915). 4, highly damaged right M2 (IPS-49926). a, occlusal; b, mesiolabial; c, mesiolingual; d, distolabial; and e, distolingual views.

The eobaatarids Eobaatar, Monobaatar and some species of between its lingual counterparts L1eL2, rather than the mesiolabial Sinobaatar have the same cusp formula 3B:4L in P4 as the specimens corner; and thick and prominent enamel radiating ridges on the under study. The latter differ from the species Sinobaatar in lacking cusps. By contrast, in the other two P1/3s of Conjunto Ib herein an asymmetrical crown outline in occlusal view, and having an assigned to E. hispanicus the cusps show much thinner and less oblique distal margin due to the distal portion of the crown being conspicuous enamel radiating ridges (Fig. 4: 1, 2 and 5). distinctly narrowed (Kusuhashi et al., 2009: 1266); they differ from P1/3 with a somewhat mesiodistally elongated crown is only the P4 of Monobaatar in lacking the completely eroded lingual row present in the species mentioned in the previous section: Eobaatar? like the P4 of the Plagiaulacidae. The?P4s of the studied collection pajaronensis, P. elongatus and I. hahni gen. et sp. nov. The P1/3 of the are similar to those of Eobaatar, while differ from the P4 of E. magnus latter, however, is four-cusped rather than tricuspid and show in having a narrower and less laterally protruding mesiolabial wing clearly different development of cingula and enamel radiating lacking cuspules. The?P4s from Galve correspond in size with the P1/ ridges. The P1/3 of Conjunto Ia has thick and prominent enamel 3s we have here assigned to E. hispanicus and are thus tentatively radiating ridges on the cusps and the crown is bordered by a sharp referred to the same species. cingulum, unlike in the P1/3s of I. hahni gen. et sp. nov. P1/3 of Plagiaulacidae or Eobaataridae indet. Conjunto Ia also differs from the other two species with P1/3s tri- Fig. 4:5 cusped in having B2 situated at the longitudinal midpoint of the 1976 Conjunto Ia of Forma 4 Crusafont-Pairó and Gibert, pp. labial margin, between its lingual counterparts L1eL2, rather than in 61e63, pl. 1, fig. 2. the mesiolabial corner as in E.? pajaronensis, or almost at (P1) or at Referred specimen. IPS-49917, right P1/3. the distolabial corner, situated in transverse position to its lingual Measurements. See Table 1. counterpart L2 (P2eP3). Moreover, in P. elongatus L1 is situated Description (emended from Crusafont-Pairó and Gibert, 1976: almost at (P1) or at (P2eP3) the transversal midpoint of the mesial 61e63). P1/3: The premolar under study shows an isosceles trian- margin and a talonid-like “heel” structure is present on the distal gular outline with the cusps confined to the vertex and a 1B:2L cusp margin unlike in the P1/3 of Conjunto Ia. Because of the scantiness of formula. The crown is slightly mesiodistally elongated, longer than the sample, the anterior upper premolar under study is here wide. B1 is confined to the longitudinal midpoint of the labial provisionally described as Plagiaulacidae or Eobaataridae indet. margin, between its lingual counterparts L1 and L2. The crown is Paulchoffatiidae indet. bordered by a thick conspicuous cingulum, but is not greatly Fig. 5:1e2 widened laterally. The enamel radiating ridges are thick and are 1976 Conjunto III of Forma 3 Crusafont-Pairó and Gibert, pp. strongly marked on the cusps. 60e61. Comparison. It was described as Conjunto Ia and included, Referred specimens. IPS-49916, right I2; and IPS-49928, fragment together with two other P1/3s referred to as Conjunto Ib, within of a left I3. Forma 4 (Crusafont-Pairó and Gibert, 1976:61e63). Both groups Remarks. Three I2s were described as Conjunto I, II, and III of (Conjunto Ia and Conjunto Ib) have been here also determined as Forma 3 in Crusafont-Pairó and Gibert (1976:60e61), but only belonging to different taxon. The P1/3 of Conjunto Ia differs from Conjunto III has been identified in the revised collection (Fig. 5: 1). the other two P1/3s of Conjunto Ib in having a somewhat mesio- Conjunto I has not been found, and the highly abraded I2 may distally elongated crown; a different labial cusp arrangement, as B1 belong to Conjunto II (Fig. 5: 3). The latter is here classified in an is confined to the longitudinal midpoint of the labial margin, open nomenclature as “Plagiaulacida” indet. (see below). Author's personal copy

A. Badiola et al. / Cretaceous Research 32 (2011) 45e57 55

Measurements. See Table 1. Description. I2 (emended from Crusafont-Pairó and Gibert, 1976: 60): there is a thick, pyramidal cusp on the mesial margin and four smaller cusps on the distal margin, arranged in a transverse ridge that encircles all the distal margin of the crown. Two thick distal cusps are situated at the middle of the ridge, while the other two, which are much smaller in size, are confined to the lingual and labial sides of the ridge. From the main mesial cusp two crests run ante- riorly and posteriorly down towards the labial and lingual margins and are linked with the afore-mentioned two small cuspules. The two crests exhibit roughly similar length, though the anterior crest is more worn than the posterior one. A small cusp seems to stick out at roughly the midway point of this worn crest (Fig. 5: 1a,b). The lingual wall of the tooth is roughly straight, though with some concave areas on both sides of the two rims, which are located at the middle and at the mesiolabial corner of the wall, respectively (Fig. 5: 1a,e). The labial wall, by contrast, is convex. I3: this is smaller than I2 and shows a subcircular or slightly subquadrate crown outline in occlusal view. The single root seems not to overhang the crown. The tooth has at least three cusps; two are situated mesiodistally at the midpoint of the crown, slightly separated from each other apically and linked by the corresponding crest to the base of the mesial and distal margins of the crown, while the third cusp, similar in size to the other two, is situated almost at the mesiodistal midpoint of the labial margin. A fourth cusp close to the base of the crown at its broken mesial margin would also be present. A prominent shelf-like concave structure encircles the lingual margin, the lingual surface of which is convex. The two afore- mentioned central cusps are relatively smooth on their lingual surface, while their labial surface is ornamented by thick and quite prominent enamel ridges. The third labial cusp also shows a lingual surface similarly ornamented by conspicuous enamel ridges. Comparison. The presence of more than one distal cusp on the distal margin linked in a transverse ridge allows us to assign the I2 under study to Paulchoffatiidae, as opposed to the other “plagiau- lacidans”. In Eobaataridae, Plagiaulacidae and Allodontidae, I2 bears one large, mesially placed cusp, and only one small distal cuspule is present immediately in line behind the main cusp. By contrast, the I2 under study is multicuspid on the distal margin. A multicuspid distal margin is also present in I2 of the Pinheir- odontidae, but the cusps are different in size and are arranged in semi-circles, distributed irregularly along the distal margin rather than linked by a transverse ridge. In Glirodon the morphology of I2 is clearly different (e.g., Hahn and Hahn, 2004: fig. 2b). The pres- ence of more than one well-differentiated cusp in the main ridge Fig. 6. Litho- and Chronostratigraphy of the Early Cretaceous “plagiaulacidan” multi- e and 0 3 mesiolabial and distolingual cusps allows us to assign the tuberculate fossil record from Galve (Teruel, Spain). See Fig. 1 and in Ruiz-Omeñaca studied I3 fragment to Paulchoffatiidae or Pinheirodontidae. et al. (2004: fig. 3) for the geological context of the sites, and Ruiz-Omeñaca et al. Morphologically I3 is also closer to the first family than to the (2004: fig. 4) and references therein for details of the rest of the vertebrate fossils second. The third upper incisor in the Pinheirodontidae bears and their distribution. Type locality of the taxa is indicated by white box, whereas other localities where these taxa also have been found are indicated by black box; the a greater number of cusps along the main ridge as well as on the grey colour box indicates one of the origin localities, whether from Colladico Blanco or labial and lingual margins of the crown than in the Paulchoffatiidae Hererro from Galve, or both localities, of the taxa we here describe on the basis of the (e.g., Hahn and Hahn, 2004: fig. 3cef). Both incisors, then, seem to revised Galve collection. The specific origin of the studied fossils is not mentioned in belong to the family Paulchoffatiidae. their original description (Crusafont-Pairó and Gibert, 1976). See locality abbreviations in the Introduction section. The size of the I2 under study is closer to that of P. herreroi than to the I2 of the other paulchoffatiid species from Galve, the tiny G. nannothus. P. herreroi, however, is only known to be represented by Paulchoffatiidae, we classify the I2 and I3 under study in an open the lost holotype (an I2 described by Crusafont-Pairó and Adrover, nomenclature as Paulchoffatiidae indet. 1966: 30) and should perhaps be considered nomem dubium.As “Plagiaulacida” indet. mentioned earlier, the other teeth subsequently assigned to P. Fig. 5:3e4 herreroi by Hahn and Hahn (2006:151,figs. 108, 109), such as an I2 Referred specimens. IPS-49915, highly abraded I2; and IPS- of Conjunto I of Forma 3 (Crusafont-Pairó and Gibert, 1976: 60, plate 49926, highly abraded right M2. 4, fig. 2) and a P4/5 of Forma 2 (Crusafont-Pairó and Gibert, op. cit. Remarks. The I2 shows similar morphological features to those 59, plate 2, fig. 2), have not been found in the revised collection. described by Crusafont-Pairó and Gibert (1976:60e61) for Conjunto Because of the scantiness of the sample and the similarities II of Forma 3, but these authors did not mention its state of pres- between the upper incisors of the different representatives of the ervation and we do not know whether it belongs to Conjunto II or to Author's personal copy

56 A. Badiola et al. / Cretaceous Research 32 (2011) 45e57 an unpublished specimen. The same question applies to M2. Three Acknowledgments M2 specimens were described in Crusafont-Pairó and Gibert (1976: 63e64) as Forma 6, and these authors mentioned that one of them The comments of the editor (Malcolm Barrie Hart) and of was highly abraded. anonymous referees have improved the quality of the paper. The Description and comparison. Both teeth are highly abraded. In the latter forms part of the project CGL2007-62469, subsidized by the I2 only the main mesial cusp and two other, smaller cusps are Ministry of Science and Innovation of Spain, the European Regional preserved on the distal margin; these seem not to be linked by Development Fund, the Government of Aragon (“Grupos Con- a transverse ridge. M2 is larger than the other two M2s described as solidados” and “Dirección General de Patrimonio Cultural”). The Forma 6 by Crusafont-Pairó and Gibert (1976:63e64) and tenta- former Institut de Paleontologia Miquel Crusafont de Sabadell (IPS), tively assigned here to I. hahni gen. et sp. nov. The two teeth studied now Institut Català de Paleontologia (ICP), loaned us the studied in this section are too damaged to make reliable descriptions and specimens of the ancient Crusafont collection. Ainara Badiola comparisons, and they have been classified as “Plagiaulacida” indet. acknowledges support from the Programa Juan de la Cierva of the Ministry of Education and Science. Rupert Glasgow revised the 4. Discussion and conclusions English Grammar.

Four species of “plagiaulacidan” multituberculates have previ- References ously been mentioned from the late Hauterivianeearly Barremian of the Galve area: the paulchoffatiids G. nannothus and P. herreroi, Ameghino, F., 1889. Contribución al conocimiento de los mamíferos fósiles de la República Argentina. In: Actas Academia Nacional de Ciencias, Córdoba, Buenos the eobaatarid E. hispanicus and the pinheirodontid L. alfambrensis Aires, 6 1e207. (Crusafont-Pairó and Adrover, 1966; Hahn and Hahn, 1992; Canudo Amo Sanjuán, O., Canudo, J.I., Cuenca-Bescós, G., 2000. First record of elongatooli- and Cuenca, 1996). In this paper, P. herreroi has been removed from thid eggshells from the lower Barremian (Lower Cretaceous) of Europe (Cuesta Corrales 2, Galve basin, Teruel, Spain). In: Bravo, A.M., Reyes, T. (Eds.), First the multituberculate faunal list of Galve and another taxon has International Symposium on Dinosaur Eggs and Babies, pp. 7e14. Extended been added to the Galve multituberculate faunal list: I. hahni gen. et Abstracts. sp. nov. The holotype of the latter species has been described from Badiola, A., Canudo, J.I., Cuenca-Bescós, G., 2008. New multituberculate mammals another locality from the Central Iberian Range, La Cantalera from the Hauterivian/Barremian transition of Europe (Iberian Peninsula). Palaeontology 51, 1455e1469 [Corrigendum: Palaeontology 52, 271]. (Maestrazgo Basin, Oliete Subbasin), late Hauterivianeearly Bar- Badiola, A., Canudo, J.I., Cuenca-Bescós, G., 2009a. Systematic reassessment of Early remian in age. This new taxon is provisionally included either into Cretaceous multituberculates from Galve (Teruel, Spain). Journal of Vertebrate Plagiaulacidae or Eobaataridae. The other isolated teeth of the Paleontology 29 (Suppl. 3), 57A. Badiola, A., Canudo, J.I., Cuenca-Bescós, G., 2009b. Plagiaulacida multituberculates revised collection have been assigned to E. hispanicus and two other from the late Barremian-early Aptian of Teruel (Spain). In: Buscalioni taxa; due to the scantiness of the sample they have been classified Delgado, A., Frenegal Martínez, M. (Eds.), Tenth International Symposium on provisionally as Paulchoffatiidae indet. and Plagiaulacidae or Mesozoic Terrestrial Ecosystems and Biota, Teruel (Spain), pp. 157e158. Abstract volume. Eobaataridae indet., respectively. The Early Cretaceous multi- Badiola, A., Canudo, J.I., Cuenca-Bescós, G., 2009c. New Early Cretaceous multi- tuberculate fossil assemblage of Galve then is diverse, consisting of tuberculate fossils from the Iberian Peninsula. In: Godefroit, P., Lambert, O. at least four taxa, such as, G. nannothus, E. hispanicus, L. alfam- (Eds.), DarwineBernissart Meeting, Brussels, February 9e13, 2009. Programme, Abstracts and Field Trips Guidebook, 16. brensis, and I. hahni gen. et sp. nov.; another two different taxa, Barco, J.L., Canudo, J.I., Cuenca-Bescós, G., Ruiz-Omeñaca, J.I., 2005. Un nuevo provisionally classified as Paulchoffatiidae indet. and Plagiaulacidae dinosaurio saurópodo Galvesaurus herreroi gen. nov., sp. nov., del tránsito or Eobaataridae indet. could be added to the list if the validity of Jurásico-Cretácico en Galve (Teruel, NE de España). Naturaleza Aragonesa 15, 4e17. these taxa can be assessed once more specimens are found. The Buscalioni, A.D., Fregenal, M.A., Bravo, A., Poyato-Ariza, F.J., Sanchíz, B., Báez, A.M., litho- and chrono-stratigraphic record of these six taxa along the Cambra, Moo O., Martín-Closas, C., Evans, S.E., Marugán Lobón, J., 2008. The Galve section is indicated in Fig. 6. vertebrate assemblage of Buenache de la Sierra (Upper Barremian of Serranía de The presence of G. nannothus and E. hispanicus in the late Bar- Cuenca, Spain) with insights into its taphonomy and palaeoecology. Cretaceous Research 29, 687e710. remian site of Uña (Hahn and Hahn, 1992, 2002, 2006) is ques- Canudo, J.I., Cuenca, G., 1996. Two new mammalian teeth (Multituberculata and tionable. Morphological differences between the P1/3s of E. Peramura) from the Lower Cretaceous (Barremian) of Spain. Cretaceous e hispanicus from Uña and Galve have been stated in the previous Research 17, 215 228. Canudo, J.I., Amo, O., Cuenca-Bescós, G., Meléndez, A., Ruiz-Omeñaca, J.I., Soria, A.R., section, and they should probably be regarded as belonging to 1997. Los vertebrados del Tithónico-Barremiense de Galve (Teruel, España). different species. The same probably applies to G. nannothus. This Cuadernos de Geología Ibérica 23, 209e241. species was described on the basis of its holotype, a I2 which came Canudo, J.I., Ruiz-Omeñaca, J.I., Aurell, M., Barco, J.L., Cuenca-Bescós, G., 2006. A e megatheropod tooth from the late Tithonian-lower Berriasian (Jurassic-Creta- from the late Hauterivian early Barremian site of Pelejón-2 in ceous transition) of Galve (Aragon, NE Spain). Neues Jahrbuch fur Geologie und Galve, together with the paratype (?p3/4) and a tiny P1/3 (Hahn Paläontologie, Monatshefte 239 (1), 77e99. and Hahn, 2002: fig. 2) from another early Barremian localities Cope, E.D., 1884. The Tertiary Marsupialia. American Naturalist 18, 686e697. fi fi Crusafont-Pairó, M., Adrover, R., 1966. El primer mamífero del Mesozoico español. from Galve (Hahn and Hahn, 1992: gs. 2,4, 2002: g. 2). A second Fossilia 5e6, 28e33 (corresponding to year 1965). I2 described also as paratype in Uña were assigned to G. nannothus Crusafont-Pairó, M., Gibert, J., 1976. Los primeros multituberculados de España. (Hahn and Hahn, 1992: fig. 3). The I2s are similar among the Nota preliminar. Acta Geologica Hispánica 11, 57e64. Cuenca-Bescós, G., Canudo, J.I., 2004. Los mamíferos del Cretácico inferior de different representatives of the Paulchoffatiidae and the other España. Geo-temas 6, 35e38. “plagiaulacidan” multituberculates, so the presence of the G. nan- Cuenca-Bescós, G., Badiola, A., Canudo, J.I., Gasca, J.M., Moreno-Azanza, M., In press. nothus in Uña with the current available specimens is also ques- New dryolestidan mammal from the Hauterivian-Barremian transition of the tionable. Moreover, between the terminal Hauterivianebasal Iberian Peninsula. Acta Palaeontologica Polonica, doi:10.4202/app2009.0157. Díaz Molina, M., Yébenes, A., 1987. La sedimentación litoral y continental durante el Barremian of the El Castellar and Camarillas Fms. e with the first Cretácico Inferior. Sinclinal de Galve, Teruel. In: Sanz, J.I. (Ed.), Geología y appearance of E. hispanicus and G. nannothus in the Galve area Paleontología (arcosaurios) de los yacimientos cretácicos de Galve (Teruel) y (Fig. 6) e and the late Barremian of the La Huerguina Fm., where the Tremp (Lérida). Estudios Geológicos, Volumen Extraordinario Galve-Tremp, Madrid, pp. 3e21. late Barremian site of Uña is located, there is a span of around 2 Myr Díez, J.B., Pons, D., Canudo, J.I., Cuenca, G., Ferrer, J., 1995. Primeros datos pal- (Gradstein et al., 2004). In the same way, the new species of Cru- inológicos del Cretácico inferior continental de Pielago, (Galve, Teruel). In: XI safontia described by Cuenca-Bescós et al. (in press) in Galve has Jornadas de Paleontología, Tremp (Spain), Abstract volume, 79e81. Eaton, J.G., Cifelli, R.L., 2001. Multituberculate mammal from near the EarlyeLate allowed us to discard the presence of C. cuencana in this area, the Cretaceous boundary, , Utah. Acta Palaeontologica holotype of which was described in Uña. Polonica 46, 453e518. Author's personal copy

A. Badiola et al. / Cretaceous Research 32 (2011) 45e57 57

Engelmann, G.F., Callison, G., 1999. Glirodon grandis, a new multituberculate Krebs, B., 1985. (Mammalia) aus der Unterkreide von Galve (Provinz Teruel mammal from the Upper Jurassic Morrison Formation. In: Gillete, D.D. (Ed.), Spanien). Berliner Geowissenschaftliche Abhandlungen A 60, 29e48. Vertebrate Paleontology in Utah. Utah Geological Survey, Special Publication 99, Krebs, B., 1993. Das Gebiss von Crusafontia (Eupantotheria, Mammalia)eFunde aus pp. 161e177. Salt Lake City. der Unter-Kreide von Galve und Uña (Spanien). Berliner Geowissenschaftliche Estes, R., Sanchiz, B., 1982. Early Cretaceous lower vertebrate from Galve (Teruel), Abhandlungen E 9, 233e252. Spain. Journal of Vertebrate Paleontology 2, 21e39. Kusuhashi, N., 2008. Early Cretaceous multituberculate mammals from the Kuwa- Falconer, H., 1857. Description of two species of the fossil mammalian genus Pla- jima formation (Tetori group), central Japan. Acta Palaeontologica Polonica 53, giaulax from Purbeck. Quarterly Journal of the Geological Society of London 13, 379e390. 261e282. Kusuhashi, N., Hu, Y., Wang, Y., Setoguchi, T., Matsuoka, H., 2009. Two eobaatarid Gasca, J.M., Gómez-Fernández, D., Moreno-Azanza, M., Canudo, J.I., 2009. Un paseo (Multituberculata, Mammalia) genera from the Lower Cretaceous Shahai and por los yacimientos del tránsito Hauteriviense - Barremiense (Cretácico Infe- Fuxin formations, northeastern China. Journal of Vertebrate Paleontology 29, rior) de Aragón. Paleolusitana 1, 211e219. 1264e1288. Gill, Th, 1872. Arrangement of the Families of Mammals. With Analytical Tables. Liesa, C.L., Soria, A.R., Meléndez, N., Meléndez, A., 2006. Extensional fault control on Prepared for the Smithsonian Institution. In: Smithsonian Miscellaneous the sedimentation patterns in a continental rift basin: el Castellar Formation, Collections 230, I-VI, pp. 1e98. Washington D.C. Galve sub-basin, Spain. Journal of the Geological Society, London 163, 487e498. Grambast, L., 1968. Evolution of the utricule in the Charophyte genera Perimneste Martill, D.M., Naish, D. (Eds.), 2001. Dinosaurs of the Isle of Wight. Palaeontological Harris and Atopochara Peck. Journal of the Linnean Society of London (Botany) Association Field Guide to Fossils, 10. The Paleontological Association, London, 61, 5e11. 433 pp. Gradstein, F.M., Ogg, J.G., Smith, A.G. (Eds.), 2004. A Geologic Time Scale 2004. Martín-Closas, C., 1989. Els caròfits del Cretaci inferior de les conques perifèriques Cambridge University Press, Cambridge, p. 589. del Bloc de l’Ebre. Ph.D. dissertation (unpublished), Universitat de Barcelona, Hahn, G., Hahn, R., 1992. Neue Multituberculaten-Zähne aus der Unter-Kreide Barcelona, Spain, 581 pp. (Barremium) von Spanien (Galve und Uña). Geologica et Palaeontologica 26, Martin, T., 1998. The premolars of Crusafontia cuencana (, Mammalia) 143e162. from the Early Cretaceous (Barremian) of Spain. Berliner Geowissenschaftliche Hahn, G., Hahn, R., 1998. Neue Beobachtungen an Plagiaulacoidea (Multi- Abhandlungen E 28, 119e126. tuberculata) des Ober-Juras. 3. Der Bau del Molaren bei den Paulchoffatiidae. In: McKenna, M.C., 1971. Multituberculata. In: McGraw-Hill Encyclopedia of Science Berliner Geowissenschaftliche Abhandlungen E 28, pp. 39e84. and Technology 8. McGraw-Hill Book Company, New York, pp. 736e737. Hahn, G., Hahn, R., 1999. Pinheirodontidae n. fam. (Multituberculata, Mammalia) Meléndez, N., Liesa, C.L., Soria, A.R., Meléndez, A., 2009. Lacustrine system evolution aus der tiefen Unter-Kreide Portugals. Palaeontographica 253, 77e222. during early rifting: el Castellar formation (Galve sub-basin, Central Iberian Hahn, G., Hahn, R., 2001. Multituberculaten-Zähne aus der Unter-Kreide (Barre- Chain). Sedimentary Geology 222, 64e77. mium) von Pié Pajarón (Prov. Cuenca, Spanien). Paläontologische Zeitschrift 74, Owen, R., 1871. Monograph of the fossil Mammalia of the Mesozoic formations. 587e589. Palaeontographical Society 33, 1e115. Hahn, G., Hahn, R., 2002. Neue Multituberculaten-Zähne aus dem Barremium Riveline, J., Berger, J.P., Feist, M., Martín-Closas, C., Schudack, M., Soulié-Märsche, I., (Unter-Kreide) von Galve (Spanien). Paläontologische Zeitschrift 76, 257e259. 1996. European mesozoic- charophyte biozonation. Bulletin de la Hahn, G., Hahn, R., 2004. The dentition of the Plagiaulacida (Multituberculata, Late Société Geologique de France 167, 453e468. Jurassic to Early Cretaceous). Geologica et Palaeontologica 38, 119e159. Ruiz-Omeñaca, J.I., Canudo, J.I., Aurell, M., Bádenas, B., Barco, J.L., Cuenca-Bescós, G., Hahn, G., Hahn, R., 2006. Catalogus Plagiaulacidorum cum figuris (Multituberculata Ipas, J., 2004. Estado de las investigaciones sobre los vertebrados del Jurásico suprajurassica et subcretacea). In: Riegraf, W. (Ed.), Fossilium Catalogus I: Superior y Cretácico Inferior de Galve (Teruel). Estudios Geológicos 60, Animalia Pars 140. Backhuys Publishers, Leiden, p. 344. 179e202. Henkel, S., Krebs, B., 1969. Zwei Säugetier-Unterkiefer aus der Unteren Kreide von Salas, R., Guimera, J., Mas, R., Martín-Closas, C., Meléndez, A., Alonso, A., 2001. Uña (Prov. Cuenca, Spanien). Neues Jahrbuch fur Geologie und Paläontologie, Evolution of the Mesozoic Central Iberian Rift System and its Cainozoic inver- Monatshefte 8, 449e463. sion (Iberian chain). Mémoires du Museum d’Histoire Naturelle 186, 145e185. Hu, Y., Wang, Y., 2002. Sinobaatar gen. nov.: first multituberculate from the Jehol Sanz, J.L., Buscalioni, A.D., Casanovas, M.L., Santafé, J.V., 1987. Dinosaurios del Cre- Biota of Liaoning, northeast China. Chinese Science Bulletin 47, 382e386 [in tácico Inferior de Galve (Teruel, España). In: Estudios Geológicos, Volumen Chinese with English summary]. Extraordinario Galve-Tremp, 45e64. Kielan-Jaworowska, Z., Ensom, P.C., 1992. Multituberculate mammals from the Schudack, M., 1989. Charophytenfloren aus den unterkretazischen Vertebraten- Upper Jurassic Purbeck formation of southern England. Palae- Fundschichten bei Galve und Uña (Otspanien). Berliner Geowissenschaftliche ontology 35, 95e126. Abhandlungen A 106, 409e443. Kielan-Jaworowska, Z., Ensom, P.C., 1994. Tiny multituberculate Simpson, G.G., 1928. A Catalogue of the Mesozoic Mammalia in the Geological mammals from the Purbeck Limestone formation of Dorset, England. Palae- Department of the BritishMuseum. British Museum Natural History, London, ontology 37, 17e31. 215 pp. Kielan-Jaworowska, Z., Dashzeveg, D., Trofimov, B.A., 1987. Early Cretaceous mul- Soria de Miguel, A.R., 1997. La sedimentación en las cuencas marginales del Surco tituberculates from Mongolia and a comparison with forms. Acta Ibérico durante el Cretácico Inferior y su control estructural. Ph.D. dissertation. Palaeontologica Polonica 32, 3e47. Servicio de Publicaciones de la Universidad de Zaragoza, Spain, 363 pp. Kielan-Jaworowska, Z., Cifelli, R.L., Luo, Z.H., 2004. Chapter 8. Allotherians. In: Soria, A.R., Martín-Closas, C., Meléndez, A., Meléndez, N., Aurell, M., 1995. Estrati- Mammals from the Age of Dinosaurs. Origin, Evolution and Structure. Columbia grafía del Cretácico inferior del sector central de la Cordillera Ibérica. Estudios University Press, New York, pp. 249e342. Geológicos 51, 141e152.