The Early Cretaceous Lizard Genus Yabeinosaurus from China: Resolving an Enigma

Home , Ardeosaurus, Bavarisaurus, Dorsetisaurus, Paramacellodus, Parviraptor, Scleroglossa

Journal of Systematic Palaeontology

ISSN: 1477-2019 (Print) 1478-0941 (Online) Journal homepage: http://www.tandfonline.com/loi/tjsp20

The early Cretaceous lizard genus Yabeinosaurus from China: Resolving an enigma

Susan E. Evans , Yuan Wang & Chun Li

To cite this article: Susan E. Evans , Yuan Wang & Chun Li (2005) The early Cretaceous lizard genus Yabeinosaurus from China: Resolving an enigma, Journal of Systematic Palaeontology, 3:4, 319-335, DOI: 10.1017/S1477201905001641

To link to this article: http://dx.doi.org/10.1017/S1477201905001641

Published online: 09 Mar 2010.

Submit your article to this journal

Article views: 107

View related articles

Citing articles: 21 View citing articles

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=tjsp20

Download by: [National Science Library] Date: 22 March 2016, At: 02:17 Journal of Systematic Palaeontology 3 (4): 319-335 Issued 30 November 2005 doi:10.1017/S1477201905001641 Printed in the United Kingdom © The Natural History Museum

THE EARLY CRETACEOUS LIZARD GENUS Yabeinosaurus FROM CHINA: RESOLVING an enigma

Susan E. Evans* Department of Anatomy and Developmental Biology, University College London, Gower Street, London WC1E 6BT, England

Yuan Wang Institute of Vertebrate Paleontology and Paleoanthropology, Chinese Academy of Sciences, 142 Xi-Zhi-Men-Wai St, P.O.Box 643, Beijing 100044, China

Chun Li Institute of Vertebrate Paleontology and Paleoanthropology, Chinese Academy of Sciences, 142 Xi-Zhi-Men-Wai St, P.O.Box 643, Beijing 100044, China

SYNOPSIS The lizard genus Yabeinosaurus was originally described more than 60 years ago from deposits of the Early Cretaceous Jehol Group of northeastern China. Its morphology and phylogenetic position have remained enigmatic for several reasons, not least the inaccessibility of the type and figured material, now lost, and the immaturity of all previously documented specimens. Recent work on the Jehol Biota has yielded a series of well-preserved Yabeinosaurus skeletons, ranging from juvenile to adult. Contrary to previous interpretations, Yabeinosaurus was not a small, weakly ossified lizard with gekkotan affinities. The adult reached a size in excess of 300 mm snout-pelvis length and had a large, heavily ossified and strongly sculptured skull. Phylogenetic analysis highlights some of the difficulties involved in determining the systematic positions of early squamate taxa. Nevertheless, the overall results suggest a position close to the iguanian-scleroglossan dichotomy (just above or just below). As such, Yabeinosaurus may represent a relict taxon in the Jehol Biota, a survivor of the Pan-Laurasian lizard fauna of the Jurassic.

KEYWORDS Jehol Biota, Squamata, Reptilia, Asia

Contents Introduction 320 Geological background 320 Downloaded by [National Science Library] at 02:17 22 March 2016 Systematic palaeontology 321 Squamata Oppel, 1811 321 Yabeinosaurus Endo & Shikama, 1942 321 Description 322 Skull 322 Mandibles 328 Postcranial skeleton 328 Soft tissue traces 331 Discussion 331 Implications 332 Acknowledgements 333 References 333 Appendix: Character lists and data matrices used 335

* Corresponding author. E-mail address: [email protected] 320 S.E. EVANS ETAL.

INTRODUCTION Rocks of the Jehol Group (Yixian and Jiufotang formations) outcrop over a large area of northeastern China from northern Hebei Province to western Liaoning Province, across the southeast part of the Inner Mongolia Autonomous Region. Although there remains some controversy over aspects of the dating, most scientists now regard the deposits as Early Cretaceous in age (Valanginian to Aptian: Smith et al. 1995; Wang & Zhou 2003; Zhou etal. 2003). These fossil horizons have yielded a spectacular array of plant, invertebrate and vertebrate fossils, including fish, salamanders, frogs, turtles, lizards, choristoderes, pterosaurs, dinosaurs, mammals and birds. The first lizard fossil described was Yabeinosaurus tenuis (Endo & Shikama 1942), but other taxa have been added more recently (Ji 1998; Ji & Ren 1999). Endo & Shikama (1942) referred Yabeinosaurus to a new family, Yabeinosaur- idae, an attribution accepted by Young (1948). Young (1958) later recorded a second specimen (IVPP V961) from Ket- zutung (Gezidong), Lingyuan, Liaoning, placing it in the type species. However, Hoffstetter (1964, 1967a) referred the specimen to a new species, Y. youngi, on the basis of minor proportional differences. This specimen remains in the collections of the Institute of Vertebrate Paleontology and Paleoanthropology (IVPP) in Beijing, but the preservation is extremely poor and its status is indeterminate. Hoffstetter (1964) placed Yabeinosaurus close to the European Ardeosaurus and Eichstaettisaurus, within the family Ardeosauridae. This position was adopted by Estes (1983a) in his brief review of the taxon. Both Hoffstetter (1964, 1967a) and Estes (1983a) classified Ardeosauridae within the Gekkota, but on pre-cladistic criteria. Neither the original descriptions, nor Estes' review, provide any significant morphological information on Yabeinosaurus beyond small size, paired frontals and relatively short limbs. There is thus little basis for a discussion of its phylogenetic position. As noted by Estes (1983a: 63), the original description implies that the holotype specimen was a very young animal since the scapula and coracoid and the astragalus and calcaneum remained unfused. This holotype is now lost. Ji et al. (2001) recently described a further specimen of Yabeinosaurus tenuis (YFM R002) and erected Figure 1 Yabeinosaurus tenuis. IVPP V13285 (adult skull) in this as a neotype. Unfortunately, this specimen is also very comparison with IVPP V12641 (juvenile skeleton). Eph, Ephemeropsis immature and the description provides no phylogenetically trisetalis (a common Jehol insect larva). Downloaded by [National Science Library] at 02:17 22 March 2016 useful data. Four new specimens in the collections of the IVPP in Beijing have permitted a detailed reassessment of Yabeino- Group to the Yixian and Jiufotang formations. Most Chinese saurus. They provide a growth series (Fig. 1), showing that workers recognise four major divisions of the Yixian Form- the adult Yabeinosaurus was a large, well-ossified lizard ation: Lujiatun, Jianshangou, Dawangzhangzi and Jingang- with a long period of post-hatchling growth. This explains shan. Overlying the last of these is the Boluchi Bed of the the difficulties encountered by previous workers. The adult Jiufotang Formation. Multiple tuff horizons have yielded a Yabeinosaurus shows no paedomorphic traits and is not re- series of consistent dates: Lujiatun Bed (c. 129-128 Ma), lated to gekkotans (contra Hoffstetter 1964; Estes 1983a). Jianshangou Bed (c. 124—125 Ma), Dawangzhangzi Bed Rather it appears to be a basal squamate. (c. 123-122 Ma), Jingangshan Bed (c. 121 Ma), top of the Jiufotang Formation (110 Ma) (Wang & Zhou 2003). Thus, under current understanding, the rocks of the Jehol Group span a period from the Valanginian to the Aptian. Underly- ing the Yixian Formation (but unconformably) is the Jurassic GEOLOGICAL BACKGROUND Tuchengzi Formation, dated at 139 Ma. The age and geology of the Jehol Group have been described The specimens of Yabeinosaurus described here were in some detail in recent years (e.g. Wang & Zhou 2003; Zhou recovered from the Jianshangou Bed (at Sihetun locality, et al. 2003) and these works should be consulted for a com- Liaoning: IVPP V13285, Figs 1 & 2), the Dawangzhangzi prehensive discussion. The current consensus limits the Jehol Bed (at Dawangzhangzi locality, Liaoning: IVPP V13284; EARLY CRETACEOUS LIZARD GENUS YABEINOSAURUS 321

Figure 2 Yabeinosaurus tenuis. IVPPV13285, adult skull, mainly in ventral view, with skull roof bones displaced and rotated to the left of the skull, and left manus. Abbreviations: l.j, left jugal; l.mx, left maxilla; l.n, left nasal; l.q, left quadrate; l.sq, left squamosal; ms, manus; par, parietal; pmx, premaxilla; r.j, right jugal; r.prf, right prefrontal; r.q, right quadrate; r.sq, right squamosal; sm, septomaxilla; ?, unknown.

Fig. 3), the Jingangshan Bed (at Jingangshan locality, Liaon- ORIGINAL HOLOTYPE. Central National Museum, Man- ing: IVPP V12641; Fig. 1) and the Jiufotang Formation churia, CNMM 3735 (now lost). (at Dapingfang locality, Liaoning: IVPP V13272). Endo & Shikama's (1942) holotype (CNMM 3735) is now lost, TYPE LOCALITY AND HORIZON. Zaocishan, Liaoning. but was recorded as coming from the Tsaotzushan Form- Jingangshan Bed, Yixian Formation. ation, Tsaotzushan (Zaocishan), 21 km southwest of Ihsien NEOTYPE. YFM R002, Yizhou Fossil Museum, Yixian (= Yixian), Chinchou Province (= Jinzhou City region, Je- County, Liaoning. hol Province). This horizon is now considered to be equi- valent to the Jingangshan Bed (Wang X. L., pers. comm., NEOTYPE LOCALITY. Jingangshan locality, Jingangshan

Downloaded by [National Science Library] at 02:17 22 March 2016 2004). The neotype specimen described by Ji et al. (2001) Bed, Yixian Formation. also came from the Jingangshan Bed, Jingangshan locality. REFERRED MATERIAL. Institute of Vertebrate Paleontology Thus Yabeinosaurus persisted through the depositional his- and Paleoanthropology IVPP V12641 (juvenile skeleton); tory of the Jehol Group, but has yet to be recorded from the IVPP V13272A,B (alarge immature skeleton); IVPP V13284 earliest level. The indeterminate holotype of'Yabeinosaurus' (a small immature skeleton); IVPP V13285 (adult or subadult youngi (IVPP V961) is recorded as coming from Ketzutung skull and left forelimb). (Gezidong), Lingyuan region, Liaoning (Hoffstetter 1964). The age of this horizon is uncertain, but it could be somewhat TEMPORAL AND GEOGRAPHICAL RANGE. Currently limited older than the Lujiatun. to the beds of the Jehol Group (Yixian Formation and Jiufotang Formation) of northeastern China, ranging in age from 121-110 Ma. DIAGNOSIS. (revised from Endo & Shikama 1942; Estes SYSTEMATIC PALAEONTOLOGY 1983a). Large lizard reaching 300-350 mm snout-vent length in adult, characterised by the following combination SQUAMATA Oppel, 1811 of characters: coarse vermiculate cranial sculpture; premax- Yabeinosaurus Endo & Shikama, 1942 illae fused but not fully co-ossified in adult, nine teeth; maxilla short, extending only part-way below the orbit; max- TYPE SPECIES. Yabeinosaurus tenuis Endo & Shikama, illa anteriorly forked, clasping premaxilla; palatal shelf of 1942. maxilla small, but may have been overlapped by expanded 322 S.E. EVANS ETAL.

toothed, bearing medial facets for pterygoid; pterygoid with long median tooth row and slender quadrate process, deep dorsal epipterygoid fossa; marginal dentition pleurodont with predominantly iguanid replacement (sensu Edmund 1960), tooth tips acuminate, with medially inflected tips; lower jaw characterised by strong, hook-like, angular process on articular; vertebrae procoelous, with broad rounded centra and large subcentral foramina; second sacral rib unusually stout; clavicle sigmoid; astragalus and calcaneum co-ossify late in development; fifth metatarsal proximally broad with strong outer process.

REMARKS. The loss of the original holotype (Endo & Shikama 1942) and the immaturity of the recently named neotype (Ji et al. 2001) make comparisons with more mature material difficult. Equally problematic is the fact that almost all of the diagnostic characters originally listed by Endo & Shikama (1942) are either generalised lizard features (procoelous vertebrae, streptostylic quadrate), evidence of immaturity (e.g. weakly developed pectoral and pelvic girdles), or probable misinterpretation (e.g. neck length, orbit morphology). However the smallest of the IVPP specimens (IVPP V12641) resembles Endo & Shikama's original photographs of CNMM 3735 (1942: pl. 6) in body proportions (skull, vertebral column and limbs) and differs from the contempor- aneous Dalinghosaurus (Ji 1998; Ji & Ji 2004) which has shorter forelimbs and very long feet. IVPP V12641 also resembles CNMM 3735 in the length and appearance of the ribs and, less certainly (because the photographs are not clear), in the form of the parietal, the parietal foramen and the frontoparietal suture. The two specimens are of roughly the same size and skeletal maturity, indicating that the original holotype was a young individual of a much larger lizard rather than a weakly ossified adult. We are therefore satisfied that the IVPP material is genuinely referable to Yabeinosaurus. The existing holotype of Y. youngi Hoffstetter, 1964, is a very poorly preserved juvenile specimen, mostly impression.

Figure 3 Yabeinosaurus tenuis. IVPP V13284, immature skeleton, DESCRIPTION in ventral view, with most of the tail lost. Skull medial border of vomer; nasals slender, parallel-edged, re- IVPP VI3285 is a large adult skull that has been prepared in Downloaded by [National Science Library] at 02:17 22 March 2016 duced contribution to narial margin; frontals paired with com- both dorsal and ventral views, but is lacking the braincase. plex median suture, deep orbital flanges that do not approach Parts are preserved in articulation, but the skull roof has been in the midline; frontals narrow between orbits; frontopari- rotated through 180 degrees about its long axis and has also etal suture strongly interdigitated; parietal single, narrow and been inverted. As recovered, the skull exposed the ventral with small deeply placed parietal foramen that was closed in surface of the mandibles and palate, but the dorsal surface of the adult; postparietal processes with juxtaposed bases, pro- the skull roof (Figs 1 & 2). The other side has been prepared, cesses posteriorly directed, long axes of upper temporal fen- exposing a dorsal view of the palate and a ventral view of estrae anteroposteriorly elongate; bases of postparietal pro- the skull roof (Fig. 4). The right maxilla remains in situ but cesses perforated ventrally by conspicuous foramina; pit for the left moved with the skull roof. Both jugals lie out to the processus ascendens of supraoccipital deep and extending side of the specimen, as does the right squamosal and a small vertically into parietal; squamosal long and slender; post- bone that may be a septomaxilla (Fig. 2). The following de- frontal clasping frontoparietal suture; postorbital and post- scription is based on IVPP V13285, unless otherwise stated. frontal separated by suture but closely associated and form- ing a single complex; postorbital posterior process restricts Premaxilla upper temporal fenestra; jugal arch complete (contra Endo & The premaxillae are partially fused but retain a trace of the Shikama 1942; Estes 1983a); jugal mostly overlain by max- midline suture on both alveolar and nasal portions (Fig. 5C). illa in orbital margin; jugal with short posterior spur; quadrate The alveolar margin bears nine small pleurodont teeth. There with large, but shallow, lateral conch and distinct pterygoid is little development of palatal flanges but the lateral edges of lappet, dorsal head perforated for 'squamosal peg'; palatine each bone bear deeply incised anterior and posterior facets EARLY CRETACEOUS LIZARD GENUS YABEINOSAURUS 323

Figure 4 Yabeinosaurus tenuis. IVPPV13285, skull in mainly dorsal view, with skull roof bones displaced and rotated to the left. Abbreviations: l.art, left articular; l.den, left dentary; l.ect, left ectopterygoid; l.fr, left frontal; l.mx, left maxilla; l.n, left nasal; l.pt, left pterygoid; l.q, left quadrate; l.sq, left squamosal; l.v, left vomer; par, parietal; pmx, premaxilla; pof, postorbitofrontal complex; r.art, right

Downloaded by [National Science Library] at 02:17 22 March 2016 articular; r.cor, right coronoid; r.den, right dentary; r.mx, right maxilla; r.prf, right prefrontal; r.pt, right pterygoid; r.q, right quadrate.

for the maxilla. Dorsally, each premaxilla extends a long The palatal shelf is relatively narrow, except posteriorly narrow nasal process. where it met the palatine. There is no evidence that the palatal shelf was sutured to the vomer, but IVPP V13285 suggests Maxilla it may have been overlapped by the broad rounded margin The maxillae are large triradiate bones with a short premaxil- of the vomer. This is the incomplete neochoanate condition lary process, a strong, rounded facial process that bears coarse (anterior vomeronasal opening separated from the choana). vermiculate sculpture and an orbital process that extended More posteriorly, the maxilla bears medial facets for the pal- part way below the orbit (Figs 5A & B). This orbital process atine and ectopterygoid. A large medial foramen provided is deep anteriorly and then tapers posteriorly; most of the the entry point for the maxillary nerve and blood vessels jugal lay medial to it, so that the maxilla made a broad entry (Fig. 5B). into the orbital rim. Each maxilla bears around 25 tooth positions. The premaxillary process is forked, with anterior and Nasal posterior rami fitting into slots in the premaxilla (Fig. 5C). A partial left nasal is all that remains of a pair of apparently The notch between the rami is rounded and may have contrib- long, narrow bones with parallel margins and a small anterior uted to a small premaxillary fontanelle. Medially, the maxilla narial border (Figs 4 & 5 A). The anteromedial edge bears a bears a large facet for the prefrontal, but further ventrally facet for the nasal process of the premaxilla, whereas the the surface is excavated by the nasal chamber (Fig. 5B). anterolateral edge met the maxilla. The dorsal surface bears S.E. EVANS ETAL.

for the prefrontal and the firm suture contributed to a strong antorbital buttress. Anteriorly, the frontals sent a shelf below the nasals but this region is obscured in dorsal view by the left maxilla and mandible. Posteriorly, the frontals had a strong, complex, interdigitating suture with the parietal. Small ventral recesses to either side of the midline accommodated weak flanges from the parietal, while the posterolateral edge is grooved by a facet for the postfrontal/postorbitofrontal.

Parietal

5 mm The parietal is single in the adult, but shows traces of a midline suture in the smallest juvenile specimen (IVPP V12641). prf.fi The parietal is a very distinctive element that permits attri- B bution of less mature specimens to the same taxon. Like f.nnx.n the frontal, the adult bone bears coarse dermal sculpture pa I.ft (Fig. 6A). It is generally narrow and rather elongate, with an expansion at the frontoparietal suture and a narrowing ect.ft between the upper temporal fenestrae. There is a small parietal foramen in the central part of the dorsal surface; ventrally, it lies at the base of a deep pit (Figs 6A & B). The foramen remains open in immature specimens (Fig. 7) but has closed, or nearly closed, in the adult. The postparietal processes pmx are unusual in being directed strongly backward so that the rear margin of the bone is deeply 'U'-shaped. In ventral view, the bases of the processes are seen to be juxtaposed in the midline and are horizontal in orientation; further distally r.mx the processes narrow and the transverse axis turns through 90 degrees. The base of each process is perforated by a conspicuous foramen (Fig. 6B); this is a consistent feature but the function is not known. Between the processes is a narrow wedge of bone that is penetrated anteriorly by a deep pit for the processus ascendens of the supraoccipital (Fig. 6B: p.pr.as). This is placed well anterior to the posterior edge of the skull roof, suggesting the supraoccipital was at least partially covered by the parietal in life. The lateral margins of the parietal are deep, delimited dorsally by a small horizontal crest so that the adductor muscles would have been restricted to the ventrolateral margins of the bone. There are I.den no descending processes or crests meeting the epipterygoids.

Prefrontal Figure 5 Yabeinosaurus tenuis. IVPP V13285. Left maxilla in lateral A single right prefrontal is preserved in IVPP V13285 view (A), and medial view (B);C, anterior part of skull, in mainly dorsal between the displaced skull roof and the left mandible (Figs 2, view. Abbreviations: ect.ft, ectopterygoid facet; f.mx.n, foramen for 4 & 6B). It is composed of a large convex body and a rel- Downloaded by [National Science Library] at 02:17 22 March 2016 the maxillary nerve; l.den, left dentary; l.n, left nasal; l.v, left vomer; atively short, deep, posterior frontal process. Much of the n.ft, facet for nasal; pal.ft, palatine facet; pmx, premaxilla; prf.ft, medial margin lay against the frontal to which it was firmly prefrontal facet; r.mx, right maxilla; su, suture between two halves of sutured. Seen in dorsolateral view (Fig. 6B), the bone has a premaxilla. rather small sculptured region close to the midline and this would have been the only part exposed in life. A large facet for the maxilla covers much of the rest of the lateral surface. conspicuous sculpture whereas the ventral surface is strongly Anteromedially, a sharp ridge separates this maxillary facet concave. from a much smaller medial facet for the nasal. In the orbital margin, the prefrontal forms a strong orbitonasal flange with Frontal a ventral sutural surface for the palatine. Laterally this flange The frontals are paired throughout ontogeny with a strong is notched by the lacrimal foramen. Whether a separate lac- midline suture (Figs 4 & 6). Dorsally they are heavily sculp- rimal completed the foramen laterally is unclear. tured. Each is narrow and parallel-sided throughout most of its length, expanding only at the frontoparietal suture. Vent- Postorbital region rally, the frontals are characterised by strong cristae cranii The expanded frontoparietal boundary bears anterior and (orbital flanges); these are inclined ventromedially but re- posterior facets showing that it was clasped by a medially main well separated in the midline (Fig. 6B). Anterolaterally, bifurcate postfrontal or postorbitofrontal. A bone fitting this each crista would have contributed to the margin of the orbit- description is preserved between the parietal and the right onasal flange. This region of each frontal bears a large facet pterygoid (Figs 4 & 8). The dorsal surface is sculptured, EARLY CRETACEOUS LIZARD GENUS YABEINOSAURUS 325

r.mx

Figure 6 Yabeinosaurus tenuis. IVPPV13285. Skull roofing bones in dorsal view (A), and ventral view (B). Abbreviations: at.a, atlas arch; cr.cr, crista cranii; l.cor, left coronoid; mx.ft, facet on prefrontal for maxilla; n.ft, facet for nasal; o.fl, orbital flange; p.pr.as, pit for the processus ascendens of supraoccipital; par, parietal; par.f, parietal foramen; pp.pr, postparietal process; prf.ft, prefrontal facet; r.fr, right frontal; r.mx, right maxilla; r.prf, right prefrontal.

Downloaded by [National Science Library] at 02:17 22 March 2016 whereas the ventral surface is smoother. In dorsal view, the of the orbital margin and to a thin presence part way below bone appears single, but the ventral surface retains traces of the orbit. The maxilla entered the orbital margin for a short a suture that divides the bone into two unequal parts. From distance. this specimen, it appears that the postfrontal was larger than the postorbital, perhaps excluding the latter bone from the or- Squamosal bital margin. The same bone complex is preserved on IVPP Both squamosals are preserved in IVPP V13285, the right V13284 (Fig. 7). lying between the right and left jugal (Figs 2 & 8D), the left lying within the main skull mass between the left pterygoid and the left mandible (Fig. 4). The bone is thin, elongate and Jugal posteroventrally curved. The lateral surface is convex, the Both jugals are preserved on IVPP V13285 (Figs 8B & C), medial surface is more flattened. Posteroventrally, the bone but they are disassociated from the main skull mass and bears a shallow curved recess. The right squamosal shows lie between it and the forelimb. The right bone is the best no trace of additional facets, but the left bone seems to bear preserved. It is a roughly 'L' shaped bone (Fig. 8B), with an a single dorsal facet, probably for the postorbital. It seems anterior maxillary process and an oblique dorsal ramus, the unlikely that the jugal and squamosal met. two meeting one another at an obtuse angle. Posteriorly the bone is drawn into a small, rounded process. The lateral facet Quadrate on the anterior process shows that most of this surface was Both quadrates are preserved. The right has been detached, covered in life by the maxilla, restricting the jugal to the rear the left is almost in situ close to the end of the left pterygoid 326 S.E. EVANS ETAL.

pof

5 mm

pof B 010 5 mm

Figure 7 Yabeinosaurus tenuis. IVPP V13284, skull in ventral view. Abbreviations: cor, coronoid; ect, ectopterygoid; fr, frontal; hy, hyoid; j, jugal; l.den, left dentary; mx, maxilla; oto, oto-occipital; par, parietal; pra, prearticular; pof, postorbitofrontal complex; prf, prefrontal; pt, pterygoid; q, quadrate; r.den, right dentary; sp, splenial; sur, surangular.

mx.ft

(Figs 2, 4 & 9A-D). The quadrate is large with a broad but shallow lateral conch and a weak tympanic crest (Fig. 9B). Medially, the pterygoid wing is quite large with a small pterygoid lappet (Fig. 9D: pt.l). The ventral head is small, with a very narrow band-like surface for the articular. The dorsal head, by contrast, is large and is perforated by a partially enclosed foramen (Fig. 9B: per.f). Seen in dorsal view (Fig. 9C), the dorsal head is divided between a smooth convex B:per surface laterally (for the squamosal) and a flattened medial surface (for the supratemporal). Another prominent articular surface juts out posteriorly, presumably for the paroccipital process of the oto-occipital (Figs 9A & C: pcc.pr.s). Downloaded by [National Science Library] at 02:17 22 March 2016

Epipterygoid/supratemporal Figure 8 Yabeinosaurus tenuis. IVPP V13285, skull bones. A, On the palatal surface, associated with the quadrate ramus postorbitofrontal complex in dorsal view (but seen from palatal of the left pterygoid, there are two small columnar bones. surface): B, right jugal, lateral view; C, left jugal, lateral view; D, right One of these (Fig. 8A: st) has a distinctly expanded end; squamosal, lateral view. Abbreviations: a.pr, anterior process; ep, the other (Fig. 8A: ep) is more tapered. These are tent- epipterygoid; l.pt, left pterygoid; mx.ft, facet on prefrontal for maxilla; atively interpreted as a supratemporal and an epipterygoid pof, postorbitofrontal complex; q.s, quadrate surface; st, respectively. supratemporal.

?Septomaxilla Vomer A thin plate-like bone lies to the side of the main skull block The vomers are partially preserved in IVPP V13285, but their in IVPP V13285 (Fig 2). It is flat, with a straight medial full structure cannot be reconstructed as they lie as a series margin and a small anterior/posterior process. It may be a of fragments over the maxilla and mandibles in a dorsal view displaced septomaxilla. of the palate (Figs 4 & 5C: v). They appear to have been rather broad with extended lateral margins that may have Palatine overlapped the palatal margins of the maxilla to create the These bones are partially preserved close to the anterior incomplete neochoanate condition. ends of the pterygoids (Fig. 10B). Only two features are EARLY CRETACEOUS LIZARD GENUS YABEINOSAURUS 327

.den I.cor

l.pt

ect.ft

•q

art

5 mm

Figure 10 Yabeinosaurus tenuis. IVPP V13285. A, posterior end of

Downloaded by [National Science Library] at 02:17 22 March 2016 left mandible, dorsal view; B, palatal region in ventral view. Abbreviations: ect.ft, ectopterygoid facet; hy, hyoid; l.cor, left coronoid; l.den, left dentary; l.ect, left ectopterygoid; l.pt, left pterygoid; l.q, left quadrate; pal, palatine; pt, pterygoid; r.prf, right 5mm prefrontal.

Figure 9 Yabeinosaurus tenuis. IVPP V13285. Right quadrate in posterior view (A), lateral view (B) and dorsal view (C); D, articular region of the skull, in mainly dorsal view; E, posterior ends of left and Pterygoid right mandibles (aligned) to show angular process, ventral view. Both pterygoids are well-preserved and are the longest bones Abbreviations: an.pr, angular process; art, articular; co, quadrate in the skull (Fig. 10B). Each has a long triangular anterior conch; l.pt, left pterygoid; l.q, left quadrate; l.sq, left squamosal; palatine plate, a small pterygoid flange with a lateral tuber- pcc.pr.s, surface for articulation of paroccipital process; per.f, osity and a long, slender quadrate process. As preserved, the perforating foramen; pt.l, pterygoid lappet of quadrate; sq.s, medial edges of the pterygoids run closely parallel to one squamosal surface on quadrate; st.s, supratemporal surface on another with only a small interpterygoid vacuity, but this is quadrate; ty.cr, tympanic crest. presumably an artefact of preservation. The pterygoids are further separated in the immature IVPP V13284. The palatal clearcut - that the bones bore a scatter of small teeth and plate bears a long pterygoid tooth row that begins on the pos- that they also bore broad medial facets for the pterygoids that teromedial edge close to the articulation for the basipterygoid presumably separated them (at least partially) in the midline. process of the sphenoid and then curves anterolaterally. The 328 S.E. EVANS ETAL.

tooth row is single posteriorly, but becomes double anteriorly. A sharp 'V’-shaped notch and facet between the palatal plate and the pterygoid flange received the medial process of the ectopterygoid (Figs 10A & B). At the dorsal junction of the pterygoid quadrate process and the pterygoid flange there is a deep excavation (fossa columellae) for the base of the epipterygoid. At this point, the medial border of the bone turns through a sharp angle and the posteromedial face met the basipterygoid process of the sphenoid, but the braincase is not preserved. The quadrate process has a strong dorsal crest but the medial and lateral faces of the bone are not deep. The lateral surface is convex, the medial one concave.

Ectopterygoid The ectopterygoids are preserved on both sides, more or less in the original position, but the left is the more clearly visible (Figs 4 & 10A). It had an anterolateral-posteromedial orientation, with the small transverse lateral head meeting the maxilla and the bifurcate medial portion clasping the pterygoid flange. Figure 11 Yabeinosaurus tenuis. IVPP V13284, mid-dorsal vertebrae in ventral view. Abbreviations: s.c.f, scapulocoracoid fenestra; v.co, vertebral condyle. Hyoid A slender element running parallel to the right pterygoid quadrate process may be part of the hyoid (Fig. 10B) in IVPP Dentition V13285, whereas IVPP V13284 preserves a pair of slender The teeth are of similar size and shape throughout the up- ceratobranchial elements posterior to the parietal and medial per and lower jaws (homodont). They are robust, quite long to the pterygoids (Fig. 7). and have acuminate tips that are posterolingually inclined and weakly striated on the lingual surface. Replacement was active and of the iguanid type described by Edmund (1960), Mandibles with the new tooth emerging immediately lingual to the old and creating a small lingual pit in the base of the old tooth. Both mandibles are in situ (Figs 2 & 4). The dentaries are exposed in ventrolateral view but their medial aspect is obscured by overlying skull bones and by thin plates of bone Postcranial skeleton that probably represent large splenials. There are roughly 24 pleurodont teeth supported by a subdental shelf, but it is The description of the postcranial skeleton is based mainly on not clear how deep this shelf is. The lateral surface is rather IVPP V13284, with additional information on the forelimb shallow and strongly convex, with at least six neurovascular from IVPP V13285 and developmental details from IVPP foramina. V12641 and IVPP V13272. Apart from the partial left forelimb on V13285, all specimens with preserved postcranial The coronoid is well-developed (Fig. 10A) with a thin material are immature, with weak ossification of the ends of lateral flange that is overlapped by a dentary process. Medi- the limb bones and of the carpus and tarsus. In the juvenile ally, the bone is divided into anteromedial and posteromedi- Downloaded by [National Science Library] at 02:17 22 March 2016 specimen (IVPP V12641), the ends of the long bones are ally processes. The coronoid process itself is anteroposteri- completely unossified. orly concave on its medial surface, with a sharp posterior crest. Behind the crest, the coronoid has a narrow recess into which jaw adductor muscles would have inserted. Vertebral column The surangular and prearticular enclose a deep adductor The vertebrae are procoelous with short, broad, robust centra fossa. The surangular bears a dorsal crest that extends back bearing large subcentral foramina (Fig. 11). The neck- to run onto a tuberosity at the end of the adductor fossa. trunk boundary is not clearcut, but there were either seven Anteriorly it extends well forward along the medial surface cervical vertebrae and 19 dorsals, or eight cervicals and of the dentary (as seen in IVPP V13284; Fig. 7). The prear- 18 dorsals. The first ribs begin on the third or fourth cer- ticular extends forward at least to the level of the coronoid vical and they persist to the end of the presacral series, al- process. Posteriorly, the articular has a short, transverse sur- though the last six dorsal ribs decrease in length while re- face for the quadrate divided into medial and lateral parts maining strongly curved. Both sacral vertebrae are strongly (Fig. 9D), although the distal head of the quadrate does not built, but the second sacral rib is unusually broad (Fig. 12). reflect this morphology). Anteriorly, the articular develops IVPP V13284 only preserves six tail vertebrae, the first five a strong tuberosity that would have prevented forward slip- of which carry long transverse processes (Fig. 3). In IVPP page of the quadrate. There is a short, pointed retroarticular V12641 and IVPP V13272 the tail is longer but the preserva- process with a small dorsal concavity on its surface. By far tion is not good enough to be certain about caudal autotomy. the most distinctive feature on the mandible is a large, hook- In the juvenile (IVPP V12641), neither the sacral ribs nor the like medial angular process, associated with the insertion of proximal caudal transverse processes are fused to the body powerful pterygoideus muscles (Figs 2, 4 & 9E). of the vertebra to which they attach. EARLY CRETACEOUS LIZARD GENUS YABEINOSAURUS 329

Pectoral girdle and forelimb The pectoral girdle is preserved in IVPP V12641 IVPP V13272 and IVPP V13284, but all of these specimens are immature, as shown by the lack of fusion of the scapula and coracoid. The girdle is best exposed on IVPP V13284 (Fig. 13). Here the combined scapulocoracoid seems to have a single anterior coracoid fenestra and a scapulocoracoid fenestra (the notch in the scapula is apparently damage). There is no trace of either clavicles or an interclavicle on this specimen, but the juvenile IVPP V12641 preserves simple sickle-shaped clavicles. The humerus is robust with expanded proximal and distal ends, but the epiphyses remain unossified in all specimens in which the element is preserved (i.e. IVPP V12641, IVPP V13272 and IVPP V13284). This is also true of the radius and ulna in these specimens. In IVPP V13285 (Fig. 14), the radius and ulna are more fully ossified, although the olecranon process of the ulna looks as though fusion was recent since it is not fully integrated with the rest of the bone and the epiphysial lines are still visible at the distal ends of both radius and ulna. This may mean that Yabeinosaurus had a relatively long slow period of skeletal growth and matura- tion (as in modern varanids). The carpals are fully ossified in IVPP V13285, with radiale, intermedium, ulnare, a large pisiform, at least one Figure 12 Yabeinosaurus tenuis. IVPP V13284, sacrum, anterior centrale and five distal carpals (Figs 14, 15B). There appear caudal vertebrae and pelvic girdle in ventral view. Abbreviations: also to be a number of additional elements (Figs 14 & 15B), a.c.v, anterior caudal vertebra; gr, groove; on ventral surface of possibly either phalangeal sesamoids or detached phalangeal anterior caudal transverse process; l.fe, left femur; l.il, left ilium; l.is, epiphyses. In IVPP V13284, the carpal count seems also to left ischium; l.pu, left pubis; r.fe, right femur; r.il, right ilium; r.is, right be complete (radiale, ulnare, intermedium, pisiform, cent- ischium; r.pu, right pubis; sv.2, sacral vertebra 2. rale and five distal carpals) but the bones are only beginning to ossify or calcify (it is not clear which). In the juvenile Downloaded by [National Science Library] at 02:17 22 March 2016

Figure 13 Yabeinosaurus tenuis. IVPP V13284, right pectoral girdle elements and right forelimb in ventral view. Abbreviations: a.c.f, anterior coracoid fenestra; h, humerus; ra, radius; sc, scapula; sc.f, scapulocoracoid fenestra; ul, ulna; 1,5, distal carpals/digits. 330 S.E. EVANS ETAL.

ca di4 mt5

Figure 15 Yabeinosaurus tenuis. A, IVPP V13284, right hind limb; B, IVPP V13285, left manus. Abbreviations; as, astragalus; ca, calcaneum; dt3, distal tarsal three; dt4, distal tarsal four; fe, femur; fi, fibula; in, intermedium; mt5, metatarsal 5; pi, pisiform; ra, radius; Figure 14 Yabeinosaurus tenuis. IVPP V13285, left manus in dorsal rad, radiale; ti, tibia; ul, ulna; uln, ulnare; 1,2,3,4,5 distal carpals/

Downloaded by [National Science Library] at 02:17 22 March 2016 view. Abbreviations: in, intermedium; pi, pisiform; ra, radius; rad, digits. radiale; ul, ulna; uln, ulnare; 1,2,3,4,5, distal carpals/digits.

IVPP V12641, the carpus is largely unossified but a few ele- short. The proximal part is perforated by a pectineal foramen. ments are beginning to calcify. The manus in IVPP V13285 The ischium is also simple and short, with a short posterior is relatively short and broad, without elongation of the meta- process. Each of these elements may have changed shape in carpals. The joint surfaces of the phalanges are complete the mature adult. and the claws are strongly built and curved, with a strong The femur, like the humerus, is a stout column of bone flexor tubercle but little development of the dorsal extensor with slightly expanded ends but no joint surfaces. The tibia attachment. In the less mature specimens, the metacarpals is more robust proximally than distally and seems to show and phalanges are simple cylinders without joint surfaces. the development of a cnemial crest (Fig. 15A). Although The phalangeal formula of the manus is 2.3.4.5.3 as inferred the distal epiphysis is not fully developed, there seems to from IVPP V13284 and IVPP V13285. be some notching where it meets the astragalocalcaneum. In IVPP V13284, the femur (20 mm) is longer than the tibia (12.5 mm), but shorter than the foot (30 mm). In the foot Pelvis and hind limbs of IVPP V13284 and IVPP V13272, the astragalus and cal- The elements of the pelvis remain separate in the three imma- caneum remain unfused (Fig. 15A), with the calcaneum as ture specimens examined (Fig. 12). The ilium has a simple, a relatively small simple bone. The fourth distal tarsal is straight blade without anterior or posterior elongation. The large and rounded and there is a small distal tarsal three. pubis is 'flask-shaped', broader proximally than distally and None of the elements looks finished and ossification was EARLY CRETACEOUS LIZARD GENUS YABEINOSAURUS 331

Neither Bavarisaurus nor Ardeosaurus show gekkotan characters. Eichstaettisaurus is more gracile and shares some general similarities with gekkotans (a weak dentition, narrow frontal with strong orbital flanges, incomplete fusion of the parietals), but each of these features is found in members of other lizard groups. The original attribution of Yabeinosaurus to this small cluster of taxa relied primarily on superficial similarity: small size, many small simple teeth, a generalised skeleton and similar body proportions (relatively short limbs). However, all previously described specimens (including the neotype erected by Ji et al. 2001) were skeletally very immature. The recognition of additional, more mature, material with good cranial remains permits some resolution of the enigma. In order to explore the phylogenetic position of Yabeino- saurus we included it in two PAUP analyses, one based on *t. the data matrix and characters in Evans & Chure (1998a), with additional fossil data from Evans & Barbadillo (1997, 1998,1999) and one based on the data matrix of Lee (1998) (see Appendix). In both cases, we included a sample of other well-known Jurassic or Early Cretaceous taxa. Due to the large size of each matrix, we were constrained to using the heuristic search option (PAUP 3.1.1: Swofford 1993), Figure 16 Yabeinosaurus tenuis. IVPP V12641, scale impressions on but tested the results with Branch and Bound searches on a tail and left hind limb; inset at bottom right shows enlargement of the smaller subset of taxa (PAUP* version 4.0b10 for Windows: leg region enclosed in the rectangle on the main figure. Abbreviations: Swofford 2001). ss, scales. Lee (1998) ordered some of his characters and we repeated both his ordering and outgroups (Marmoretta, clearly incomplete. In the juvenile IVPP V12641 the as- Kuehneosauridae, Rhynchocephalia) for consistency. How- tragalus and calcaneum remain as small rounded elements, ever, we added the fossil genera Ardeosaurus (Late Juras- the fourth distal tarsal is round and incompletely ossified, the sic, Germany: Mateer 1982); Bavarisaurus (Late Jurassic, third distal tarsal has not yet appeared. In all specimens the Germany: Cocude-Michel 1963; Evans 1994a), Dorset- metatarsals, like the metacarpals, are simple cylinders, with isaurus (Late Jurassic-Early Cretaceous England, North the exception of the fifth metatarsal which has a strong outer America and, possibly, Asia: Hoffstetter 1967b; Prothero & process and an expanded dorsal head (barely expanded in Estes 1980; Alifanov 1993; Evans & Chure 1998b), IVPP V12641). The phalangeal epiphyses are unfused and Eichstaettisaurus (Late Jurassic-Early Cretaceous Germany, the phalangeal formula is 2.3.4.5.4. Spain and Italy: Broili 1938; Evans et al. 2001, 2004), and Parviraptor (Middle Jurassic-Early Cretaceous, England, Soft tissue traces Portugal and North America: Evans 1994b,1996). Figure 17 shows a strict consensus tree derived from 14 equally parsi- Both IVPP V12641 and IVPP V13272 show traces of the monious trees (Tree length (TL) = 1010, Consistency Index integument in the form of a dark stain around the body, but (CI) = 0.677, Rescaled Consistency Index (RC) = 0.442); the the scales are most clearly visible on the hind limbs and tail semistrict and 50 % Majority rule trees were the same. Al- in IVPP 12641 (Fig. 16). The limb scales appear to have been though all modern families were included individually, the Downloaded by [National Science Library] at 02:17 22 March 2016 uniformly small and ellipsoid; the caudal scales are elongate tree has been simplified by grouping together all anguimorph and rather rectangular. taxa (as referred by Estes et al. 1988, but also including Amphisbaenia and Dibamidae), scincoids (scincids, cordyli- forms), lacertoids (lacertids, teiids, gymnophthalmids) and DISCUSSION iguanians (agamids, iguanids, chamaeleonids) where they Yabeinosaurus has never been included in a phylogenetic formed monophyletic clades. Yabeinosaurus falls into a gen- analysis because so little was known of its anatomy. Pre- eralised cluster of Jurassic-Cretaceous fossil squamates that vious authors (e.g. Hoffstetter 1964, 1967a; Estes 1983a,b) forms a sister taxon to Scleroglossa. This result matches placed it in Gekkota with the Late Jurassic Solnhofen gen- earlier, precladistic, proposals (e.g. Hoffstetter 1964,1967a; era Ardeosaurus, Bavarisaurus and Eichstaettisaurus. How- Estes 1983a,b) of a relationship between Yabeinosaurus and ever, most of the characters used to link these four taxa are the Solnhofen taxa Ardeosaurus and Eichstaettisaurus. How- either superficial or primitive. Bavarisaurus is a stem squam- ever, no derived characters link the members of this clade ate (Evans 1993; Reynoso 1996, 1998; Evans & Barbadillo and Eichstaettisaurus is particularly unstable; with differ- 1997, 1998, 1999; Reynoso & Callison 2000; Evans et al. ent combinations of taxa, it moves around the tree, group- 2004; this paper), whereas Ardeosaurus is somewhat more ing variably with xenosaurs, gekkotans, amphisbaenians, derived in having heavy dermal sculpture, partial closure or scincomorphs. In previous analyses (e.g. Evans 1994b; of the upper temporal fenestrae and vertebral procoely. In Nydam 2000, 2002), Parviraptor grouped with varanoid an- recent analyses (Reynoso 1996, 1998; Evans 2003), it has guimorphs rather than basal forms. In order to test the result been placed either as a stem lizard or a basal scleroglossan. further, we reran the analysis using a Branch and Bound 332 S.E. EVANS ETAL.

Anguimorpha and joins Ardeosaurus and Bavarisaurus as successive outgroups to Iguania + Scleroglossa. As previously, we Scincoidea tested the heuristic search result for Yabeinosaurus by using a Branch and Bound search (PAUP version 4.0b 10 for Dorsstisaurvs Windows) on a smaller subset of taxa (modern families + Lacertoidea Dorsetisaurus, Paramacellodus, Yabeinosaurus and the Solnhofen taxa). The analysis produced three Maximum Xantusiidae Parsimony trees (MPTs) (TL = 776, CI = 0.776, RC = 0.422) in which the positions of all except Eichstaet- Gekkota tisaurus remained unchanged from the heuristic search (Fig. 18B); Eichstaettisaurus grouped with Amphisbaenia. Ardeosaurus Thus although the analyses using the two matrices yiel- Yabeinosaurus ded slightly different results for Yabeinosaurus (just above or just below Iguania), both support the hypothesis that Yabeino- Parviraptor saurus is a basal lizard unrelated to any modern clade. Neither analysis provides any strong support for a close relationship Eichstaettisaurus with Solnhofen lizards, although they are all of similar grade, and neither supports placement of Yabeinosaurus within Iguania Gekkota (contra Young 1948, Hoffstetter 1964,1967a; Estes 1983a,b). Outgroitp It should perhaps be noted that while these analyses support a relatively conservative phylogenetic arrangement for Figure 17 Cladogram (strict consensus tree of 14 equally squamates, with an early dichotomy into Iguania and Sclero- parsimonious trees resulting from Heuristic search) derived using the glossa, a recent molecular analysis (Townsend et al. 2004) data matrix of Lee (1998) with the addition of Ardeosaurus, has proposed a radically different arrangement in which Bavarisaurus (in outgroup), Dorsetisaurus, Eichstaettisaurus, iguanians are nested within a paraphyletic Scleroglossa. Parviraptor and Yabeinosaurus. Some well-accepted branches have been minimised to simplify the tree. See the text for further explanation. IMPLICATIONS search (PAUP version 4.0b 10 for Windows) with a smal- The Late Cretaceous lizard faunas of eastern Asia are well ler subset of taxa (living families with Parviraptor and the known and were clearly exceptionally diverse (e.g. Gao & three Solnhofen taxa). The search yielded 8 trees (TL = 744, Hou 1996; Gao & Norell 2000). At least part of this diversity CI = 0.788, RC = 0.44), but consensus trees were almost- can be traced back into the Early Cretaceous assemblages of completely unresolved (except for iguanians, amphisbaeni- Mongolia (Alifanov 1993) and Central Asia (Nessov 1988; ans + snakes and scincoids). Gao & Nessov 1998), from which acrodont iguanians, xen- The second analysis used the data matrix of Evans & osaurid anguimorphs, scincids, macrocephalosaurid teioids Chure (1998a) with the addition of a wider range of fossil and gekkotans have been recorded. In Europe and North taxa, taken from Evans & Barbadillo (1997,1998,1999). The America, known Early Cretaceous lizard assemblages more first heuristic search (PAUP 3.1.1: Swofford 1993) included closely resemble those of the Late Jurassic (Evans 1998b, only Aigialosaurus (as a basal mosasauroid: Carroll & De 2003; Nydam 2002) and it has been suggested that migration Braga 1992) and Yabeinosaurus. Yabeinosaurus emerged at from Asia in the mid-Cretaceous was responsible for the in- the base of Scleroglossa, consistent with the first Lee ana- creased diversity seen in North America around this time lysis (TL = 478, CI = 0.54, RC = 0.31: Fig. 18A). A second, (Nydam 2002). However, Early Cretaceous Asian faunas Downloaded by [National Science Library] at 02:17 22 March 2016 more extensive, analysis included also the Late Jurassic reportedly retained rare survivors of more primitive taxa, Ardeosaurus (Mateer 1982) and Bavarisaurus (Evans including paramacellodids and dorsetisaurs, both of which 1994a), the Late Jurassic to Early Cretaceous Dorsetisaurus have been described from the Aptian-Albian of Mongolia (e.g. Hoffstetter 1967b) and Eichstaettisaurus (Broili 1938; (Alifanov 1993), as well as from the Late Jurassic and Evans et al. 2001, 2004), the Early Cretaceous Hoyalacerta Early Cretaceous of Europe and North America (e.g. Seiffert (Barremian, Spain: Evans & Barbadillo 1999) and 1973; Prothero & Estes 1980; Evans 1998a,b, 2003; Meyasaurus (Barremian, Spain: Evans & Barbadillo 1997), Averianov & Skutchas 1999; Nydam 2002; Evans & the Late Jurassic to Early Cretaceous Paramacellodus Searle 2002). Yabeinosaurus has always been an enigma be- (Evans & Chure 1998a); the Middle Jurassic to Early cause it was so poorly known. As noted above, previous Cretaceous Parviraptor (Evans 1994b; Nydam 2000, 2002) workers (e.g. Young 1958; Hoffstetter 1964; Estes 1983a,b) and the Early Cretaceous Scandensia (Evans & Barbadillo had suggested a link between it and the Solnhofen taxa Ar- 1998). This resulted in four equally parsimonious trees deosaurus, Bavarisaurus and Eichstaettisaurus, but on the (strict consensus, Fig. 18B), but with very poor tree stat- basis of superifical similarity and, to at least some degree, istics (TL = 568, CI = 0.46, RC = 0.24). Eichstaettisaurus, skeletal immaturity. More recent work (e.g. Reynoso 1996, Hoyalacerta, Parviraptor and Scandensia group together at 1998; Evans & Barbadillo 1997, 1998, 1999; Reynoso & the base of Scleroglossa, but apart from Eichstaettisaurus, Callison 2000) has shown that the three Solnhofen taxa are these genera are incompletely known, have few comparable not close relatives, although they all appear to lie close to the parts and share no significant derived characters. In this more base of the squamate radiation. Taken together, the results of inclusive analysis, Yabeinosaurus drops out of Scleroglossa our analyses suggest that Yabeinosaurus occupies a similar EARLY CRETACEOUS LIZARD GENUS YABEINOSAURUS 333

B Iguaria Gekkota Cordylidae Seine idae Paramacettodus Iguania Lacertidae Gekkota Teiidae Anguidae Gymnophthalmidae Helodermatidae Msyasaurus — Varanus Xantusiidae — Lanthanotus Anguidae Sarpenles Amphisbaenia Varanus Dibamidae Lanth0notu$ Aigialosaurus Serpsntss Cordylidae Amphisbaenia Scinddaa Dibamidae Laoerlidae Aigiafos&urus I— Teiidae Dorsetisaoms 1— Gymnophthalmidae Sbinisaurus Xanlusiidae Xenosaurus •— Shinissurus Scandensia 1— Xenossurus Yabeinosaurus Parviraptor Outgroup Hoyalscena Yabeinosaurus Ardeosaurus Bavarisaurus Outgroup

Figure 18 Cladograms derived from heuristic searches using a data matrix based on Evans & Chure (1998) with additional data from Evans & Barbadillo (1997,1998,1999). A, first analysis (single tree) including only Aigialosaurus and Yabeinosaurus; B, strict consensus of four equally parsimonious trees resulting from a more inclusive analysis incorporating a range of Jurassic and Early Cretaceous fossil taxa. See the text for further explanation. Downloaded by [National Science Library] at 02:17 22 March 2016 position, perhaps as a relative of Ardeosaurus. As such, it REFERENCES would represent another relict of the original Pan-Eurasian, or Pan-Laurasian, Jurassic squamate assemblage. Alifanov, V. 1993. Some peculiarities of the Cretaceous and Palaeogene lizard faunas of the Mongolian People's Republic. Kaupia 3: 9-13. Averianov, A. O. & Skutchas, P. P. 1999. Paramacellodid lizard (Squamata, Scincomorpha) from the Early Cretaceous of Trans- ACKNOWLEDGEMENTS baikalia. Russian Journal of Herpetology 6: 115-117. Broili, F. 1938. Ein neuer fund von ? Ardeosaurus H. von Meyer. Sitzungs- This manuscript stems from work completed as part of a berichte der Bayerischen Akademie der Wissenschaften, München joint Anglo-Chinese project funded by The Royal Society, 1938:97-114. UK and the National Natural Science Foundation of China. Camp, C. L. 1923. Classification of the lizards. Bulletin of the American We thank Drs Zhong-he Zhou and Xiou-lin Wang for discus- Museum of Natural History 48: 289-481. sions on the Jehol Biota and the geology of the Yixian and Carroll, R. L. & De Braga, M. 1992. Aigialosaurs: mid-Cretaceous Jiufotang Formation, Jane Pendjiky (UCL) who assisted in varanoid lizards. Journal of Vertebrate Paleontology 12: 66-86. Cocude-Michel, M. 1963. Les rhynchocephales et les sauriens des the preparation of figures, Wei Gao and Zhao Wang (IVPP) calcaires lithographiques (Jurassique supérieur) d'Europe occidentale. who photographed and prepared some of the specimens de- Nouvelles Archives du Muséum d'Histoire Naturelle de Lyon 7: 1— scribed here and Ryoko Matsumoto (National Science Mu- 187. seum, Tokyo) who also took some of the photographs. We Edmund, A. G. 1960. Tooth replacement phenomena in the lower ver- are also grateful to the editors and reviewers for their advice tebrates. Contributions of the Royal Ontario Museum, Life Sciences and suggestions. Division 52: 1-190. 334 S.E. EVANS ETAL.

Endo, R. & Shikama, T. 1942. Mesozoic reptilian fauna in the Jehol Colloques Internationaux du Centre National de la Recherche Scien- mountainland, Manchoukuo. Bulletin of the Central National Museum tifique 163: 349-371.

of Manchoukuo 3: 1-19. Ji, S. A. 1998. A new long-tailed lizard from the Upper Jurassic of Liaon- Estes, R. 1983a. Sauria Terrestria, Amphisbaenia. Pp. 1-245 in P. ing, China. Pp. 496-505 in Geology Department of Peking University Wellnhofer (ed.) Handbuch der Palaoherpetologie¨. Volume 10A, (ed.) Collected Works of International Symposium on Geological Sci- Gustav Fischer Verlag: Stuttgart and New York. ence, Peking University, Beijing, China, 1998. Seismological Press: — 1983b. The fossil record and the early distribution of lizards. Pp. 365- Beijing, China. 398 in A. G. J. Rhodin & K. Miyata (eds) Advances in herpetology — & Ji, Q. 2004. Postcranial anatomy of the Mesozoic Dalinghosaurus and evolutionary biology: essays in honor of E.E. Williams. Museum (Squamata): evidence from a new specimen of Western Liaoning. Acta of Comparative Zoology, Harvard University: Cambridge, MA. Geologica Sinica 78: 897-906. —, Gauthier, J. & de Queiroz, K. 1988. Phylogenetic relationships within — & Ren, D. 1999. [First record of lizard skin fossil from China with Squamata. Pp. 119-281 in R. Estes & G. Pregill (eds) Phylogenetic a description of a new genus (Lacertilia, Scincomorpha)]. Acta Zoo- relationships of the lizard families: essays commemorating Charles L. taxonomica Sinica 24: 114-120. Camp. Stanford University Press: Stanford, CA. —, Lu, L. W. & Bo, H. C. 2001. [New material of Yabeinosaurus tenuis Evans, S. E. 1993. Jurassic lizard assemblages. Revue de Paléobiologie (Lacertilia)], [Land and Resources] 2001(3): 41-43. [In Chinese]. (Suisse) 7: 55-65. Lee, M. S. Y. 1998. Convergent evolution and character correl- — 1994a. The Solnhofen (Jurassic: Tithonian) lizard genus Bav- ation in burrowing reptiles: towards a resolution of squamate arisaurus: new skull material and reinterpretation. Neues Jahrbuch relationships. Biological Journal of the Linnean Society 65: 369- fur Pal¨aontologie und Geologie¨, Abhandlungen 192: 37-52. 453. — 1994b. A new anguimorph lizard from the Jurassic and Lower Creta- Mateer, N. 1982. Osteology of the Jurassic lizard Ardeosaurus brevipes ceous of England. Palaeontology 37: 33-49. (Meyer). Palaeontology 25: 461-469. — 1996. Parviraptor (Squamata, Anguimorpha) and other lizards from the Nessov, L. 1988. Late Mesozoic amphibians and lizards of Soviet Middle Morrison Formation at Fruita, Colorado. Museum of Northern Arizona, Asia. Acta Zoologica Cracoviensis 31: 475–486. Bulletin 60: 243-248. Nydam, R. L. 2000. A new taxon of helodermatid-like lizard from the — 1998a. Crown group lizards from the Middle Jurassic of Britain. Pa- Albian-Cenomanian of Utah. Journal of Vertebrate Paleontology 20: laeontographica, A 250: 123-154. 285-294. — 1998b. Lepidosaurian faunas from the Early Cretaceous: a clade in — 2002. Lizards of the Mussentuchit local fauna (Albian-Cenomanian transition. Bulletin of the New Mexico Museum of Natural History and boundary) and comments on the evolution of the Cretaceous lizard Science 14: 195-200. fauna of North America. Journal of Vertebrate Paleontology 22: 645- — 2003. At the feet of the dinosaurs: the origin, evolution and early di- 600. versification of squamate reptiles (Lepidosauria: Diapsida). Biological Oppel, M. 1811. Die Ordnungen, Familien und Gattungen der Repti- Reviews, Cambridge 78: 513-551. lien, als Prodrom einer Naturgeschichte derselben. Joseph Lindauer: — & Barbadillo, L. J. 1997. Early Cretaceous lizards from las Hoyas, Munich, 87 pp. Spain. Zoological Journal of the Linnean Society 119: 23–49. Prothero, D. & Estes, R. 1980. Late Jurassic lizards from Como Bluff, — & — 1998. An unusual lizard from the Early Cretaceous of Las Wyoming, and their palaeobiogeographic significance. Nature 286: Hoyas, Spain. Zoological Journal of the Linnean Society 124: 235-265. 484–486. — & — 1999. A short-limbed lizard from the Early Cretaceous of Spain. Reynoso, V. H. 1996. Early Cretaceous lepidosaurs (Reptilia: Diap- Special Papers in Palaeontology 60: 73-85. sida) from Central Mexico and the phylogeny of lepidosaur- — & Chure, D. 1998a. Paramacellodid lizard skulls from the Jurassic omorphs. PhD Thesis: McGill University, Montreal, Canada, Morrison Formation at Dinosaur National Monument, Utah. Journal 369 pp. of Vertebrate Paleontology 18: 99-114. — 1998. Huehuecuetzpalli mixtecus gen. et sp. nov: a basal squamate — & — 1998b. Morrison lizards: structure, relationships and biogeo- (Reptilia) from the Early Cretaceous of Tepexi de Rodr´ıguez, Central graphy. Modern Geology 23: 35-48. Mexico. ´Philosophical Transactions of the Royal Society, Series B 353: — & Searle, B. 2002. The lepidosaurian assemblage of the Purbeck 477-500. Limestone Group. Special Papers in Palaeontology 68: 145-159. — & Callison, G. 2000. A new scincomorph lizard from the Early Creta- —, Lacasa-Ruiz, A. & Erill Rey, J. 2001. A lizard from the Early ceous of Puebla, Mexico. ´Zoological Journal of the Linnean Society Cretaceous (Berriasian-Hauterivian) of Montsec. Neues Jahrbuch fur 130: 183-212. Palaontologie und Geologie¨, Abhandlungen 215: 1-15. Seiffert, J. 1973. Upper Jurassic lizards from Central Portugal. Memoria, Downloaded by [National Science Library] at 02:17 22 March 2016 —, Raia, P. & Barbera, C. 2004. New lizards and sphenodontians from Serviços Geológicos de Portugal 22: 1-85. the Early Cretaceous of Italy. Acta Palaeontologica Polonica 49: 393- Smith, P. E., Evensen, N. M., York, D., Chang, M. M., Jin, F., Li, J. L., 408. Cumbaa, S. & Russell, D. 1995. Dates and rates in ancient lakes: Ar- Gao, K. Q. & Hou, L. H. 1996. Systematics and diversity of squamates Ar evidence for an Early Cretaceous age for the Jehol Group, north-east from the Upper Cretaceous Djadochta Formation, Bayan Mandahu, China. Canadian Journal of Earth Sciences 32: 1426-1431. Gobi Desert, People's Republic of China. Canadian Journal of Earth Swofford, D. L. 1993. PAUP (Phylogenetic Analysis Using Parsimony). Sciences 33: 578-598. Version 3.1.1. University of Illinois: Urbana-Champaign, IL.. — & Nessov, L. A. 1998. Early Cretaceous squamates from the — 2001. PAUP* (Version 4.06b10 for 32-bit Microsoft Windows). Sinauer Kyzylkum Desert, Uzbekistan. Neues Jahrbüch für Geologische und Associates Incorporated, Publishers: Sunderland, MA. Palaontologische Abhandlungen 207: 289-309. Townsend, T. M., Larson, A., Louis, E. & Macey, J. R. 2004. Molecular — & Norell, M. A. 2000. Taxonomic composition and systematics of phylogenetics of Squamata: the position of snakes, amphisbaenians, Late Cretaceous lizard assemblages from Ukhaa Tolgod and adjacent and dibamids, and the root of the squamate tree. Systematic Biology localities, Mongolian Gobi Desert. Bulletin of the American Museum 53: 735-757. of Natural History 249: 1-118. Wang, X. L. & Zhou, Z. H. 2003. Mesozoic Pompei. Pp. 19-35 in Hoffstetter, R. 1964. Les Sauria du Jurassique supérieur et specialement M.M. Chang, P.J. Chen, Y.Q. Wang & Y.Wang (eds) The Jehol Biota. les Gekkota de Bavière et de Mandchourie. Senckenbergiana Biologie Shanghai Scientific and Technical Publishers: China. 45,281-324. Young, C. C. 1948. A review of the Lepidosauria from China. American — 1967a. A propos des genres Ardeosaurus etEichstaettisaurus (Reptilia, Journal of Science 246: 711-719. Sauria, Gekkonoidea) du Jurassique Supèrieur de Franconie. `Bulletin — 1958. On a new locality of Yabeinosaurus tenuis Endo and Shikama. de la Societé Géologique de France 1966: 592-595. Vertebrata PalAsiatica 2: 151-156. — 1967b. Coup d'oeil sur les Sauriens (= lacertiliens) des couches de Zhou, Z. H., Barrett, P. M. & Hilton, J. 2003. An exceptionally preserved Purbeck (Jurassique superieur d'Angleterre. Résumé d' un Mémoire). Lower Cretaceous ecosystem. Nature 421: 807-814. EARLY CRETACEOUS LIZARD GENUS YABEINOSAURUS 335

APPENDIX: CHARACTER LISTS AND DATA MATRICES USED Two different character lists and data matrices were used in the analyses; they are not reproduced here due to their total length. The relevant character lists and matrices can be found in Evans & Chure (1998a) and Lee (1998), with additional data in Evans & Barbadillo (1997, 1998, 1999). Below we list the codings for those taxa that were added to each of the original matrices for this analysis.

Evans & Chure (1998a) 9 Aigialosaurus (data froom Carroll & De Braga 19 92) 11010 100??0211001?? 0?110 10?1 0100?01 10100? 1?? 02? 01 02? 01 01??1 01 10000 10001 o??oi lolll 02023 11231 00000 ?0???0 000 00 0000? 11111 11111 11010 mil " 1' 1 211?? 11111 J? ? ? J 11222 mi? ' ' 011 01112 ?2

Dorsetisaurus 100?0 00000 ollll 01000 01001 '01' ' ?100? 00? ?0 00010 00010 000?? ???1? 01121 00??? 020?? ????? ????? ????? 0000? 1???? 111?? 2?1?1 1???? ????? ????? ????? ????? 77777 0?01? ?? ? ?? ?2

Parviraptor 1 ill 0 00?11 00?? 00002 00?? ' 01" 00100 00012 11111 11111 02221 00??? 0201? 1' 0" ????? ? J J ?? 1???? 0???? ?2

Yabeinosaurus 00010 00?00 10?? 0??00 0?? 00 10?? 00??? 10??0 ???1?? 000? 0 ollio 0?010 0000101 01120 00??? 00?0 ?0 20? 0?00 ?00?0 ???00 0002? 11111

? ????? 77777 ??221 1111 ?30?1 0?122

Lee (1998) Ardeosaurus llllo ollll ?0??? ? ? 10? 0000? 011011 0000?0 000? 0 11011 00? 1? ollll ollll 0???? 77777 77777 ????? ????? ????? ????? ????? ????? ?????

77777 77777 ????? ????? ????? ????? ????? ????? ????? ?????

Bavarisaurus (in outgroup) 00000 00000 010011 00?1? 0 10011 00? 1? ollll 0???? ? ?n? ' ? ?? ? 1 ? 1 0' ' ????? ????? ????? ????? ????? 1?000 0???? ???? ?? ? ?? ? ?? ? ? ollll ?? ?? ?

Dorsetisaurus llloo o?ooo 00101 001010 ???0?000 ?0001 ????? ????? 0?000 O?O11 OOO?0 0??01 000? 1 010?? ?0001 01100 ?0100 Downloaded by [National Science Library] at 02:17 22 March 2016 1O?OO 1010? 001? 0000? 1 ? ? 1 ? 11001 001?? ?0000 00??? ? ? ? ?? 000?? 00??? 2.1111

Eichstaettisaurus llllo oilll ?0???? lloil 10011 0011 100?1?0 00?1?0 10011 ?0 1? ollll ollll ??001 02??? 11111 11111 ????? 11111 ????? 1?000 ????? ????? ????? 11111

Parviraptor llllo ollll llioo ? ? ? ? 00001 00? ?0 100?? ???1? 0?000 0???? 012??? ? ? ? ?? 00? 00 0?100 0?001 00? 00 12??? 20011 ?001? 1??00 01001 000?? 00000 ? 00? ? ????? ????? ?????

Yabeinosaurus ?1OOO OO??O 00001 00100? 0100?0 1 00001 00000 100?0 ????? 1000? 0?000 0?00 01??? 001?? ?0?00 00010 010010 100100 00010 000100 00? 00 0???? 120?? ??0?0 o?oo? ?oooo 0?010 10? 01 0?010?? 010?? 000?? 21111 ??0?0