<<

Society for the Study of and

The Oldest of Scincid Lizard () from the Etadunna Formation of South Australia Author(s): James E. Martin, Mark N. Hutchinson, Robert Meredith, Judd A. Case and Neville S. Pledge Reviewed work(s): Source: Journal of , Vol. 38, No. 2 (Jun., 2004), pp. 180-187 Published by: Society for the Study of Amphibians and Reptiles Stable URL: http://www.jstor.org/stable/1566212 . Accessed: 29/10/2012 11:56

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

.

Society for the Study of Amphibians and Reptiles is collaborating with JSTOR to digitize, preserve and extend access to Journal of Herpetology.

http://www.jstor.org Journalof Herpetology,Vol. 38, No. 2, pp. 180-187, 2004 Copyright2004 Society for the Study of Amphibiansand Reptiles

The Oldest Genus of Scincid Lizard (Squamata)from the TertiaryEtadunna Formation of South Australia

JAMESE. MARTIN,1'2 MARK N. HUTCHINSON,3 ROBERTMEREDITH,4 JUDD A. CASE,5 AND NEVILLES. PLEDGE3

1Museumof Geology,South Dakota School of Minesand Technology, Rapid City, SouthDakota 57701, USA; E-mail: [email protected] 3SouthAustralian Museum, North Terrace, Adelaide, South Australia 5000, Australia 4BiologyDepartment, University of California,Riverside, California 92521, USA 5Departmentof Biology,St. Mary'sCollege of California,Moraga, California 94575, USA

ABSTRACT.-Recentexpeditions to the LakePalankarinna area of South Australiaresulted in the oldest and only known extinct genus in Australia. The holotype of a new genus and , Proegernia palankarinnensis,was collected from the portion of the Late Oligocene EtadunnaFormation from the Minkina Local Fauna.Additional scincid fossils previously recoveredfrom higher levels in the formation include materialthat may be referableto Proegeria. Proegerniais placed in the Egeria group within the AustralianLygosominae, based mainly on its closed Meckeliangroove with the apex of the splenial notchlow on the lingual surface.However, the apex is markedlymore anteriorin Proegerniathan in any living Egernia group taxon. Proegerniapossesses characterssuggestive of its position as a stem taxon for later occurring of the Egerniaspecies groups. Even so, overall stage of evolution suggests that the Scincidae of Australiahad a long evolutionaryhistory prior to the Late Oligocene, a contention supportedby previous molecularstudies.

The Scincidae are the most abundant and continent. However, because of the isolation of diverse lizards in Australia. At least 38 genera Australia, one might also postulate rapid di- and 378 species are recognized (Cogger, 2000) in versification caused by reduced competition and Australia, and additional species are known but limited predators. Based upon immunological not yet described (e.g., Storr et al., 1999). Skink data, phylogenetic divergence of major scincid species range from 0.4 g and 35 mm snout-vent groups is postulated to have occurred during the length (SVL) to 1000 g and 371 mm SVL earliest Tertiary, suggesting divergence of the (Hutchinson, 1993). The diversity of skink habitat family during the (Baverstock and preferences includes terrestrial, fossorial, arbo- Donnellan, 1990). Unfortunately, the fossil record real, saxicoline, and semiaquatic. Skink diets is limited. A medial to late Eocene femur from the are also diverse as reflected by various Rundle Formation, central east Queensland was morphologies, some of which are characteristicof reported by Hocknull (2000), who felt that the lineages, whereas others are related to prey type specimen warranted assignment to scinco- (e.g., Estes and Williams, 1984). morphs but not enough morphological data were Skink systematics emphasize cranial features preserved for assignment to the Scincidae. Estes and include a cranium covered with enlarged der- (1984) described skinks from the Late Oligocene mal plates, paired premaxillae, and parietals with Etadunna Formation, represented by dentary, a descending process that abuts against the epi- maxillary, and parietal elements, which until pterygoids (Estes et al., 1988). In fossil deposits, now were the oldest well-documented skink cranial material is often scarce or fragmentary fossils. Estes identified these specimens as (Smith, 1976), so most fossil skinks are identified Egerniaand Tiliqua,but both were derived from by the peglike teeth and enlarged coronoid pro- higher in the stratigraphic section (Ditjimanka cess of the dentary (Estes, 1983). Although these L.F.) than that described herein. Skinks from the characters are shared with related groups (cor- early to medial Miocene Riversleigh assemblages dylids and anguids; Lee, 1998), most skinks are were described by Hutchinson (1992) and in- distinguishable by mode of tooth replacement and clude the Egerniafrerei (= Egerniamajor) species tooth-crown shape (Edmund, 1969; Lang, 1991). group, the Egernia striolata species group, and Given the diversity of Australian scincids, one Tiliqua(Hutchinson, 1992; Shea and Hutchinson, might expect a long evolutionary history on the 1992). These Egernia comparisons were based upon plesiomorphic morphology or characters dentition) that are simi- 2 (especially phenetically CorrespondingAuthor. lar to those of certain Egernia but of unclear AUSTRALIAN TERTIARYSCINCID 181

TABLE1. Comparative species examined and Eger- University of California, Berkeley, and their nia species groups assignments. colleagues from the South Australian Museum, Adelaide. The major stratigraphic unit is the Numberspecimens Etadunna Formation, named for the station Speciesgroup Speciesincluded examined (ranch) upon which the deposits and fossils were E. major E. major 2 found (Stirton et al., 1961). The formation is E. 3 frerei approximately 40 m thick (Woodburne et al., E. whitii E. whitii 4 1993:484) and consists predominately of greenish E. kintorei 2 tuffaceous siltstones and claystones interbedded E. inornata 2 with tan to white dolomitic units and some E. margaretae 3 into E. multscutata 2 matrix-supported conglomerates grading sandstones. These units lacustrine E. pulchra 2 represent E. striata 2 deposits interrupted by intervals of fluvial E. E. 3 incursions in a somewhat rhythmical succession. cunninghami cunninghami The Etadunna formational name was extended E. depressa 1 E. hosmeri 1 northward to include deposits at Lakes Kanunka, E. stokesii 4 Pitikanta, and Ngapakaldi. As a result, the E. striolata E. striolata 8 Ngapakaldi Fauna was an inclusive term for all E. napoleonis 2 in the Etadunna Formation (e.g., E. richardi 2 Stirton et al., 1961, 1968; Rich et al., 1982). E. saxatilis 2 However, Woodburne, Tedford, and others (see E. luctuosa E. coventryi 3 Aplin and Archer, 1987, and articles in that E. kingii E. kingii 1 volume) noted that the mammalian assemblages were distinct between the northern lakes and Lake Palankarinna. Consequently, the term Fauna was restricted to the northern (1992) de- Ngapakaldi evolutionary polarity. Finally, Pledge above a dolomite marker and scribed extant taxa from Plio-Pleistocene assemblages bed, depos- the Local Fauna was for the its. Because of the of the Ditjimanka proposed fragmentary vertebrates from Lake Palankarinna below a unit scincid the occurrence of well- record, relatively referred to as the dolomite" from the late "upper (Woodburne preserved specimens Oligocene, et differentiated the fossiliferous Etadunna Formation are al., 1993). Pledge (1984) richly L.F at the northwestern corner of Lake because occur in some of the Ngama important they Palankarinna from levels than those that oldest terrestrial known from higher Tertiary deposits the L.F. work Australia. produced Ditjimanka Subsequent (see Woodburne et al., 1993) indicated that an older local fauna occurred at the southwestern MATERIALSAND METHODS corner of Lake Palankarinna from the lowest The fossil described in this contribu- specimen stratigraphic levels. Consequently, Woodburne et tion was procured through field campaigns in al. (1993) subdivided the Etadunna Formation the institutions cooperation among represented into five faunal "zones" (A-E) preliminarily the authors. of tuffa- by Quarrying techniques based on stage of evolution of marsupial taxa, ceous detrital from low in the layers preserved rather than first or last appearances of taxa: the section the lower The geological produced jaw. Minkina, Ditjmanka, Ngapakaldi, Ngama, and jaw was then compared with 49 osteological Treasure local faunas. The lizard described here is of the specimens Egerniagroup (Greer, 1979) and, from the oldest fossiliferous exposures at Lake in the particular, Egerniaspecies groups (Horton, Palanakarinna, is part of the Minkina L.F. and Table of the 19 1972; Storr, 1978; 1). Specimens was presumably deposited approximately 25-26 were measured and to Egernia species analyzed Ma (= late Oligocene). quantify morphological characteristics of the splenial notch (49 specimens) and internal SYSTEMATICPALEONTOLOGY septum (24 specimens). Because of space limi- tations, these measurements and analyses are Class: Reptilia available from the authors. : Squamata Family: Scincidae STRATIGRAPHICAND CHRONOLOGIC POSITION Subfamily: Lygosominae The skink described herein was collected from Proegerniapalankarinnensis the Lake Palankarinna area in the Tirari of new genus and species the eastern Lake Eyre Basin of South Australia. Holotype.-SAM (South Australian Museum) This fossiliferous region was discovered in 1953 P39204, right dentary with 10 complete teeth and by R. A. Stirton and R. H. Tedford from the 12 alveoli or broken teeth for a total of at least 22 182 J. E. MARTIN ET AL.

l

I I

FIG. 1. Holotype of Proegerniapalankarinnensis new genus and species, SAM P39204.(A) lingual view of right dentary;(B) dorsal (occlusal)view of dentary;(C) buccal view of 18th tooth from the anteriorend; (D) occlusal view of 18th tooth; (E) lingual view of 18th tooth. teeth (Fig. 1). Abrasion within the symphyseal lian groove and a large, elongate inferior alveolar region may have removed another alveolus, sug- foramen located relatively low and anterior. The gesting the possibility of a maximum of 23 teeth. genus is distinguished from other members of Etymology.-Derived from the Greek and Latin the Egernia group (Egernia, Cyclodomorphus, pro, meaning before or forward, indicating the Tiliqua,and Corucia)by the anteriorly positioned early occurrence of an Egernia-liketaxon. Named apex of the splenial notch at 61% the anteropos- for the Lake Palankarinna area, source of the terior length of the dentary; a patent suture of the holotype; derived from Palankarinna, an aborig- Meckelian groove trending forward from the inal (Diyari) word, meaning place of intercourse, inferior alveolar foramen; and the anteriorly and the Latin suffix ensis, meaning native or excavated caudal margin of the exposed internal resident in. septum. Type Locality.-Young Bucks Quarry (Univer- Description.-The type specimen (Fig. 1A) is sity of California, Riverside Locality RV-9002) in 10.05 mm from the anterior tip to the end of the Lake Palankarinna area, 30 km south of the angular process and exhibits several cycles Etadunna Station; 28?47' S Latitude, 138o25' E of tooth replacement. The Meckelian canal is Longitude; South Australia, Australia. closed anterior to the inferior alveolar fora- Stratigraphic Position.-Lower Etadunna For- men by overgrowth of the dentary. However, mation, Zone A (Woodburne et al., 1993). a remnant millimeter-long groove occurs at Age.-Late Oligocene, 24-26 Ma (Woodburne the symphysis. The contact between the dentary et al., 1993). and coronoid is robust and rises steeply as a Diagnosis.-The holotype is a member of the distinct, posterodorsally directed coronoid Egerniagroup (Greer 1979), with a closed Mecke- process. AUSTRALIAN TERTIARYSCINCID 183

Lingually (Fig. 1A), the splenial notch is group (3 genera), and the Eugongylus group (16 completely open 6.5 mm from the distal end of genera). The Sphenomorphusgroup is character- the angular process (5.3 mm from the indenture ized by the retained plesiomorphic open Mecke- between the angular and coronoid processes and lian groove occurring along the ventromedial 6.1 mm from the end of the coronoid projection) margin of the dentary below the anterior half of to a position below the 12th preserved tooth from the tooth row (lost several times within the the anterior end (llth from the posterior end). At group), as well as by an apomorphic parallel- this point the dentary is 2.5 mm deep; its sided retroarticular process. The Egernia group maximum height at the posterior end between (following the restricted of Greer, 1979; see the angular and coronoid processes is 2.8 mm. Hutchinson, 1993:278) exhibits an anteriorly However, a patent groove extends anteriorly for closed Meckelian groove and a large, elongate a further 2.1 mm. The lower edge of the internal inferior alveolar foramen but is otherwise ple- septum that separates the Meckelian groove and siomorphic in having the foramen located rela- alveolar canal terminates at a point below the tively low and anterior, possessing a relatively posterior end of the tooth row, but its margin is large splenial, and exhibiting an inflected retro- deeply excavated. The excavation reaches the articular. The Eugongylus group has a reduced or 14th tooth, 4.25 mm from the distal end of the absent angular, reduced splenial, and an inflected angular process (3.3 mm from the indenture retroarticular. Within the Lygosominae, Proeger- between the coronoid and angular processes and nia is excluded from the Sphenomorphusgroup 3.75 mm from the tip of the coronoid process). owing to the anteriorly closed Meckelian canal Labially, the jaw is smooth with eight nutrient and from the Eugongylus group because of its foramina extending 4.1 mm from the anterior jaw large splenial notch. Proegerniamay be referred to tip to below the 13th preserved tooth from the the Egerniagroup because of its anteriorly closed anterior end. Meckelian groove and large splenial. The jaw has alveoli for 22 pleurodont teeth, The Egernia group includes approximately 50 although only 10 are preserved (Fig. 1B). How- species within three Australian genera (Cyclo- ever, a small portion of the symphyseal region domorphus,Tiliqua, and Egernia) and one in the has been abraded as the anterior opening of the Solomon Islands (Corucia). Corucia has laterally Meckelian canal is visible. Perhaps a single tooth flattened teeth with fan-shaped crowns (Greer, was lost during abrasion, suggesting that the 1976), unlike the columnar teeth of Egernia or maximum tooth number may have been 23. Proegernia(Estes, 1984:339-340). Cyclodomorphus Some crowding of the first four teeth occurs. The and Tiliquaare sister groups of large skinks with teeth are seemingly more widely spaced than durophagous dentitions (Estes and Williams, those of most Egerniaspecies. All Proegerniateeth 1984), again unlike Egernia or Proegernia.Within are wide at the bases and taper slightly upward. the Egernia group, the majority of species are The tooth crowns are thumb-shaped, with placed in the genus Egernia. Members of this curved lateral face and flattened medial face. genus are currently defined by lacking the The two surfaces meet and form an anteropos- specializations of dentition and by the vertebral teriorly oriented apical ridge. In lateral view, the and cranial features that diagnose the other three profile of the apical ridge forms a shallow, genera (e.g., Shea, 1990). The species of Egernia inverted asymmetrical vee, with the apex offset have been placed into six species groups (Horton, posteriorly (Fig. 1C-E). No distinct apical cusp 1972; Storr, 1978), based on external morphology occurs, nor any trace of the bicuspid crown and scalation: the Egernia cunninghamii, Egernia morphology regarded as plesiomorphic for scin- major, Egernia striolata, Egernia kingii, Egernia cid lizards (Sumida and Murphy, 1987). Distinct whitii, and Egernia luctuosa species-groups (See striations trend vertically on both lingual and Table 1). labial faces of SAM P39204. Relatively little The size and shape of the holotype dentary differentiation of crown pattern exists along the compares favorably with members of several tooth row. species groups of Egernia, particularly those in Comparisons.-The majority of extant Austra- the E. striolataand E. luctuosa species groups that lian skink species are part of the Lygosominae possess tooth crowns with slightly flared profiles, (Greer, 1970), which is characterized by fused with a continuous lingual apical ridge lacking an frontals, a secondary palate formed by expanded apical cusp, and with only modest differentiation ventral laminae of the palatines, and an angular from the crowded anterior teeth to larger contact between the ventrolateral ridge of the posterior teeth. The fossil specimen is approxi- frontal or its ventral process and the prefrontal mately the size of most specimens of the E. (Greer, 1970, 1986). Lygosomines comprise five striolata, E. whitii, and E. luctuosa species groups major lineages (Honda et al., 2000), of which but much smaller (normally -50%) than the three are Australian (Greer, 1979, 1989): the E. major, E. cunninghami, and E. kingii species Sphenomorphusgroup (13 genera), the Egernia groups. With 22 teeth represented in the 184 J. E. MARTIN ET AL. dentition, the fossil lies within the range of most from the posterior end of the coronoid process. dentitions of Egernia, except E. cunninghami, E. This compares with 37% of the dentary length in kingii, and some members of the E. whitii species the fossil taxon. The extant species in which the groups, which normally possess more. Within extent of the septum is most similar to that of the Egernia, tooth morphology varies markedly. fossil taxon is E. frerei. However, statistical Members of the E. whitii species group have analysis indicates that the length of the internal narrow-crowned teeth, with an apical cusp and septum measured from the coronoid process to angular rather than smoothly rounded corners at the septum apex compared to dentary length of the anterior and posterior ends of the apical ridge this living species is significantly smaller than where it meets the tooth shaft. The E. luctuosa that of the fossil species (probability 0.0019 at species group has similar tooth morphology with 95% C.L.). an apical cusp, but the marginal corners are more With greater sample size, two other characters rounded and the teeth are more laterally flat- may become important in differentiating Pro- tened. Largely herbivorous species of the E. egernia from other species groups. Observed cunninghamii species group have small, closely specimens of E. coventryi, E. kingii, and some E. packed teeth with laterally compressed crowns stokesii have a relatively longer, thin angular having a sharp cutting edge. In some of the larger process compared to the short, stout process of species, such as E. kingii and E. frerei (E. major Proegerniaand most other Egernia species. Also, species group), the teeth are robust with crowns the holotype dentary exhibits eight mental expanded both anteroposteriorly and lingually. foramina, whereas most other species exhibit This morphology also develops later ontogenet- six to seven. ically in E. striolata (Hutchinson, 1990), but in Overall, the dental morphology and closure of both E. striolataand the larger species mentioned, the Meckelian groove indicate inclusion of the younger individuals show the same tooth mor- fossil taxon in the Egernia group, but the phology exhibited by the fossil specimen. The combination of features in the type specimen, tooth crown shape of SAM P39204 is the basic especially the anteriorly extended splenial notch pattern that has been modified within Egernia,as and the anteriorly positioned posterior margin of currently defined. the internal septum differentiate this taxon from The splenial notch and the internal septum are known Australian skinks. The anterior extent of distinctive in P. palankarinnensis,although the the splenial notch may be plesiomorphic, repre- characteristics are approached by some members senting a lesser extent of closure of the Meckelian of the Egerniaspecies groups. The anterior extent groove than in living Egernia.The margins of the of the splenial notch and anterior position of the bone at the apex of the splenial notch are smooth, inferior alveolar foramen are more extreme than without articular facets for the surangular and in any Egerniaspecies we examined. Most species splenial, and indicate the opening of the inferior of Egernia show the anterior termination of the alveolar foramen. This implied position for the foramen/splenial notch at a point between 20 inferior alveolar foramen is further forward and 52% (average = 34%) from the posterior compared to that of any living Egernia,and also margin of the coronoid process, rather than the compared to other skinks, even those without 61% as in the fossil. The closest approach to the a closed Meckelian groove. Simple closure of the condition in the fossil is seen in Egerniacoventryi Meckelian groove would not be expected to leave (Egernia luctuosa species group), in which the the foramen in this topological position, imply- anterior alveolar foramen/splenial notch is 40% ing that the anteriorly placed inferior alveolar the dentary length measured from the posterior foramen of P. palankarinnensisis an apomorphic tip of the coronoid process and in E. major in character state. which the splenial notch may be 52% of the DISCUSSION dentary length; however, the latter species is three times the size of P. palankarinnensis. Proegerniarepresents the oldest, unequivocal Moreover, the splenial notch length-overall jaw skink known from Australia, based upon the length ratio of E. majorspecies group and E. major smoothly closed Meckelian groove, enlarged itself (exhibiting the largest noted ratio among coronoid process, and dental morphology (Estes, extant species) is significantly smaller than that 1983, 1984). Although a closed groove occurs in of P. palankarinnensis(probability 0.0001 at 95% members of the , relatively few, Confidence Level). The internal septum dividing robust teeth differentiate Proegernia.Moreover, the Meckelian and alveolar canals is deeply features such as the anteriorly open splenial excavated anteriorly on the holotype, and the notch, incomplete internal septum, generalized ventral portion extends to near the end of the tooth crowns, and small size suggest this species tooth row. In other species of Egernia,the anterior could be ancestral to later skinks of the Egernia extent of the septum ranges from 14-32% group. However, having more than 60% of the (average 22%) of the dentary length measured dentary exposed by the splenial notch, compared AUSTRALIAN TERTIARYSCINCID 185 to 50% or less for known species of Egernia and The occurrence of P. palankarinnensisin depos- an incomplete internal septum separating the its of late Oligocene age sheds some light Meckelian canal from the alveolar canal suggests concerning possible ancestry and biogeography. a lineage separated from the extant genus. As mentioned above, the closed Meckelian Estes (1984) considered specimens from the groove and dental morphology indicate the Etadunna Fm. (Ditjimanka L.F.) as skinks based taxon is a basal member of the Egernia group. on one apomorphic character state, closure of the This occurrence confirms that this lineage dates Meckelian groove, combined with plesiomorphic to at least 25-26 Ma, and some time may have character states, notably separation of the parie- elapsed between the arrival of the scincid tal scales by the interparietal, as evinced from the ancestor in Australia and the appearance of adherent to the parietal, relatively Proegernia. Living skinks are concentrated in unmodified teeth, and lack of obvious special- the Southern Hemisphere, with most of the izations. The holotype of P. palankarinnensisand surviving stem-group "scincines" restricted to Estes' specimens are distinguishable from any Gondwanan land masses (southern , living Egerniaspecies but strikingly similar to one Madagascar, and ). The oldest fossils re- another. We can assess this similarity only in ported to be scincids are from North America, a preliminary manner as Estes provided a single but the evidence for their familial assignments is line drawing of each element (dentary in lingual weak. Estes (1969) described Contogenys sloanei view, maxilla in lateral view, and parietal in from the () dorsal view), and these illustrations leave a num- and Tongue River Formation (Medial Paleocene) ber of characters indeterminable. He also did not of Montana as a scincid, but later (Estes, 1983) provide meristic data aside from a bar treated the scincid attribution of both Contogenys accompanying the drawings. Moreover, searches and Paracontogenys (Late Eocene) as tentative, of both the University of California, Riverside, mainly because of the paucity of characters and San Diego State University collections have provided by the material. Gao and Fox (1996) failed to locate Estes' material. Given these discussed these genera and concluded that Para- limitations, the list of morphological similarities contogenys was most likely a xantusiid (probably that can be gleaned from the single drawing is congeneric with Palaeoxantusia)and Contogenys significant. The two are apparently identical in also was more similar to xantusiids than scincids. length (estimated 10 mm) and are immediately Gao and Fox (1996) also allocated one of their distinguished from extant Egernia species by the new taxa, Penemabuyaantecessor, to the Scincidae, anteriorly placed inferior alveolar foramen situ- and identified two others, Orthrioscincusmixtus ated near the midpoint of the tooth row (in- and Aocnodromeuscorrugatus as probable scin- complete anteriorly in Estes' specimen). The cids. In our view, familial assignment of any of foramen (apex of the splenial notch) terminates the three genera is unwarranted, as the differen- below the 11th tooth locus from the rear of the tial characters used are not restricted to the tooth row in both and is approximately 5 mm scincids, nor even typical of scincids in some from the anterior limit of the indenture between cases. These taxa are best regarded only as the coronoid and angular processes. The depth of scincomorphans and may be closer to cordylids the jaw at the level of the inferior alveolar fora- than to scincids. men (approximately 2 mm in Estes' specimen, 2.5 Molecular divergence among major skink in SAM P39204) and at the posterior end of the lineages is profound (Baverstock and Donnellan, tooth row (approximately 3 mm in Estes' speci- 1990; Honda et al., 2000) and suggests divergence men, 2.8 in SAM P39204) indicates that pro- of the family during the Paleocene or earlier. The portions as well as absolute size are comparable. likelihood is that this divergence occurred in Additional qualitative similarities include a dor- Gondwana where great numbers of skinks now sal groove for contact with the splenial above the occur, although generally recognized less-de- inferior alveolar foramen, steeply rising contact rived skinks such as Eumeces and their relatives facet for the coronoid, and relatively strong are from the Northern Hemisphere. The fossil striations on the lingual faces of the tooth crowns. record is becoming better known, and the Other potential similarities, such as the morphol- recognition of extant lineages already differenti- ogy of the internal septum and detailed mor- ated 25 million ago pushes the ultimate phology of the tooth crowns are not determinable origin of these lineages, and therefore of skinks as from Estes' illustrations. On current evidence, a whole, further back in time. The possible Estes' assignment to Egernia sp. should be existence of North American fossil scincids regarded as Proegerniarather than Egernia (s.l.). suggests that a northern origin is a persistent Specific attribution is not advisable until the possibility, but much of the differentiation specimens are found, given that important evidently occurred in the south. If a northern distinguishing features of P. palankarinnensisare origin is substantiated, a corridor through - not yet known for Estes' younger specimens. arctica (Case et al., 2000) is now recognized 186 J. E. MARTIN ET AL. during the Late Cretaceous and may have pro- ing Charles L. Camp. Stanford Univ. Press, vided the route by which skinks dispersed. Stanford, CA. GAO, K., AND R. C. Fox. 1996. of late Acknowledgments.-We thank A. Greer and Cretaceous lizards (Reptilia: Squamata) from west- anonymous reviewers, who made valuable ern Canada. Bulletin of Carnegie Museum of Natural 33:1-107. suggestions that resulted in a much better History GREER,A. E. 1970. A subfamilial classification of scincid contribution. We also thank M. Gabel, Depart- ment of Black Hills State lizards. Bulletin Museum of Comparative Zoology, Biology, University, Harvard 139:151-184. who a amount of time and lent spent great . 1976. On the evolution of the giant Cape Verde expertise to produce the SEM photographs for scincid lizard Macroscincuscoctei. Journal Natural illustrations herein. The Soci- Royal Geographic History 10:691-712. ety of London/Discovery Channel-Europe pro- . 1979. A phylogenetic subdivision of Australian vided partial field funding through a prize skinks. Records Australian Museum 32:339-371. awarded to the first author. The National Science 1986. Lygosomine (Scincidae) monophyly: Foundation (OPP 0087972) provided ancillary a third corroborating character and a reply to funding for study of Gondwanan fossil reptiles critics. Journal of Herpetology 20:123-126. . 1989. The and Evolution of Australian through grants awarded to the first author. Biology Lizards. and Additional funding through the Faculty De- Surrey, Beatty Sons, Pty. Limited, Fund was St. Chipping Norton, New South Wales, Australia. velopment provided by Mary's S. A. 2000. Remains of an Eocene skink to the fourth author. We thank K. HOCKNULL, College Aplin, from Queensland. 24:63-64. Western Australia Museum, for loan of Recent Alcheringa HONDA, M., H. OTA, M. KOBAYASHI,J. NABHITABHATA, H.- and M. comparative specimens O. Woodburne, S. YONG,AND T. HIKIDA.2000. Phylogenetic relation- of for loan of University California, Riverside, ships, character evolution, and biogeography of the additional squamate specimens. subfamily Lygosominae (Reptilia: Scincidae) in- ferred from mitochondrial DNA sequences. Molec- ular and Evolution 15:452-461. LITERATURECITED Phylogenetics HORTON, D. R. 1972. Evolution in the genus Egernia. APLIN, K. P., AND M. ARCHER.1987. Recent advances in Journal Herpetology 6:101-109. marsupial systematics with a new syncretic classi- HUTCHINSON,M. N. 1992. Origins of the Australian fication. In M. Archer (ed.), Possums and Opos- scincid lizards: a preliminary report on the skinks sumes: Studies in Evolution, pp. xv-lxxi. Surrey of Riversleigh. Beagle, Records Northern Beatty and Sons, Sydney, New South Wales, Museum of Arts and Sciences 9:61-70. Australia. . 1993. Family Scincidae. In C. J. Glasby, G. J. B. BAVERSTOCK,P. R., AND S. C. DONNELLAN. 1990. Ross, and P. L. Beesley (eds.), Fauna of Australia. Molecular evolution in Australian and dragons Vol. 2A. Amphibia and Reptilia, pp. 261-279. skinks: a Memoirs progress report. Queensland Australia Government Publishing Service, Can- Museum 29:323-331. berra, Australian Capital Territory,Australia. E. MARTIN,D. S. CHANEY,M. REGUERO,S. CASE, J. A., J. LANG, M. 1991. Generic relationships within Cordyli- A. S. M. AND M. O. WOOD- MARENSSI, SANTILLANA, formes (Reptilia: Squamata). Bulletin de l'Institute BURNE. 2000. The first Duck-Billed (Family Royal des Sciences Naturelles de Belgique, Biologie Hadrosauridae) from . Journal 61:121-188. 20:612-614. Paleontology LEE,M. S. Y. 1998. evolution and character and of Convergent COGGER,H. G. 2000. Reptiles Amphibians correlation in towards 6th ed. Reed New burrowing reptiles: [sic] Australia. Holland, Sydney, resolution of New South Wales, Australia. squamate relationships. Biological Journal Linnean 65:369-453. EDMUND, G. 1969. Dentition. In C. Gans, A. d'A. Society PLEDGE,N. S. 1984. A new Miocene vertebrate faunal and T. Parsons (eds.), of the Bellairs, Biology from the Lake Basin: a 117-200. Academic Press, London. assemblage Eyre preliminary Reptilia, pp. Australian 21:345-355. R. 1969. A scincoid lizard from the Cretaceous report. Zoologist ESTES, .1992. The Curramulka local fauna: a late and Paleocene of Montana. Breviora 331:1-9. . 1983. Handbuch der Tertiary fossil assemblage from Yorke Peninsula, palaoherpetologie. Ency- South Australia. The 9:115-142. clopedia of Palaeo-herpetology. Part 10A. Beagle RICH,T. H., M. ARCHER,M. PLANE,T. F. FLANNERY,N. S. terrestria, . Gustav Fischer Verlag, PLEDGE,S. HAND,AND P. V. RICH.1982. Australian Stuttgart, Germany. localities. In P. V. Rich and E. M. . 1984. Fish, amphibians and reptiles from the Tertiary Etadunna Formation, Miocene of South Australia. Thompson (eds.), The Fossil Vertebrate Record of 526-572. Monash Univ. Offset Australian Zoologist 21:335-343. Australasia, pp. ESTES,R., AND E. E. WILLIAMS.1984. Ontogenetic Printing Unit, Clayton, Victoria, Australia. variation of molariform teeth of lizards. Journal of SHEA, G. M. 1990. The genera Tiliqua and Cyclo- Vertebrate Paleontology 4:96-107. domorphus(Lacertilia: Scincidae): generic diagnoses ESTES, R., K. DE QUEIROZ, AND J. GAUTHEIR. 1988. and systematic relationships. Memoirs Queensland Phylogenetic relationships within Squamata. In R. Museum 29:495-520. Estes and G. Pregill (eds.), Phylogenetic Relation- SHEA, G. M., ANDM. N. HUTCHINSON.1992. A new ships of the Lizard Families: Essays Commemorat- species of Tiliqua (Lacertilia: Scincidae) from the AUSTRALIAN TERTIARYSCINCID 187

Miocene of Riversleigh. Memoirs Queensland STORR,G. M., L. A. SMITH,AND R. E. JOHNSTONE.1999. Museum32:303-310. Lizards of Western Australia. I Skinks. Western SMITH,M. J. 1976.Small fossil vertebratesfrom Victoria Australia Museum, Perth, Western Australia, Cave, Naracoorte, South Australia IV. Reptiles. Australia. Transactionsof the RoyalSociety of SouthAustralia SUMIDA,S. S., AND R. W. MURPHY.1987. Form and 100:39-51. function of the tooth crown structurein gekkonid STIRTON,R. A., R. H. TEDFORD,AND A. H. MILLER.1961. lizards (Reptilia, Squamata, Gekkonidae).Cana- Cenozoicstratigraphy and vertebratepaleontology dian Journalof Zoology 65:2886-2892. of the TirariDesert, South Australia. Records South WOODBURNE,M. O., B. J. MACFADDEN,J. A. CASE,M. S. AustralianMuseum 14:19-61. SPRINGER,N. S. PLEDGE, J. D. POWER,J. M. WOOD- STIRTON,R. A., R. H. TEDFORD,AND M. O. WOODBURNE. BURNE,AND K. B. SPRINGER.1993. Land mammal 1968. AustralianTertiary deposits containing ter- biostratigraphy and magnetostratigraphy of the restrialmammals. Univ. of CaliforniaPublications Etadunna Formation (Late Oligocene) of South GeologicalSciences 77:1-30. Australia. Journal Vertebrate Paleontology 13:483- STORR,G. M. 1978. The genus Egernia (Lacertilia: 515. Scincidae) in Western Australia. Records of the South Australian Museum 6:147-187. Accepted: .

Journalof Herpetology,Vol. 38, No. 2, pp. 187-196, 2004 Copyright 2004 Society of Amphibians and Reptiles

Taxonomy and Geographic Variation in the Leaf-Nosed Phyllorhynchus decurtatus (Squamata:)

STEPHANIEA. GARDNER AND JOSEPHR. MENDELSON III1

Departmentof Biology,Utah State University, Logan, Utah 84322-5305, USA

ABSTRACT.-Thepreviously recognized subspecies (used here as a heuristic tool) of Phyllorhynchus decurtatus artificially compartmentalize the inconsistent geographic variation in this species. Principal component analysis did not identify consistent geographic groupings of individuals. Regression Analysis revealed a longitudinal dine in number of ventral scales, as well as some weaker clinal trends in other characters.There is extensive overlap in all morphological and color-patterncharacters examined between four subspecies of P. decurtatusand their intergrades.There are differences between mean scale and blotch counts between previously recognized subspecies, but they do not representdiscrete differences among these taxa. Discriminant Analysis demonstratedthe inconsistency in the observed geographic variation in this species. We conclude that P. decurtatus represents a geographically variable species lacking consistent geographic patternclasses.

Phyllorhynchus decurtatus was first described chus decurtatus decurtatus and found overlap in by Cope (1868) as Phimothyradecurtata. Stejneger the diagnostic characters of these two subspecies; (1890) later referred the taxon to the genus he placed P. d. arenicolaas a junior synonym of P. Phyllorhynchus. Five subspecies of P. decurtatus decurtatus. Stebbins (2003) did not incorporate and their distributions have been described (Fig. any subspecific designations for P. decurtatus,and 1; McCleary and McDiarmid, 1993). McCleary Grismer (2002) referred all subspecific taxa on the and McDiarmid (1993) suggested that most of Baja California Peninsula to simple pattern these putative subspecies may actually be eco- classes. However, neither of these two authors morphs resulting from differences in incubation included explicit analyses of data. temperature. Similarly, Grismer (1999) compared Distinct evolutionary species may go unrecog- Phyllorhynchusdecurtatus arenicola to Phyllorhyn- nized among the variation in widespread species, and closer evaluation of subspecific taxa may reveal the existence of species (see discussions by 1 CorrespondingAuthor. E-mail: [email protected]. Burbrink et al., 2000). In any case, studies of edu geographic variation inform the discovery of