A new giant python from the Pliocene Bluff Downs ...

A new giant python from the Pliocene Bluff Downs Local Fauna of northeastern Queensland JOHN D. SCANLON AND BRIAN S. MACKNESS SCANLON,1. D.' & MACK NESS, B. S.'. 2002 :0 1:31. A new giant python from the Pliocene Bluff Downs Local Fauna of northeastern Queensland. Alcheringa 25, 425-437. ISSN 0311-5518 .

Liasis dubudingala n. sp., described on the basis of isolated vertebrae from the Early Pliocene Bluff Downs Local Fauna, is the largest snake known from Australia. Dependance of vertebral proportions on intracolumnar position indicates that the fossil taxon can be excluded from the MorelialPython clade. High neural spines suggest possible affinity with Liasis olivacea, whereas a posterior dentary fragment with small teeth is unlike L. olivacea and more similar to Liasis mackloli or species of Bothrochilus and Leiopython. As these extant species have all recently been treated as members of Liasis, the new species is assigned to that genus. 'John D. Scanlon & 2Brian S. Mackness . School oj Biological Sciences, University oj New South Wales. NSW, 2052. 'Present addresses: Department oj Palaeontology, South Australian Museum, North Terrace. Adelaide. South Australia 5000; and 2P.o. Box 560. Beerwah.Queensland 4519; receil'e d 16 Jun e. 1999; accepted 2 September. 2 ()()O. 'corresponding author; fe-mail: ·scanlonjohn @Sa ugol..sa.gol·. au} Keywords: Serpentes, Booidea, pythonines, Bluff Downs Local Fauna, Pliocene,

FOSSIL SNAKES have been recovered from Tertiary deposits in many parts of Australia, representing three distinct lineages: madtsoiids, pythons and elapids. The earliest records are madtsoiids from the Eocene Tingamarra Fauna of Eastem Queensland (Scanlon 1993), with other madtsoiids coming from: the Miocene of the Northern Territory (Scanlon 1992) and Queensland (Scanlon 1992, 1997; Scanlon & Lee 2000); the Pliocene of South Australia (Pledge 1992) and Queensland (Mackness & Scanlon 1999); and the Pleistocene of Western Australia (Merrilees 1979, Flannery 1989), South Australia (Smith 1976, Barrie 1990), New South Wales (Scanlon 1995) and Queensland (McNamara 1990). Two large pythons, Montypythonoides riversleighensis and Morelia antiqua, were described by Smith & Plane (1985) from the Miocene deposits of Riversleigh, Queensland and Bullock Creek, Northern Territory respectively. Kluge (1993) synonymisedMorelia 0311/5518/2001 /04425-\3

$3.00

© AAP

antiqua with Liasis olivacea Gray, 1842 and Montypythonoides riversleighensis with Morelia spilota (Lacepede, 1804) but Scanlon (in press) recognised both of the Smith & Plane species as representing a single extinct species Morelia riversleighensis, distinct from but closely related to extant Morelia spp. Archer (1976) noted, but did not name, three vertebrae (including QM F7775) which he compared with modern Morelia Gray, 1842, as well as two small vertebrae (including QM F7826) which he compared with species of Pseudechis Wagler, 1830. These smaller vertebrae are presently under study by one of the authors (BM). Further elapid fossils have been reported from the Miocene of Queensland (Scanlon 1992, 1995b), Pliocene of South Australia (Pledge 1992) and several Pleistocene deposits (Smith 1975, Archer, 1976, McNamara 1990, Scanlon 1995a). Liasis dubudingala n. sp. is described here on the basis of isolated vertebrae from the Early Pliocene freshwater fluviatile and lacustrine deposits of the Allingham Formation, northwest

}()III' II. SCANt ON AND ORlAN S. MACKNI SS

ALCIiEIt l.VGA

ALCHERINGA

GIANT PLIOCENE PYTHON FROM QUEENSLAND

of Charters Towers, northeastern Queensland. A number of reptilian taxa have been reported from this formation (Archer 1976, Mackness & Hutchinson 2000, Mackness & Sutton 2000, Thomson & Mackness 1999, Willis & Mackness 1996). Collectively this assemblage has been called the Bluff Downs Local Fauna (Archer 1976).

Materials and methods Fossil remains of reptiles were obtained through quarrying or through wet sieving of sediments. Terminology for vertebrae follows Auffenberg (1963) and Hoffstetter & Gasc (1969) . Abbreviations for specimen numbers: QMF, Queensland Museum Fossil Collection; MM, Macleay Museum, University of Sydney; AR, Research Collection, University of New South Wales.

Systematic palaeontology Order SERPENTES Linnaeus, 1758 Family BOIDAEGray, 1825 Subfamily PYTHONINAE Fitzinger, 1826 Liasis Gray, 1842 Type species. L. mack/oli Dumeril & Bibron, 1844 Included species. For the purposes of the present paper, the genus is broadly defined (cf. Stimson 1969, McDowell 1975 and Cogger el a/. 1983) as equivalent to the genera Anlaresia (L. childreni Gray, 1842; L. macu/osa Peters, 1873; L. perthensis Stull, 1932; L. slimsoni Smith, 1985), BOlhrochilus (L. boa Schlegel, 1837), Leiopylhon (L. a/berlisii Peters & Doria, 1878), Liasis (L. mack/oli Dumeril & Bibron, 1844, L. olivacea Gray, 1842) and Apodora (L. papuana Peters & Doria, 1878) in Kluge (1993). The best-fitting phylogenetic hypotheses based on a large amount of character evidence (Kluge 1993) indicate that Liasis, in the

427

sense used here, is paraphyletic. However, a hypothesis of monophyly is most parsimonious when non-skeletal characters are excluded (Kluge 1993, fig. 23) and the group may be diagnosed on this basis. Kluge's genera can be considered subgenera, but only Antaresia will be so used here. Contrary to most authors (and the International Commission on Zoological Nomenclature 1988), we take the first use of the generic name (in Liasis olivacea Gray, 1842) as establishing the gender of Liasis to be feminine. Diagnosis. The following diagnosis is based on data in Kluge (1993), with the cladistic relationships of species as in his fig. 23 (species noted as exceptions represent reversals) . Numbers refer to characters and states defined by Kluge; synapomorphies are listed only if they do not depend on assumptions of character state order (additivity).

5(1), ventral openings for the premaxillary channels anterior to the posterior margin of the premaxillary teeth (except usually L. a/bertisii, see Kluge 1993: 14); 31 (I), supraorbital (postfrontal of most authors) widely separated from the parietal; 45(1), quadrate short, less than 21 % of total jaw length. 75(5), average adult tail length 4.5 or more times total head length (less than 4.0 (3) in L. a/bertisii and L. boa, less than 3.0 (1) in species of Antaresia); 76(2), single pair of enlarged parietal scales usually in contact on the midline (more than one pair (3) in L. boa and most L. mack/oli); 81 (0), preocular scale single (species of Antaresia usually with two preoculars (I »; 115(2 or 3), awn originating abruptly from hemipenial lobe (gradually (I) in species of Anlaresia); 116( 1), calyces limited to distal end ofhemipenial lobes (except species of Antaresia); 1I8( 1), three or fewer distal flounces on the

Fig. I. Liasis dubudingala n. sp. A-E, holotype QM F9132, mid-trunk vertebra. A, lateral; B, anterior; C, posterior; D, dorsal; E, ventral views. F-J, QM F2366 I , anterior trunk vertebra. F, lateral; G, anterior; H, posterior; I, dorsal; J, ventral views. K-O, OM F30586, juvenile posterior trunk vertebra. K, lateral; L, anterior; M, posterior; N, dorsal; 0, ventral views. P, QM F 30587, juvenile dentary. Seale bar = I em

428

JOHN D. SCANLON AND BRIAN S. MACKNESS

hemipenis (four or more (0) in L. albertisii, L. maculosus and L. stimsoni). Liasis dubudingala n. sp. (Fig. 1, Table 1) 1976

?Morelia sp. Archer & Wade, p. 395, fig. 54k

Holotype. QMF 9132, a mid-trunk vertebra collected by M. Archer in 1975. Type Locality. Main Quarry (Lat. 19° 43 's, Long. 145° 36'E), Allingham Formation, BIuffDowns Station, northeastern Queensland. Age. Early Pliocene, based on the interpreted age of the overlying Allensleigh Basalt (Archer & Wade 1976, Mackness et al. 2000). Etymology. The specific name is a conjunction of two words from the Gugu-Yalanji dialect - dubu, meaning 'ghost' and dingalmeaning 'to squeeze' (Oates et al. 1964); i.e. 'ghost squeezer', in reference to the presumed constricting habits of the snake when alive. Diagnosis. A very large pythonine, with neural spine higher than long, back-sloping, and usually with blunt (not acute) dorsoposterior angle in lateral view. Junction of neural spine with zygosphene roof not sharp. Zygosphene roof steeply sloping, not sharply demarcated from anterior face of zygosphene or from base of neural spine. Zygapophyses close to horizontal (sloping at less than 5° at mid-trunk). Hypapophysis depth reduces gradually from anterior to posterior, rather than sharply at transition to mid-trunk region, and haemal keel well-defined throughout trunk. Referred material. QM F7773 (ptv, Main Quarry); QM F7775 (ptv, Main Quarry); QM F9138 (mtv, Main Quarry); QM F9134 (ptv, Main Quarry); QM F9139 (mtv, Site 2, bend 9km SE of homestead); QM F23661-7 (atv, DML site); QM F23668 (mtv, DML site); QM F23669 (neural arch fragment, DML site); QM F23670 (ptv, DML site); QM F23671 (neural arch fragment, DML site); QM F23672 (ptv centrum, DML site); QM F23673 (ptv DMLsite); QMF23674 (ptv,JHY site); QM F23675

ALCHERINGA

(4 vert fragments, 3 rib heads, DML site); QM F23677 (3 fragments, Main Quarry); QM F23678 (1 rib head, DML site); QM F23679 (3 bags of fragments, DML site); QM F30586 ptv, Main Quarry); QM F30587 (dentary, DML site); QM F30588 (isolated teeth, Main Quarry); QM F30589 (vertebra, Main Quarry); QM F30590 (four vertebral fragments, Main Quarry). Description. The largest vertebra (holotype QM F9132) evidently comes from the middle trunk region. The right prezygapophysis, left postzygapophysis and right side of the zygosphene are broken, and cracks extend through the neural arch. The vertebra is massive and high, with the centrum, neural arch and neural spine each contributing about equally to the total height. In lateral view, the vertebra is much higher than long. The neural spine is slightly higher than long, beginning anteriorly near the anterior edge of the zygosphene; its anterior and posterior margins are nearly straight (slightly convex), and practically parallel to each other. The dorsal edge of the spine is smoothly continuous with the anterior edge, convex dorsally and slightly lower posteriorly, and forms an obtuse angle with the posterior edge. The posterior edge of the neural arch is vertical; its oblique dorsoposterior edge (above the zygantrum) extends dorsally as a crest on the lateral face of the neural spine to just dorsal and posterior to the centre. The zygosphenal facets are subelliptical, with the long axes approximately vertical. The interzygapophyseal ridge is thick and well-developed. The paradiapophyses are partially worn, but clearly divided (by a posterior constriction) into convex dorsal and convexo-concave ventral portions; the subcentral ridges are short and ventrally concave,extending to near the base of the cotyle. The haemal keel projects below the centrum in its posterior half, with a sigmoid ventral edge and vertical posterior edge below the centre of the condyle. In dorsal view, the vertebra is distinctly wider than long. The posterior border of the neural arch is broadly notched, but strongly overhung by the neural spine. The dorsal surface of the spine is subrectangular, about three times as long as

ALCHERINGA


broad. The interzygapophyseal ridges are nearly straight but converging posteriorly and sharply demarcated from pre- and postzygapophyses. The zygosphene has prominent lateral and median anterior lobes; its dorsal surface is not flat but divided by transverse ridges which sweep up to join the anterior edge of the neural spine. From the zygosphene to the posterior edge, the dorsal part of the neural arch roof slightly overhangs deep lateral depressions and is therefore sharply defined in dorsal view. The preserved prezygapophyseal facet is concave anteriorly, slightly convex posteriorly, and with a distinct anteromedial prominence; its long axis is directed laterally and slightly anteriorly. The lateral part of the facet is rounded, but with a slight anterolateral concavity. The prezygapophyseal process projects slightly lateral to the facet. In ventral view, the centrum is much wider than long, the subcentral ridges at close to 45° to the midline. The haemal keel is smoothly rounded in cross-section, and well-defined laterally by faint ridges and deep depressions containing small subcentral foramina. The depressions diverge anteriorly to form a broad cotylar 'lip', but the keel itselfbegins somewhat posterior to the cotyle, with a distinct transverse edge (step-like, but hidden in direct lateral view by the Specimen

pzw

eml

zaw

nse

IIsh

enw

OM OM OM OM OM OM OM OM OM OM

37 .5 38 .8 40 .0 40 .4 38 .7 42 .4 42 .0 47 .6 59 . 1

15.2 15.8 15.6 15 .7 15 .9 16.8 17.0 19.0 19.6 18 . 1

19 .7 20.0 20 .0 20 .0 22 .2 24 .2 23 .2 244 25 .7 24 .3

42 .7 44.9 44 .5

50.0

10 .8 11 .0 11 .8 114 12 .8 14 .5 14 .5 15.4 16.1 15.2

F2 3661 F23662 F2 3665 F23663 F23664 F23667 F23668 F23669 F9132 F91J3

45 .8 52.1 52.2 53 .9 54 .3 49 .5

57.2 57.2 55.6 56 . 1 50 .5

429

paradiapophysis). The keel widens posteriorly, then narrows in the posterior third of its length between posteroventrolateral facets converging to a point. The postzygapophyseal facet has anterior and posterior edges convex, forming an ellipsoid with axis directed slightly posterior to the transverse plane, with the medial edge parallel to the midline and angularly defined. In anterior view, the vertebra is dominated by the massive zygosphene, wider than the cotyle and almost as high as wide. The cotyle is slightly wider than high, its widest point below its vertical centre. The paradiapophyses are located low on the centrum, extending ventrally below the cotyle and facing more ventrally than laterally. The prezygapophyses are nearly horizontal, defining planes which intersect at the base of the neural canal. The zygosphenal facets are slightly convex laterally. The anterior face of the zygosphene (cracked, but similar to other more complete specimens) is divided by ridges into anterodorsal, lateral and ventral regions, the ridges converging to form a vertically oriented median prominence (median lobe in dorsal view). The ventral portion of the anterior face passes smoothly into the roof of the neural canal. The canal is relatively small and trapezoidal in shape. There are deep paracotylar depressions , lacking foramina. In posterior view, the dorsal edge of the neural arch is divided into strongly concave portions dorsal and lateral to the large zygantrum. The dorsal portions extend well up onto the neural spine, making the zygantral roof very thick medially. The lateral portions are nearly vertical medially, becoming convex distally (preserved on right only) so that the postzygapophysis is quite blunt. No parazygantral foramina are present in this specimen.

Referred material Table I. Measurements (in millimetres) for individual vertebrae of Liasis dubudingala n. sp. Abbreviations: pzw, maximum width across prezygapophyseal processes; cml, centrum length in ventral midline, measured from rim of cotyle to rim of condyle; zaw, maximum internal width of zygantrum; nsc, distance in midline from anterodorsal tip of neural spine to furthest point of condyle; nsh, from anterodorsal tip of neural spine to furthest point of hypapophysis or haemal keel; cnw, maximum width of condyle.

The vertebrae available represent each of the major regions of the trunk, but the most numerous and best-preserved are from the anterior trunk (precardiac, or cervical sensu lato) . The wellpreserved precardiac vertebrae were found in a small area at the same level, so are assumed to represent a single individual; other remains representing similar-sized individuals come from

430


ALCHERINGA

ALCHERINGA


other sites. Cervical (atlas and axis), cloacal and caudal vertebrae have not been discovered. The precardiac vertebrae have the neural spine distinctly narrower, higher and anteroposteriorly somewhat shorter than at mid-trunk (more than twice as high as long in some); smaller and rounder condyle and cotyle; and haemal keel narrower anteriorly (triangular rather than semicircular in section transverse to the vertebral axis) and posteriorly forming a prominent hypapophysis (broken in some specimens). The keel is sigmoid in lateral view and the hypapophysis extends obliquely below the condyle in the most anterior vertebrae, but both anterior and posterior edges are more vertical in those nearer mid-trunk; the ventral extremity is angular but thick and relatively blunt. One of the anterior vertebrae has the dorsoposterior angle of the neural spine produced backwards and acute in lateral view, but the anterior edge is also deflected posteriorly so that the condition represents irregular growth of a single vertebra (the whole dorsal edge shifted posteriorly) rather than the usual pythonine 'hatchet-shaped' spine. Vertebrae from the posterior trunk have the neural spine only slightly higher than long, zygapophyses sloping at up to about 10° from the horizontal, relatively longer centrum, and more depressed condyle and cotyle. The haemal keel does not extend as close to the cotyle, but is more angular in section than at midbody, with more distinct lateral ridges and a median ridge defmed by slight depressions; the posterior facets face more ventrally and contact each other along the midline (rather than only at the tip as at midbody), forming a sharp keel. This difference in the haemal keel between middle and posterior trunk regions is seen in most pythonines, although the shape and degree of prominence of the keel varies between species. The sub central ridges do not become more strongly defined in available posterior vertebrae, but the subcentral depressions extend more broadly to the cotylar rim, eliminating the broad 'lip' seen in anterior and middle trunk. A few of the vertebrae possess small pits lateral to the zygantrum (parazygantral foramina).

431

A number of rib heads referred to this species do not differ significantly from those of other pythons examined, except in size. One posterior trunk vertebra (QM F30586) and a fragment of a dentary (QM F30587) apparently represent smaller or juvenile individuals. The vertebra is missing most of the neural spine, but its anterior edge begins to rise steeply, level with the posterior edge of the zygosphenal facets; in dorsal view, the interzygapophyseal ridges are straight for most of their length but slightly less strongly demarcated from the zygapophyses than in the larger specimens; the lateral faces of the neural arch are not vertical except immediately posterior to the zygosphene; the haemal keel begins at the cotylar rim, between deep subcentral depressions; the posterior margin of the right postzygapophysis is straight (left missing); the zygosphene is barely wider than the cotyle and much wider than deep, with flat (right) or slightly concave (left) facets; the cotyle and condyle are round dorsally but bluntly angular ventrally, so about as high as wide; the paradiapophyses thus do not extend below the cotyle; the zygapophyses slope at less than 10° from the horizontal; the dorsal and lateral segments of the posterior edge of the neural arch are only slightly concave, and the postzygapophysis is acute in posterior view. In other respects, this specimen is similar to others from the same site. The jaw fragment represents most of the posterior tooth-bearing process of a left dentary; length 17.1 mm, maximum width 4.3 mm, depth (excluding teeth) 2.9 mm. Nine small round alveoli (and part of the tenth anteriorly) occur on a distinct raised ridge along the lateral edge, occupying half the width of the fragment. The medial edge is intact posteriorly (the last four alveoli are thus on a more or less distinct freeending process) but broken anteriorly, and another small broken area on the ventral surface, below the most anterior alveolus, indicates the position of the anterior end of the fossa for the surangular (level with ninth or between ninth and tenth from the rear of dentary). Teeth are ankylosed in all but the second, fourth and sixth alveoli, but partial crowns are

432


retained only in the third and fifth. An anterolateral ridge is present on these teeth, but no lingual ridge is observed (lingual ridges are usually restricted to the distal part of the tooth in pythons: Frazzetta 1966). The alveoli and teeth are considerably smaller in diameter (especially buccolingually) than in specimens ofL. olivacea, but with similar posteromedial curvature. The ventrolateral margin is produced ventrally as a sharp crest, rounded in outline, below the last four alveoli. Otherwise the fragment is of approximately uniform depth laterally, but thin medially. It is similar in size to a specimen of L. olivacea (AR 4422) with a total mandible length of70.2 mm, but deeper laterally (depth 2.2 mm in L. olivacea) due to the more prominent crest.

Discussion Diagnoses of the Pythoninae and included clades were given by Kluge (1993) without reference to vertebral morphology. Possibly the group cannot be diagnosed by uniquely derived characters of the vertebrae, as the close vertebral similarity of recent pythons to some Boinae suggests that most shared features are plesiomorphic (Rage 1984). The material described here is therefore referred to the Pythoninae on the basis of overall similarity to recent species (cf. Szyndlar 1984) and because no boines are known from Australia. It can be confidently excluded from all other families including the extinct Madtsoiidae (e.g., Scanlon 1992, Rage 1998) and from all boine genera examined or known from the literature. The following characters are found in most pythonines: zygapophyses inclined only slightly above horizontal (steepest in Python molurus [Linnaeus, 1872]), less than 10% in posterior trunk); neural spine steep anteriorly, overhanging posteriorly and with dorsal surface for tendinous attachment sharply defined anteriorly and extending for full length of spine; anterior face of zygosphene with vertical median ridge forming a distant prominence in dorsal view; no paracotylar foramina; several small parazygantral foramina often present; haemal keel of middle and posterior trunk vertebrae defined by grooves or

ALCHERINGA

depressions beginning at the cotylar rim, but projecting below centrum only in the posterior part of each vertebrae. Variations from this widespread pattern may be polarised by mapping them onto a cladogram, and for the present we use the phylogenetic hypothesis of Kluge (1993; but see also Scanlon in press). All of the characters listed in the generic diagnosis remain unknown in the new species described here; it is referred to Liasis on the basis of overall similarity to the included species, a possible synapomorphy with L. olivaceus and L. mackloti (height of neural spines), and lack of apomorphies of Aspidites Peters, 1876; Morelia and Python Daudin, 1803, the only other pythonine genera recognised here. This new species is distinguished from all pythonines except Python regius (Shaw, 1802) by the back-sloping anterior edge and blunt dorsoposterior angle of the neural spine, and the strongly sloping anterodorsal face of the zygosphene. It is further distinguished from all except P. regius and Morelia viridis (Schlegel, 1872) by the greater relative height of the neural spine at equivalent positions in the column; from Python reticulatus (Schneider, 1801) and P. molurus by less steeply sloping zygapophyses; from Morelia and these species of Python by the hypapophysis depth reducing gradually in the anterior trunk (condition not known in P. regius); and from species of all other genera by its much larger size. It differs from P. regius in its much greater size, and the presence of a median prominence on the anterior face of the zygosphene. The vertebrae from Bluff Downs differ strongly in morphology from all recent Australasian pythons examined and equally from those known from Miocene sites (Riversleigh and Bullock Creek; Smith & Plane 1985, Scanlon, in press). The relative height and morphology of the neural spine, and its lack of sharp demarcation, in lateral view, from the zygosphene roof, are matched only in the small African species Python regius (two vertebrae figured by Rage 1984, fig. 3; no data provided). However, Python appears to be monophyletic (Underwood & Stimson 1990,

ALCHERINGA


Kluge 1993) and these features must have arisen within that genus in an ancestor of P. regius, presumably outside the Australasian region. P. regius also has a median notch in the anterior edge of the zygosphene, also probably an autapomorphy. Assignment to Python appears to be precluded by a lack of a derived character shared by all available specimens of Python and Morelia. This character is not observable on a single vertebra, but depends on an inference of relative position in the column for several elements from a single skeleton. Hypapophysis depth (H) is measured as the difference between distances from anterodorsal point of neural spine to, respectively, the furthest points of condyle and hypapophysis (Hoffstetter 1960). In the condition found in Aspidites and Liasis s.l. (i.e. the six basal lineages of pythonines according to Kluge 1993), which is assumed here to be plesiomorphic, H decreases steadily through the precardiac region. In Morelia and Python species examined (the derived condition), H remains nearly constant through much of this region, and reduces sharply just before the mid-trunk region. This can be seen in a plot ofH (or as used in Fig. 2, -H) against vertebra number (if available) or maximum vertebral width; the latter figure steadily increases up to the mid-trunk, and is therefore a useful estimator of sequential position. Plotted values for the Bluff Downs snake (inferred to represent one individual) seem to approach zero gradually as pzw (width across prezygapophyses) increases (Fig. 2), like species of Aspidites but contrasting with the steeper change in species of Morelia and Python. This polarity is adopted provisionally based on comparisons within Pythoninae, but requires further testing using other lineages. Hoffstetter (1960) plotted curves for the value ofH against vertebral sequence for three boines (species of Boa sensu Kluge 1991), which show sharp reductions at the transitions at the transition to mid-trunk as in Morelia and Python. Hence this condition, even if unambiguously apomorphic within Pythoninae, may be homoplastic when the character is applied more

433

broadly. Among pythonines outside the Morelia + Python clade, the highest neural spines are found in Liasis sensu stricto (i.e L. olivacea and L. mackloti). The low neural spines of Aspidites, Antaresia spp., and L ias is papuana are assumed to represent the primitive condition for pythonines (other Melanesian species not examined). The high neural spines of the Bluff Downs python are thus a possible synapomorphy with Liais s.s., but this feature has certainly been subject to convergence within Morelia and Python. Neural spine height is proportionally much greater in the fossil material than in either Liasis species. The shape of the prezygapophyseal facets is quite variable between pythonine species; midtrunk vertebrae show this feature most clearly. Among those examined, an anterior concavity and distinct anteromedial prominence are present in P. reticulatus, Morelia amethistina and L. olivacea, absent in P. molurus and P. regius; L. mackloti and L. papuana have nearly straight anterior and medial edges, but without a sharp angle between them. Polarity is uncertain, but again the phenetically closest species is L. olivacea. The limited number of characters identified here for pythonine vertebrae thus imply a position either within, or as a sister taxon to Liasis (sensu stricto). On present evidence, L. dubudingala is the largest snake to have lived in Australia. Scanlon (1993) used the proportions between maximum vertebral width and total length of the vertebral column in an Aspidites melanocephalus skeleton (1 : 13 7) to estimate the length of fossil snakes. The largest vertebra from Bluff Downs (QMF 9132) is 59.1 mm wide (Table 1),so the preliminary estimate is 8096 mm (excluding the head), or roughly 8.35 m in total. However, the Aspidites used for this comparison has a total of 394 vertebrae whereas large Liasis spp. (olivacea and papuana) average close to 450, suggesting the length should be increased proportionally. Allometric and other forms of variation have yet to be thoroughly investigated, but these figures imply that, of the small number

434


of individuals of L. dubudingala yet known, one was roughly 9 metres in total length; the maximum size of the species may have been even greater. Previously, the largest snake vertebrae known in Australia were those of the madtsoiid Yurlunggur camfieldensis (Scanlon, 1992).

Palaeoecology Several large reptilian predators have been recorded from the Bluff Downs Local Fauna including three crocodilians (Archer 1976, Molnar 1979, Willis & Mackness 1996, Willis & Molnar 1997, Mackness & Sutton 2000) and a giant varanid (Mackness & Hutchinson 2000). The giant python is the third large terrestrial reptilian carnivore (Quinkana babarra Willis & Mackness, 1996 is interpreted as a terrestrial crocodile) recorded from the site. Several authors (Archer & Bartholomai 1978, Molnar 1981, Flannery 1994) have argued that the absence of large terrestrial mammalian predators in the Australian Tertiary and Pleistocene deposits may be a reason for the predominance of these large reptiles although Wroe et al. (1999) have presented an alternate hypothesis. Only two mammalian carnivores have been described from Bluff Downs: Thylaeoleo crassidentatus (Archer & Dawson, 1982) and Dasyurus dunmalli Barthlomai, 1971 (Wroe & Mackness 1998,2000). Most pythons detect and capture their prey using heat sensitive labial pits. Any active ectothermic prey usually has a body temperature higher than the ambient temperature and is therefore as readily detectable as endothermic prey (Ehmann 1993). There are several factors determining the type of prey this extinct giant python may have eaten. The first is the size of the head, or more correctly the size of the gape (Shine 1986). Unfortunately no adult skull material is available and there are varying opinions as to the size of the heads of giant extinct snakes. Flannery (1994, p. 113) refers to the madtsoiid Wonambi naracoortensis as having the head' the size of a shovel'. However, Barrie (1990) described a relatively complete skull ofa specimen estimated to be 6 m long, the skull being 135 mm

A LCHERINGA

in length - about the size of a child's toy spade (fragments of even larger specimens imply that the mandible length reached about 160 mm; Scanlon & Lee 2000). The extant Morelia amethistina is relatively slender with reported lengths of up to 8.5 m (Ehmann 1992). Even with a comparatively small head, individuals of less than 6 m are still capable of feeding on adult wallabies (Ehmann 1993). More massive snakes such as Python reticulatus of southeast Asia are probably closer to the body size of the extinct python described here. They are able to feed on relatively large prey items such as pigs, deer and primates, (Shine et al. 1998) including humans (Murphy & Henderson 1997). Second, the size and diversity of mammalian prey in the Miocene and Pliocene may have had a significant effect on the evolution of size and prey selection in Australian pythons. Shine (1987) has demonstrated a relationship between the size of island tiger snakes with large prey items (mutton birds) available and those without. The palaeoenvironment of the Bluff Downs Local Fauna has been interpreted as being similar to that found in Kakadu today (Boles & Mackness 1994) although the presence of a ringtail possum (Mackness, unpublished data) and a land snail assemblage normally associated with vine thickets (Mackness, unpublished data) indicates that there may have been patches of closed forest either in gullies or as part of a riparian fringe. Thirdly, climatic and seasonal factors may have come into play. During the Pliocene, Australia was within 10 to 20 of its present position with a warm climate (Wildford & Truswell 1990). The presence of deeply weathered laterites (Mackness & Archer 2002) and extensive permanent water (Boles & Mackness 1994) indicates that there was a high level of precipitation and humidity at the site. Such a climatic regime would have been suitable for the digestion of larger prey items by the python. Digestion in cooler climates takes longer and the bacterial fermentation of large prey items in such conditions would have posed a considerable risk to the snake. There were also increasing cycles of seasonality towards the end of the Pliocene,

ALCHERINGA


which perhaps would have favoured more opportunistic feeding behaviour. There was a considerable selection of fauna available including presumably juvenile diprotodontids (Archer 1976, Mackness 1995a, b). Given that snakes like P. reticulatus can 'process' quadrupeds such as pigs and deer, there seems to be little doubt that juvenile diprotodontids may have been within the size range of prey items for this snake. The presence of high neural spines in the fossil, on the other hand, indicates strong epaxial muscles and a climbing ability (Johnson 1955). This may point to a wider range of prey including birds, reptiles and arboreal mammals. Regardless of what prey items L. dubudingala consumed, there is little doubt that it would have been a formidable predator in the Bluff Downs palaeoenvironment.

Acknowledgments We wish to thank Professor Michael Archer and Dr Suzanne Hand for their helpful comments on the manuscript. Jack, Rhonda, Bram, Troy and Selesti Smith of Bluff Downs Station continue to provide tremendous help and support for the ongoing research into the Bluff Downs Local Fauna. Collection of the Bluff Downs material was supported in part by an ARC Program Grant to M. Archer; a grant from the Department of Arts, Sport, the Environment, Tourism and Territories to M. Archer, S. Hand and H. Godthelp; a grant from the National Estate Program Grants Scheme to M. Archer and A. Bartholomai; and grants in aid to the Riversleigh Research Project from Wang Australia, ICI Australia and the Australian Geographic Society. JS also thanks Glen Shea for information on the taxonomic literature and acknowledges financial support from the Australian Research Council through grants to M.S. Y. Lee. References ARCHER, M. , 1976. Bluff Downs local fauna. Results of the Ray E. Lemley Expeditions , Part I. The Allingham Fonnation and a new Pliocene vertebrate fauna from northern Australia, p. 383-396. In M.

435

Archer & M . Wade, eds, Memoirs of the Queensland Museum 17. ARCHER, M. & BARTHOLOMAI, A., 1978 . Tertiary mammals of Australia: a synoptic view . Aleheringa 2, 1-19. ARCHER, M . & DAWSON, L., 1982 . Revision of marsupia l lions of the genu s Thylaeoleo Gervais (Thylacoleonidae, Marsupialia) and Thylaco leonid evolution in the late Cainozoic. In Carnil'oro us Marsupials . M . Archer, ed., Royal Zoological Soc iety of New South Wales, Sydney, 477-494. ARCHER, M. & WADE, M., 1976. Results of the Ray E. Lemley Expeditions, Part I. The Allingham Fonnation and a new Pliocene vertebrate fauna from northern Australia . Memoirs of th e Queensland Museum 17, 379-97 . AUFFENBERG, W ., 1963. The fossil snakes of Florida. Tulane Studies in Zoology 10, 131-216. BARRIE, DJ ., 1990. Skull elements and additional remains of the Plei stoc ene boid snake Wonambi

naracoortensis. Memoirs of the Queensland Museum 28, 139-151. BARTHOLOMAI, A., 1971. Dasyurus dunmali, a new species of fossil marsupial (Dasyuridae) in the upper Cainozoic deposits of Queensland . Memoirs of the Queensland Museum 16 , 19-26. BOLES, W.E. & MACKNESS, B.S ., 1994 . Birds from the Bluff Downs Local Fauna, Allingham Formation, Queensland. Records of the South Australian Museum 27, 139-149. COGGER, H.G ., CAMERON, E.E. & COGGER, H.M ., 1983. Zoological Catalogue of Australia. Vol. I. Amphibia and Reptilia. Australian Government Publishing Service, Canberra, 313 p. DAUDIN, F.M., 1803. Histoire naturelle generale et particuliere des reptiles. Volumes 5-7. Dufart, Paris . DUMERIL, A.M .e. & BIBRON, G., 1844. Erpetologie generale ou histoire naturelle complete des reptiles. Vol. 6. Librairie Encyclopedique de Roret, Paris. EHMANN, H.F.W ., 1992. Encyclopedia of Australian Animals - Reptiles. Angus & Robertson, Sydney. 423 p. EHMANN, H.F.W., 1993. Family Boidae. In Fauna of Australia Volume 2A Amphibia & Reptilia, CJ . Glasby, GJ.B. Ross & P.L. Beesley, eds, Australian Government Publishing Service, Canberra, 284-289. FITZINGER, LJ ., 1826. Neue Classification del' Reptilien nach ihren natiirlichen Verwandtschaften. Heubner, Vienna. (Facsimile reprint, Society for the Study of Amphibians and Reptiles, 1997 .) FLANNERY, T .F., 1989. A new species of Wallabia (Macropodinae: Marsupialia) from Pleistocene deposits in Mammoth Cave, southwestern Western Australia. Records of the Western Australian Museum 14, 299-307 . FLANNERY, T.F. 1994. The Future Eaters. Reed, Melbourne, 423 p. FRAZZETTA, T.H ., 1966. Studies on the morphology and function of the skull in the Boidae (Serpentes), Part II. Morphology and function of the jaw apparatus in Python sebae and Python molurns. Journal of Morpho-

436


logy" 8, 217-296. GRAY, J.E., 1825. A synopsis of the genera of reptiles and amphibians, with a description of some new species. Annals of Philosophy. new series 10, 193-217. GRAY, J.E., 1842. Synopsis of the species of prehensiletailed snakes, or family Boidae. Zoological Miscellany, 41-46. HOFFSTETTER, R., 1960. Sur la classification des bOldes de Madagascar et des Mascareignes. Bulletin du Museum National dHistoire Naturelle 32,131-138. HOFFSTETTER, R. & GASC, J.P., 1969. Vertebrae and ribs of modem reptiles. In Biology of the Reptilia. Volume I, C. Gans, ed., Academic Press, London, 210-310. INTERNAnONAL COMMISSION ON ZooLOGICAL NOMENCLA11JRE, 1988. Opinion 1514. Liasis Gray, 1842 (Reptilia, Serpentes): Liasis mack/oti Dumeril & Bibron, 1844 designated as the type species. Bulletin of Zoological Nomenclature 45, 244. JOHNSON, R.G., 1955. The adaptative and phylogenetic significance of vertebral form in snakes. Evolution 'I, 367-388. KLUGE, A.G., 199 \. Boine snake phylogeny and research cycles. Miscellaneous Publications, Museum of Zoology. University of Michigan 178, I-58. KLUGE, A.G ., 1993 . Aspidites and the phylogeny of pythonine snakes. Records of the Australian Museum, Supplement 19, 1-77. LAcEPEDE, B.G. E. L., 1804. Memoire sur plusieurs animaux de la Nouvelle-Hollande dont la description n'a pas encore ete publiee. Annales du Museum National d 'Histoire Naturelle, Paris 4, 184-21 \. LINNAEUS, e., 1758. Systema naturae per regna tria naturae. secundum classes, ordines, genera, species cum characteribus, difJerentiis, synonymis, locis. Editio decima, reformata, Tom. f. Laurentii Salvii, Stockholm, 824 p. McDoWELL, S.B., 1975. A catalogue of the snakes of New Guinea and the Solomons, with special references to those in the Bernice P. Bishop Museum. Part II. Anilioidea and Pythoninae. Journal of Herpetology 9, 1-79. MACKNESS, B.S., 1995a. The Bluff Downs Local Fauna - a new synopsis. CA VEPS '95 : Conference on Australasian Vertebrate Evolution, Palaeontology and Systematics. Canberra, 19-21 April 1995. MACKNESS, B.S., 1995b. A review of late Cainozoic Diprotodontids. CAVEPS '95: Conference on Australasian Vertebrate Evolution, Palaeontology and Systematics. Canberra, 19-21 April 1995. MACKNESS, B.S. & ARCHER, M., 2002. A new petauroid possum (Marsupialia, Pseudocheiridae) from the Pliocene Bluff Downs Local Fauna, northern Queensland. Alcheringa 25, 439-454. MACKNESS, B.S. & HUTCHINSON, M., 2000. Fossil lizards from the early Pliocene Bluff Downs Local Fauna. Transactions of the Royal Society of South Australia 124, 17-30.

MACKNESS, B.S. & SCANLON, J.D., 1999. First Pliocene record of the madtsoiid snake genus Yurlunggur

A LCHERINGA

Scanlon, 1992 from Queensland. Memoirs of the Queensland Museum 43, 783-785. MACKNESS, B.S. & SIITTON, R.H., 2000. Possible evidence for intraspecific aggression in a Pliocene crocodile from north Queensland. Alcheringa 24, 55-62 MACKNESS, B.S., WIlITEHEAD, P.W. & McNAMARA. G.e. , 2000. A new potassium-argon hasalt date in relation to the Pliocene Bluff Downs Local Fauna, northern Australia. Australian Journal of Earth Sciences 47, 807 -8 II. McNAMARA. G.C .. 1990. The Wyandotte Local Fauna, a new dated Pleistocene vertebrate fauna from northern Queensland. Memoirs of the Queensland Museum 28, 285-297. MERRILEES, D., 1979. The prehistoric environment in Western Australia. Journal of the Royal Society of Western Australia 62. 109-128. MOl.NAR, R.E., 1979. Crocodylus porosus from the Pliocene Allingham Formation of North Queensland. Results of the Ray E. Lemley Expeditions. Part 5. Memoirs of the Queensland Museum 19, 357-365. MOLNAR, R.E., 1981. Pleistocene ziphodont crocodilians of Queensland. Records of the Australian Museum 33, 803-834. MURPHY, J.e. & HENDERSON, R.W., 1997. Tales of Giant Snakes: A Historical Natural HistolY of Anacondas and Pythons. Krieger Publishing Company, Florida. OATES, W., OATES, L., HERSHBERGER, H., HERSHBERGER, R., SAYERS, B. & GODFREY, M., 1964. Gugu-Yalangi and Wik-Munkan language studies. Occasional Papers in Aboriginal Studies. 2. Australian fnstitute of Aboriginal Studies. Canberra. PETERS, W., 1873. Uber eine neue Schildkrotenart, Cinosternon eJJeldtii und einige andere .}Ieue oder weniger bekannte Amphibien. Monat~berichte der Koniglich Preussischen Akademie der Wissenscha!ten zu Berlin (1873), 603-618. PETERS, W. & DoRIA, G., 1878. Catalogo dei rettili e dei batraci raccolti da O. Beccari, L.M. D'Albertis e A.A. Bruijn nella sotto-regione Austro-malese. Annali del Museo Civico di Staria Naturale di Genova 13. 323-450. PLEDGE, N.S., 1992. The Curramulka Local Fauna: a new late Tertiary fossil assemblage from Yorke Peninsula, South Australia. The Beagle, Records of the Northern Territory Museum of Arts and Sciences 9, 49-60. RAGE, J.-C., 1984. Handbuch der Paliioherpetologie. II. Serpentes. Gustav Fisher, Stuttgart. RAGE, J.-e., 1998. Fossil snakes from the Palaeocene of Sao Jose de Itaborai, Brazil. Part I. Madtsoiidae, Aniliidae. Palaeo vertebrata 27, 109-144. SCANLON, J.D., 1992. A new large madtsoiid snake from the Miocene of the Northern Tenitory. The Beagle, Records of the Northern Territory Museum of Arts and Sciences 9, 49-60. SCANLON, J.D., 1993. Madtsoiid snakes from the Eocene Tingamarra Fauna of eastern Queensland. Kaupia,

ALCHERINGA


Darmstadter Beitrage zur Naturgeschichte 3, 3-8. SCANLON, 1.0., 1995a. First records tium Wellington Caves, New South Wales of the extinct madtsoiid snake Wonambi naracoortensis, Smith 1976. Proceedings of the Linnean Society of New South Wales 115, 233-238. SCANLON, J.D., I 995b. An elapid snake (Hydropheinae) from the Middle Miocene Encore Site, Riversleigh.

5th Conference on Australian Vertebrate EI'olution. Palaeontology & Systematics. Collected Abstracts. SCANLON, J.D., 1997. Nanowana gen. nov. , small madtsoiid snakes from the Miocene of Riversleigh: sympatric species with divergently specialised dentition. Mem-

oirs of the Queensland Museum 41. 393-412. SCANLON, J.D., in press. Montypythonoides: the Miocene snake Morelia riversleighensis (Smith & Plane, 1985) and the geographic origin of pythons. Memoirs of

the Association of Australasian Palaeontologists. SCANLON , J.D. & LEE, M.S.Y., 2000. The Pleistocene serpent Wonambi and the early evolution of snakes.

Nature 4tJ3, 416-420 . SCHLEGEL, H., 1837. Essai sur la physionomie des serpens. M.H . Schonekat, Amsterdam. SCHLEGEL, H., 1872. De dierentuin van het koninklijk

zoologisch genootschap Natura Artis Magistra te Amsterdam zoOiogis ch geschetst door Prof H. Schlegel directeur van het Rijks Museum van Natuurlijke Historie te Leiden met historische herinneringen van P.H. Witkamp. Gebruder van Es, Amsterdam, xxviii (general introduction), vi + 3-292 (birds), vii + 3-174 (mammals), iii + 3-64 (reptiles). SHAW, G., 1802. General Zoology. or Systematic Natural History. G. Kearsley, London. SHINE, R. 1986. Ecology of a low-energy specialist: food habits and reproductive biology of the Arafura filesnake (Acrochordidae). Copeia 1986, 424-437 . SHINE, R., 1987. Ecological comparisons of island and mainland populations of Australian tiger snakes (Notechis: Elapidae). Herpetologica 43, 233-240. SHINE, R., HARLOW, P.S., KEOGH, 1.S. & BOEADI, 1998. The influence of sex and body size on food habits of a giant tropical snake, Python reticulatus. Functional

Ecology 12,248-258. SMITH, L.A., 1985. A revision of the Liasis childreni species-group (Serpentes: Boidae). Records of the

Western Australian Museum 12, 257-276. SMITH , M.J., 1975. The vertebrae offour Australian elapid snakes (Squamata: Elapidae) Transactions ofthe Royal Society of South Australia 99, 71-84. SMITH, M.1., 1976. Small fossil vertebrates tium Victoria Cave, Naracoorte, South Australia. IV Reptiles. Transactions of the Royal Society of South Australia 100, 39-5\.

437

SMITH, M.J. & PLANE, M., 1985. Pythonine snakes (Boidae) from the Miocene of Australia. BMR Journal of Australian Geology and Geophysics 9, 191-195. STIMSON, A.F., 1969. Liste der rezenten Amphibien und Reptilien. Boidae. Das Tierreich 89, 1-49. STULL, O.G., 1932. Five new subspecies of the family Boidae. Occasional Papers of the Boston Societv of Natural History 8, 25-30. SZYNDLAR, Z., 1984 . Fossil snakes of Poland. Acta Zoologica CracOI'iensia 28, 1-156 . THOMSON, S. & MACKNESS, B.S., 1999. Fossil reptiles from the early Pliocene Bluff Downs Local Fauna, with a description of a new species of Elseya. Transactions of th e Royal Society of South Australia 123, 101105. UNDERWOOD, G. & STIMSON, A.F., 1990. A classification of pythons (Serpentes, Pythoninae). Journal of the

Zoologica l Society of London 221, 565-603. WA GLER, J.G. , 1830. Natiirliches System del' Amphibien.

mit I'oranghender Classification del' Siiugethiere und Vogel. Cotta'schen, Miinchen. WI\.Ll S, P.M.A. & MACKNESS, B.S., 1996 . Quinkana baban'a, a new species of ziphodont mekosuchine crocodile from the early Pliocene Bluff Downs Local Fauna, northern Australia with a revision of the genus.

Proceedings ofthe Linnean Society of New South Wales 1/6, 143-151. WI\.LlS, P.M.A. & MOLNAR, R.E., 1997. A review of the Plio-Pleistocene crocodilian genus Pallimnarchus.

Proceedings of the Linnean Socie ty of New South Wales 117, 224-242. WILFORD , G.E. & TRUSWELL, E.M., 1990. Cainozoic. In Australia: EI'olution of a Continent, A.G.L. Paine, ed., Bureau of Mineral Resources, Australia. 82-88. WROE, S. & MACKNESS, B.S., 1998. Revision of the Pliocene dasyurid, Dasyurus dunmalli (Dasyuridae: Marsupialia). Memoirs of th e Queensland Museum 42, 605-612. WROE, S. & MACKNESS, B.S., 2000. Additional material of Dasyurus dunmalli from the Pliocene Chinchilla Local Fauna of southeastern Queensland and its phylogenetic implications. Memoirs of the Queensland Museum 45, 457-461. WROE, S., MYERS, T. 1., WELLS, R.T. & GILLISPIE, A., 1999. Estimating the weight of the Pleistocene Marsupial Lion (Thylacoleo carnifex: Thylacoleonidae): implications for the ecomorphology of a marsupial super-predator and hypotheses of impoverishment of Australian marsupial carnivore faunas. Australian Journal of Zoology 47, 489-498.