52949750

7/29/2019 52949750

1/22

Cranial osteology and preliminary phylogenetic

assessment ofPlectrurus aureus

Beddome, 1880(Squamata: Serpentes: Uropeltidae)

REBECCA S. COMEAUX, JENNIFER C. OLORI* and CHRISTOPHER J. BELL

Department of Geological Sciences, Jackson School of Geosciences, 1 University Station C1100,

The University of Texas at Austin, Austin, TX 78712, USA

Received 11 July 2008; accepted for publication 31 March 2009

Uropeltid snakes are among the most poorly understood clades within Alethinophidia. Their small size, limited

geographic distribution, high incidence of endemism, and fossorial behaviour all contribute to the general paucityof systematic collections of these snakes, especially of adequate skeletal preparations, in most museum collections.Their hypothesized position within the higher-order phylogeny of alethinophidian snakes calls attention to the needfor additional morphological work on the group. Hypotheses of uropeltid phylogenetic relationships based onmorphological analyses are few, and continue to be hampered by limited taxon-sampling and character matricesthat rely predominantly on features that are visible on articulated skulls. We utilized high-resolution X-raycomputed tomography (HRCT) to investigate the cranial osteology of Plectrurus aureus Beddome, 1880, a speciesfor which no osteological data were previously available. We provide a detailed description of the skull andmandible, and comment on morphological characters and potential phylogenetic relationships. Clarity in characterdescriptions is of paramount importance, and additional morphological characters are desirable. HRCT provides anondestructive way to identify new systematically informative morphological characters from digitally disarticu-lated specimens. The small size of many uropeltid species, including P. aureus, will help to frame a greaterappreciation of the limitations of traditional HRCT protocols for revealing detailed anatomical features of small

vertebrates.

2010 The Linnean Society of London, Zoological Journal of the Linnean Society, 2010, 160, 118138.doi: 10.1111/j.1096-3642.2009.00595.x

ADDITIONAL KEYWORDS: computed tomography (CT) plectrurus skull anatomy snake phylogeny.

INTRODUCTION

Uropeltids are a diverse clade of fossorial snakes that

inhabit tropical montane forests, agricultural fields,

and occasionally wet lowlands in southern India and

Sri Lanka (Rajendran, 1970, 1977, 1985; Gans, 1976).

Little is known of their evolutionary history and no

fossil specimens are yet recognized. They range in

body size from 20 to 80 cm in total length, and are up

to 2 cm in diameter (Gans, 1973). Their heads are

small and pointed, but their tails are often thick and

reinforced with a robust skeletal structure, including

a uniquely formed bony caudal plate, lending an

overall shape that is easily mistaken for the head,

and from which the name uropeltid, meaning rough-

tailed is derived (Gans, 1976).Higher-order snake phylogeny remains an active

and controversial area of research (Caldwell, 2007).

Recent phylogenetic hypotheses consistently place

uropeltids near the base of the evolutionary tree of

alethinophidian snakes, but their exact relationships

remain obscure. Morphological analyses consistently

yield hypotheses in which uropeltids are closely

related to Anomochilus Berg, 1901, Cylindrophis

Wagler, 1828, and Anilius Oken, 1816, either with a

stepwise succession of basal alethinophidians*Corresponding author. E-mail: [email protected]

Zoological Journal of the Linnean Society, 2010, 160, 118138. With 32 figures

2010 The Linnean Society of London, Zoological Journal of the Linnean Society, 2010, 160, 118138118

7/29/2019 52949750

2/22

(Anomochilus, uropeltids, Cylindrophis, and Anilius;

e.g. Cundall, Wallach & Rossman, 1993; Lee &

Scanlon, 2002; Lee et al., 2007) or with uropeltids as

a sister group to Anomochilus, with that clade having

variable hypothesized relationships with Cylindro-

phis and Anilius (e.g. Scanlon & Lee, 2000; Tchernov

et al., 2000; Lee & Scanlon, 2002; Lee et al., 2007). Anexception is a hypothesis based in part on morphology

and presented by White, Kelly-Smith & Crother

(2005), in which Uropeltis Cuvier, 1829 was some-

times recovered as sister to Anilius, but Anomochilus

and Cylindrophis were not included in the analysis.

Molecular estimates of the position of uropeltids vary

considerably, depending on the data set analysed and

the taxon sampling. Uropeltids (variably represented

by Rhinophis Hemprich, 1820, and/or Uropeltis)

were placed as sister group to the African Calabaria

reinhardtii (Schlegel, 1848) by Heise et al. (1995), to

(Tropidophis + Casarea) by Slowinski & Lawson

(2002; based on c-mos), to Caenophidia by Vidal &Hedges (2004), to the anomalepidid Liotyphlops

Peters, 1881 by White et al. (2005; but Cylindrophis

was not included in the analysis), to Cylindrophis

(Slowinski & Lawson, 2002; Vidal & Hedges, 2002;

Wilcox et al., 2002; Lee et al., 2007), and to

(Cylindrophis + Anomochilus) (Gower et al., 2005).

Eight genera and 47 species of uropeltids are cur-

rently recognized (McDiarmid, Campbell & Tour,

1999), but their interrelationships are unresolved

(Dessauer, Gartside & Gans, 1976; Dessauer, Cadle &

Gans, 1987; Cadle et al., 1990; Rieppel & Zaher,

2002), and alternative hypotheses of relationships

cannot meaningfully be compared. The two mostrecent hypotheses of relationships among uropeltids

were presented by Cadle et al. (1990; based primarily

on protein data) and Rieppel & Zaher (2002; based on

skull morphology). In the aggregate, only 19 species

were included in the phylogenetic hypotheses pre-

sented by Cadle et al. (1990) and Rieppel & Zaher

(2002), and only three species were common to both

analyses. The small size of uropeltids, their limited

geographic distribution, high incidence of endemism,

fossorial behaviour, and local superstitions regarding

their handling all contribute to the general paucity of

systematic collections of these snakes (Rajendran,

1990), especially of adequate skeletal preparations, inmost museum collections.

The skull morphology of uropeltid snakes is not

well studied. The most comprehensive taxonomic

surveys are those by Cundall & Irish (2008) and

Rieppel & Zaher (2002). Early discussions and illus-

trations of the skull and mandibles of uropeltids

were provided by Dumril (1853), Peters (1861), Jan

& Sordelli (1865), and Boulenger (1893). The first

detailed description was that by Baumeister (1908)

for Rhinophis philippinus (Cuvier, 1829) and Rhino-

phis homolepis (Hemprich, 1820) (see McDiarmid

et al., 1999, for a discussion of taxonomic syn-

onymy). Subsequent authors in the early half of the

20th century included uropeltids as part of larger

surveys of squamate or snake anatomy and evolu-

tion (Radovanovic, 1937; Mahendra, 1938; Bellairs,

1949; Bellairs & Underwood, 1951), and the samewas true in the reviews by Underwood (1967) and

Bellairs & Kamal (1981). A few other authors dedi-

cated their attention to the specific anatomical fea-

tures of uropeltids, notably the mandible (Rieppel &

Zaher, 2000), cranialvertebral joint (Hoffstetter,

1939; Williams, 1959), or the configuration of the

skull (Gans, 1973), and Smith (1943) briefly

reviewed cranial features of the group and provided

illustrations of the skull of Uropeltis smithi (Gans,

1966) (reported as Uropeltis grandis by Smith; see

McDiarmid et al., 1999 for a discussion of taxonomic

synonymy).

A resurgence of interest in anatomical features ofbasal snakes (including uropeltids) and their impor-

tance for elucidating phylogeny is reflected in a series

of papers from the last quarter of the 20th century, all

of which include descriptions of uropeltid anatomy

(Rieppel, 1977, 1978, 1979, 1980a, b, 1983; Groom-

bridge, 1979a, b, c; Cundall & Rossman, 1993;

Cundall et al., 1993; Zaher & Rieppel, 1999). These

papers provide the foundation upon which our

modern understanding of uropeltid skulls is being

constructed (Cundall & Irish, 2008). The analysis by

Rieppel & Zaher (2002) remains the only in-group

phylogeny of uropeltids based entirely on morphology.

As a result of the framework provided by that land-mark paper we can place basic anatomy in a phylo-

genetic context, and begin to test uropeltid

morphological characters. In this paper we fully

describe the osteology of the skull of Plectrurus

aureus Beddome, 1880, and use this information to

evaluate existing characters and comment on poten-

tial phylogenetic relationships.

MATERIAL AND METHODS

A single specimen of P. aureus [California Academy of

Sciences (CAS) 17177] was scanned at The University

of Texas High Resolution X-ray Computed Tomogra-phy Facility (UTCT Facility). The specimen was col-

lected by R.H. Beddome from Wynad, Kerala State,

India, and is preserved in alcohol (Fig. 1). The date of

collection is unknown. The total skull length is

9.4 mm from the tip of the rostrum to the distal edge

of the occipital condyle. A wet specimen was utilized

for scanning because in dry skeletal preparations

elements may pull together creating false contacts.

The use of an alcohol-preserved specimen produces

CT scans that more accurately represent the gaps

CRANIAL OSTEOLOGY OF PLECTRURUS AUREUS 119


7/29/2019 52949750

3/22

between bones, and render the intervening soft tissue

clearly visible in unmodified CT slices.

High-resolution CT scans were taken in the coronal

(axial) plane, and resulted in 527, 16-bit TIFF imageswith an image size of 1024 1024 pixels. The original

slices had an interslice spacing (slice thickness) of

0.01811 mm, with a field of reconstruction of 7 mm.

The completed scan of P. aureus yielded a data set of

reasonably high quality, but the somewhat grainy

nature of the renderings of the tiny, individual ele-

ments indicate that this uropeltid was near the size

limit for the standard scanning and image-processing

protocols used by UTCT at the time the specimen was

scanned.

The 3D graphics volume software VGStudioMax

1.21 was used to obtain slices in the other two

orthogonal planes, producing a total of 744 sagittalslices and 503 frontal slices. 3D models of the skull

were rendered using the same software package. Soft

tissue was digitally removed by optimizing the

density histogram for the greyscales representing

bone. Where extremely thin bone exists in the skull

(e.g. the notch in the nasal bone) this optimization

can be challenging, because thin bone and some soft

tissues may be rendered with similar greyscales.

Thus, when the soft tissue is removed or rendered

completely transparent, regions of extremely thin

bone may also disappear.

Individual cranial elements were digitally disar-

ticulated using the manual segmentation tool withinVGStudioMax. The resulting amplified images can

be rotated and digitally manipulated to provide

novel anatomical views, allowing a detailed analysis

of each element. The entire, fully segmented skull is

depicted in Figures 2 and 3 in dorsal, ventral, right-

lateral, and anterior views. A major advantage of

this technology is that it is non-invasive, permitting

detailed data on skeletal anatomy to be gathered in

a nondestructive manner. Digital disarticulation of

the skull is particularly useful for identifying or

scoring characters that are not visible in articula-

tion, especially in cases where specimens may be too

small, fragile, or rare for traditional methods of

study, such as histological sectioning. For example,

Rieppel & Zaher (2002) proposed a number of poten-

tially informative characters (e.g. 27, 28, 31) that

can only be scored on partially or fully disarticu-lated material. This is exactly a case where CT data

have huge advantages. In sum, CT data reveal

details of anatomy that are not generally accessible

to many researchers.

RESULTS

GENERAL DESCRIPTION OF THE SKULL AND

MANDIBLE

Premaxilla

The edentulous premaxilla (Fig. 4) has a distinctly

triangular form. The nasal process of the premaxilla

is posterodorsally directed, and is exposed in dorsal

view as a narrow wedge separating the nasals ante-

riorly. The vomerine process is a midline structure

projecting posteriorly between the vomers. The

medial edge of each vomer contacts the corresponding

lateral edge of the vomerine process of the premaxilla.

Ventrally, the premaxilla slots into a shallow notch

on the ventral surface of the vomers; the vomer

extends dorsal to the premaxilla in a horizontal, over-

lapping contact. The nasal process also extends above

the medial portion of each septomaxilla, although

actual bone-to-bone contact does not occur. A single

premaxillary foramen is present, appearing as aslight circular indentation centered on the ventral

surface of the bone. Distinct transverse processes

(lateral processes of Rieppel, 1977) form a flat, weakly

buttressing contact with the anterior tip of each

maxilla (= shizarthrotic contact of Cundall et al.,

1993; Rieppel & Zaher, 2002; Fig. 5). There are

grooves on the dorsal and ventral sides of the trans-

verse process, the ventral side of the vomerine

process, and on the rostrum. The anteromedial

surface of the premaxilla is emarginated to form a

distinctly bipartite rostrum (Rieppel & Zaher, 2002)

that projects anteriorly a short distance beyond the

transverse processes.

Septomaxilla

The posterolateral edge of the septomaxilla (Fig. 6) is

directed dorsomedially and contacts the prefrontal.

The nasal process of the premaxilla is positioned

between the medial flange of each septomaxilla, and

the vomerine process is ventral to, but not in contact

with, the anteroventral edge. The lateral edge of the

septomaxilla is medially inflected, and is separated

from lateral contact with the maxilla by a gap filled

Figure 1. Plectrurus aureus (CAS 17177), whole animal.

Scale bar: 1 cm.

120 R. S. COMEAUX ET AL.


7/29/2019 52949750

4/22

with soft tissue. However, the anteroventral edge of

the septomaxilla overlies and directly contacts the

anteromedial process of the maxilla (Fig. 7).

The dorsal surface of the septomaxilla is concave

ventrolaterally, with a deep lateral edge curving

upwards and medially, and a medial flange located on

the anterodorsal medial edge. Dorsomedially, the sep-

tomaxilla is overlain by the ventromedial edge of the

nasal. In the articulated skull this contact obscures a

relatively large foramen for the vomeronasal nerve

that is located on the posterodorsal medial tip of the

septomaxilla (Cundall & Irish, 2008). The ventral

surface of the septomaxilla forms a deep concavepocket for the vomeronasal organ. A strongly curved

flange of bone from the lateral edge of the septomax-

illa (lateral wall of Rieppel, 1977) forms the floor of

the lateral side of the vomeronasal capsule, and con-

tacts the dorsal side of the vomer.

Maxilla

The transverse process of the premaxilla abuts the

anterior end of the maxilla (Fig. 8) in a weakly but-

tressing (but not fully buttressed, or shizarthrotic)

contact (Rieppel & Zaher, 2002), where the edges of

the two bones are in articulation (Baumeister, 1908),

just anterior to the anteromedial process. The dorsal

surface of the anteromedial process of the maxilla

underlaps the ventral side of the medially curved

lateral edge of the septomaxilla, and closely

approaches, but does not meet, the anterolateral

vomerine process of the premaxilla.

The palatine (or medial) process of the maxilla is

posterior to the anteromedial process, and is in

contact with the ventral anterolateral curvature of

the choanal process of the palatine. The maxilla

tapers gradually as it projects posteriorly to meet thelateral side of the maxillary process of the ectoptery-

goid in a mediolaterally overlapping contact. The

ascending process (Rieppel, 1977; Rieppel & Zaher,

2002) (= prefrontal process of Baumeister, 1908;

dorsal process of Cundall & Irish, 2008) of the maxilla

contacts the prefrontal in a strong, interlocking

articulation. The open superior alveolar nerve canal is

exposed dorsally, just medial to the ascending process

(Rieppel & Zaher, 2002: 124). Seven tooth positions

are located ventrally: five functional teeth are in place

Figure 2. Articulated skull, anterior to the right. A, dorsal view. B, ventral view, lower jaws digitally removed.

Abbreviations: cb, compound bone; co, coronoid; dt, dentary; ept, ectopterygoid; f, frontal; mx, maxilla; n, nasal; ot-occ,

otico-occipital complex; pa, parietal; pf, prefrontal; pl, palatine; pm, premaxilla; pt, pterygoid; q, quadrate; sm, septomax-

illa; vo, vomer. Scale bar: 1 mm.



7/29/2019 52949750

5/22

Figure 3. Articulated skull. A, right-lateral view, anterior to the right. B, anterior view. Abbreviations: ang, angular;

cb, compound bone; co, coronoid; dt, dentary; ept, ectopterygoid; f, frontal; mx, maxilla; n, nasal; ot-occ, otico-occipital

complex; pa, parietal; pl, palatine; pf, prefrontal; pm, premaxilla; q, quadrate; sm, septomaxilla; V2, opening for the

maxillary branch of the trigeminal nerve; V3, opening for the mandibular branch of the trigeminal nerve. Scale bar:

1 mm.

Figure 4. Isolated premaxilla, anterior to the right except

in (C). A, dorsal view. B, ventral view. C, anterior view. D,

right-lateral view. Abbreviations: np, nasal process; pmf,

premaxillary foramen; ro, rostrum; tp, transverse process;

vop, vomerine process. Scale bar: 1 mm.

Figure 5. Dorsal view of the articulated maxillae and

premaxilla with other skull elements digitally removed.

Abbreviations: mx, maxilla; pmx, premaxilla. Scale bar:

1 mm.



7/29/2019 52949750

6/22

on the right side and appear to be fully ankylosed.

The teeth are relatively large and distinctly recurved.

A well-developed replacement tooth is visible in the

second tooth position, but is not yet ankylosed

(Fig. 8B, C); additional replacement teeth are visible

for some of the distal tooth positions (Fig. 8C). Two

foramina are present: the first is ovoid in shape and

located at the anterior end of the maxilla; the second,

smaller opening is below the ascending process.

Nasal

The paired nasals (Fig. 9) are separated anteriorly by

the premaxilla for approximately one-third of their

length. A distinct medial process (not visible in the

articulated skull) is located on the posteromedial

portion of each nasal: it forms a weak vertical contact

with the other nasal along the midline, but declines in

Figure 6. Isolated septomaxilla. A, posterior view, medial to the left. B, dorsal view, anterior to the right. C, ventral view,

anterior to the right. D, right-lateral view, anterior to the right. E, right-medial view, anterior to the left. Abbreviations:

mf, medial flange; vnc, vomeronasal capsule; vnf, vomeronasal nerve foramen. Scale bar: 1 mm.

Figure 7. Articulated septomaxillae and right maxilla

with other skull elements digitally removed, anterior view.

Abbreviations: mx, maxilla; sm, septomaxilla. Scale bar:

1 mm.

Figure 8. Isolated maxilla. A, dorsal view, anterior to the

right. B, right-lateral view, anterior to the right. C, right-

medial view, anterior to the left. Abbreviations: amp,

anteromedial process; asp, ascending process; ep, ectop-

terygoid process; plp, palatine process; sanc, superior

alveolar nerve canal. Scale bar: 1 mm.



7/29/2019 52949750

7/22

prominence anteriorly, and is completely diminished

at the level of the posteriormost extent of the nasal

process of the premaxilla. A small notch or emargin-

ation is present on the anterolateral edge of the nasal,

a short distance behind the anteriormost tip (Fig. 9A,

B). Although this region is notched in some uropeltids

(Rieppel & Zaher, 2002), in the CT data set of P.aureus this probably represents an area of thin bone

and not an actual notch. The nature of the notch

mentioned by Rieppel & Zaher (2002) for their char-

acter 2 is unclear, but presumably refers to the nasal

margin of the external naris(?). We interpret the

notch in P. aureus as an artifact of digital image

processing, resulting from the loss of bone when soft

tissues were digitally removed from the original data

set to reveal the skull.

The nasal overlies the posterodorsal tip of the sep-

tomaxilla, with the lateral edge of the nasal curving

ventrally inward towards the dorsally curving lateral

edge of the septomaxilla. The lateral edge of theposteroventral surface of the nasal contacts the ven-

tromedial curvature of the prefrontal. The posterior

surface of the nasal directly meets the anterior face of

the medial flange and the anterodorsal surface of the

frontal in an overlapping contact.

Prefrontal

The frontal process (supraorbital process of Cundall &

Irish, 2008) of the prefrontal (Fig. 10) is located pos-

terodorsally, and articulates with the frontal via a

shallow notch on the anterolateral surface of the

frontal. The supraorbital process of the parietal does

not meet the frontal process of the prefrontal, afeature that was reported as polymorphic for Plectru-

rus perroteti Dumril & Bibron, 1854 (Rieppel &

Zaher, 2002: character 7). The posteroventral surface

of the prefrontal, ventral to the frontal process, abuts

the anteroventral surface of the frontal and forms the

anterior margin of the orbit. The prefrontal wraps

around a protrusion, the preorbital ridge (Rieppel,

1978), located on the anterolateral edge of the frontal.

The dorsomedial edge of the prefrontal is positioned

next to the posterolateral edge of the nasal; the

lateral curvature of the septomaxilla contacts the

anteromedial edge of the prefrontal.A broad, bipartite maxillary process extends from a

lateral foot process (Rieppel, 1977) anteromedially to

a medial foot process (Fig. 10A); the notch between

these two accommodates the ascending process of the

maxilla. The lateral foot process is finger-like, and

extends posterolaterally to form a slight overlapping

contact with the posterior edge of the ascending

process of the maxilla (Fig. 3A). Medial to that articu-

lation, the prefrontal is notched to form the dorsal

portion of a lacrimal duct. In the articulated skull, the

lateral wall of that duct is formed by the prefrontal,

and is floored by the maxilla, whereas the palatine

contributes to the medial margin of the duct (Fig. 11).However, a narrow line of soft tissue prevents the

prefrontal from directly contacting the palatine.

Vomer

The dorsal surface of the anterolateral process of the

vomer (Fig. 12) is situated beneath the ventral

surface of the septomaxilla, and projects laterally

towards the anteromedial process of the maxilla

(although no actual contact occurs; Fig. 13). The dor-

sally concave pocket of the ventral surface of the

septomaxilla overlies the dorsal surface of the vomer,

and completes the vomeronasal capsule. Anterome-

dially, a premaxillary process meets the vomerineprocess of the premaxilla in a complex articulation.

The vomerine process of the premaxilla is situated

within a recess between the two vomers, but dorsally

each vomer overlaps the caudal tip of the vomerine

process of the premaxilla.

Figure 9. Isolated nasal, anterior to the right. A, ventral

view. B, dorsal view. C, right-lateral view. Abbreviation:

mp, medial process. Scale bar: 1 mm.

Figure 10. Isolated prefrontal. A, right-lateral view, ante-

rior to the right. B, posterior view, lateral to the right.

Abbreviations: fp, frontal process; lc, notch for lacrimal

canal; lfp, lateral foot process; mfp, medial foot process.

Scale bar: 1 mm.



7/29/2019 52949750

8/22

The posterolateral process appears midway along

the length of the vomer, curving laterally and dorsally

towards the maxilla, although the two do not meet

(Fig. 13). It is widest anteriorly, and is connected to

the medial ridge via a low crest that marks the

posterior surface of the vomeronasal capsule. A small,

rounded foramen transmits the vomeronasal nerve,and penetrates the crest near where it meets the

medial ridge (Figs 12D, 14). The posterolateral

process tapers posteriorly to a pointed palatine

process that is directed posteriorly and positioned

inside the choanal process of the palatine; the

palatine process of the vomer fills a narrow ventral

gap running along the lateral edge of the choanal

process of the palatine (Fig. 15).

The medial edge of the vomer forms a dorsally

projecting ridge that serves as a vertical contact

between the two vomers anteriorly. The interchoanal

process of the sphenoid extends anteriorly along the

midline to the level of the posteromedial ends of each

vomer, but does not contact either vomer (Fig. 16).

Palatine

The palatines (Fig. 17) are edentulous, as in all uro-

peltids studied so far except for Melanophidium punc-

tatum Beddome, 1871 (Rieppel & Zaher, 2002:

Figure 11. Posterior view of the snout region with theotico-occipital, frontals, parietal, pterygoids, ectoptery-

goids, quadrates, and lower jaws digitally removed. Abbre-

viations: lc, lacrimal canal; mx, maxilla; n, nasal; pf,

prefrontal; pl, palatine. Scale bar: 1 mm.

Figure 12. Isolated vomer, anterior to the right except in (D). A, ventral view. B, dorsal view. C, right-lateral view. D,

anterior view, medial to the right. Abbreviations: alp, anterolateral process; mr, medial ridge; plp, palatine process; pmp,

premaxillary process; polp, posterolateral process; vnc, vomeronasal capsule; vnf, vomeronasal nerve foramen. Scale bar:

1 mm.

Figure 13. Ventral view of the articulated maxillae and

vomers with the other skull elements digitally removed;

anterior to the right. Abbreviations: mx, maxilla; vo,

vomer. Scale bar: 1 mm.



7/29/2019 52949750

9/22

Figure 14. Posterior view of the septomaxillae (top) and

vomers (bottom) in life position, with other skull elements

digitally removed. Abbreviations: crest, vertical crest along

dorsal surface of posterolateral process of the vomer; mr,

medial ridge of vomer; vnf, vomeronasal nerve foramen.

Scale bar: 1 mm.

Figure 15. Anterolateral view of the complex articulation

between the palatines and vomers; anterior to the front-

right. Other skull elements have been digitally removed;

Abbreviations: pl, palatine; vo, vomer. Scale bar: 1 mm.

Figure 17. Isolated palatine, anterior to the right, except in (D). A, ventral view. B, dorsal view. C, right-lateral view. D,

anterior view, medial to the right. Abbreviations: chp, choanal process; ich, internal choana; lp, lateral process of Rieppel

& Zaher (2002); mxf, facet for maxilla; ptp, pterygoid process; V2f, foramen for the maxillary branch of the trigeminal

nerve; vop, vomerine process. Scale bar: 1 mm.

Figure 16. Dorsal view of the otico-occipital complex and palate with the frontals, nasals, parietal, pterygoids, ectop-

terygoids, quadrates, and lower jaws digitally removed; anterior to the right. Abbreviations: mx, maxilla; ot-occ,

otico-occipital complex; pf, prefrontal; pl, palatine; pmx, premaxilla; sm, septomaxilla; vo, vomer. Scale bar: 1 mm.



7/29/2019 52949750

10/22

character 3), and are separated from one another

along their entire midline length.

The choanal process is broad, arching in a medial

curvature to form the internal choana. Medially the

ventral edge of the process is elongated into distinct

anterior and posterior extensions. The anterior end of

the palatine is broadened, and underlaps the poste-rior portion of the vomer along its medial edge. On its

lateral side the anterior surface of the palatine over-

laps the maxilla. A distinct facet on the ventrolateral

surface of the anterior palatine marks the articula-

tion with the palatine process of the maxilla.

A large foramen for the maxillary branch of the

trigeminal nerve is present on the anterolateral

surface of the palatine. Its ventral margin is formed

by what Rieppel (1977) called the lateral process of

the palatine. In an articulated skull, the complex

articulation of the palatine and maxilla create the

appearance of a distinct process in ventral view, but

in the disarticulated skull ofP. aureus, it is clear thatthe process is merely the ventral portion of the

palatine ossification surrounding the foramen for the

nerve (Rieppel, 1977; Figs 2B, 17). The anterior tip of

the pterygoid inserts into a posterior groove on the

pterygoid process of the palatine (Figs 2B, 17A). The

palatine closely approaches, but does not contact, the

frontal dorsally: the space separating the elements is

small, presumably filled with soft connective tissue,

and would probably be seen as a clear contact in a

dried skull. Similarly, the dorsolateral edge of the

palatine is positioned ventral to the prefrontal, but

the two elements do not make direct contact.

Pterygoid

The anterior end of the pterygoid (Fig. 18) forms a

palatine process that articulates with the grooved

surface of the pterygoid process of the palatine. The

quadrate ramus extends posteriorly to the level of the

mandibular articulation, beneath the crista circum-

fenestralis of the otico-occipital complex. The ectop-

terygoid process is an anterolateral projection that

extends beneath the ectopterygoid, overlapping it in a

horizontal contact (anteriorly, the pterygoid underlies

the ventral side of the ectopterygoid). No teeth or

foramina are present.

Ectopterygoid

The anterior maxillary process of the ectopterygoid

(Fig. 19) meets the maxilla in an overlapping contactalong the posteromedial surface of the maxilla. The

pterygoid process is located on the posterior end of

ectopterygoid, and overlaps the pterygoid in a hori-

zontal contact. No foramina are present.

Frontal

The frontal (Fig. 20) closely approaches, and in dried

skull may contact, the palatine anteroventrally: in

our digital renderings, there is a narrow gap between

them that is filled with connective tissue, as in other

alethinophidian snakes. In medial view the frontal

forms an open, anteriorly tapering chamber thataccommodates the olfactory bulbs of the brain. Ante-

riorly, an olfactory tract canal is formed lateral to the

open medial contact between the dorsal and ventral

medial edges of each frontal (Rieppel, 1977).

A ridge located ventrolaterally and directed

towards the palatine forms the palatal process of the

frontal. It contacts the parasphenoid region of the

sphenoid bone medially: a flat surface formed

between the sphenoid and the palatal process pre-

sumably overlies the trabecula, extending from the

base of the ossified crista trabecularis, although the

cartilage is not visible in the scans.

Figure 18. Isolated pterygoid in dorsal view, anterior to

the right. Abbreviations: ep, ectopterygoid process; plp,

palatine process; qr, quadrate ramus. Scale bar: 1 mm.

Figure 19. Isolated ectopterygoid, anterior to the right.

A, ventral view. B, dorsal view. Abbreviations: mxp, max-

illary process; ptp, pterygoid process. Scale bar: 1 mm.



7/29/2019 52949750

11/22

The open posterior edge of each frontal contacts the

open anterior surface of the parietal. The supraorbital

process of the parietal projects anteriorly into a

groove on the dorsolateral surface of the frontal. A

large optic foramen is enclosed entirely within the

frontal, anterior to the frontalparietal contact. Ante-

rolaterally, a slight dorsal groove marks the articula-

tion with the frontal process of the prefrontal. Ventral

to that groove, the preorbital ridge forms an addi-

tional contact surface with the prefrontal. The ridge

extends anteriorly past the margin of the dorsal expo-

sure of the frontal, a feature that is shared with other

uropeltids and related taxa (Rieppel & Zaher, 2002:

character 27).

Although there is a narrow gap between the dorsal

surfaces of the nasal and frontal (filled with soft

tissue), the anteromedial and anterodorsal edges of

the frontal directly contact the nasal.

Postorbital and supraorbital

The postorbital and supraorbital are absent as dis-

crete ossifications in P. aureus and other uropeltids.

ParietalThe parietal (Figs 21, 22) is an unpaired element with

pronounced descending flanges that form much of the

body of the bone and contact the sphenoid region of

the otico-occipital complex. The posterodorsal surface

extends farther posteriorly from the level of the

descending flanges, and overlaps the otic region of the

otico-occipital complex. The paired supraorbital pro-

cesses (possibly homologous with the postfrontal;

Rieppel, 1977; Cundall & Irish, 2008) are located on

the lateral sides of the parietal and project anteriorly,

articulating with the frontal via a shallow groove on

the lateral surfaces of the frontal, dorsal to the optic

foramen (Figs 3A, 22B). The supraorbital processes donot contact the prefrontals (Fig. 3A). The anterior

surface of the main body of the parietal contacts the

open posterior edge of each frontal, extending the

open cavity and enclosing the central portion of the

brain. A narrow shelf of bone between the supraor-

bital processes forms the articulation facet for the

frontals.

A faint sagittal crest is located on the posterodorsal

surface of the parietal along the sagittal midline

(Fig. 21A). The parietal does not contribute to the

margin of the optic foramen in P. aureus. Posterolat-

erally, there is a shallow notch in the wall of the

parietal that marks the passage of the V2 branch ofthe trigeminal nerve (Figs 21, 22); in the articulated

skull, the parietal thus forms the anterior margin of

the opening for the passage of the V2 branch

(Fig. 3A).

Otico-occipital complex

The otico-occipital complex (Figs 2325) is a single

element in P. aureus, presumably composed of paras-

phenoid, basisphenoid, basioccipital, laterosphenoid,

prootic, opisthotic, exoccipital, and possibly supraoc-

Figure 20. Isolated frontal. A, anterior view. B, right-

lateral view, anterior to the right. C, right-medial view,

anterior to the left. Abbreviations: opf, optic foramen; otc,

olfactory tract canal; palp, palatal process; por, preorbital

ridge; sog, groove for the supraorbital process of the pari-

etal. Scale bar: 1 mm.

Figure 21. Isolated parietal, anterior to the right. A,

dorsal view. B, ventral view. Abbreviations: aff, articula-

tion facet for the frontal; sag, sagittal crest; sop, supraor-

bital process; V2f, fenestra for the V2 branch of the

trigeminal nerve; vlf, ventrolateral flange. Scale bar:

1 mm.



7/29/2019 52949750

12/22

cipital ossifications, all in a state of complete fusion

with one another (no sutures can be seen). The occipi-

tal condyle, otic region, and sphenoid region are dis-

tinct regions of the element.

In dorsal view, the parietal articulation facet is

extensive, extending back to the position of the

common crus between the anterior and posteriorsemicircular canals. A low, midline sagittal crest

extends from the posterior end of the synotic tectum,

anteriorly to the midline emargination of the roof. On

either side of the sagittal crest, a distinct foramen

opens into a canal that traverses the bone posteroven-

trally. These foramina were illustrated by several

authors (e.g. Smith, 1943; Gans, 1973; Rieppel, 1977;

Rieppel & Zaher, 2002; Cundall & Irish, 2008), but

remain unnamed, and their function is unknown. We

here name these as Rieppels canal, in honor of

Olivier Rieppel, who has done so much to advance our

understanding of uropeltid snakes. In some uropeltid

taxa, these canals are incompletely formed (Rieppel &Zaher, 2002: fig. 5C), and are instead developed as a

posteriorly positioned notch (Rieppels notch); in some

species this may be asymmetrical within an indi-

vidual [e.g. Uropeltis ocellata (Beddome, 1863);

Cundall & Irish, 2008].

The exoccipitals and basioccipital are fused to form

the occipital condyle (Baumeister, 1908; Rieppel &

Zaher, 2002). The occipital condyle forms a hemi-

spherical knob that is positioned on an elongated

neck of bone (Baumeister, 1908; Williams, 1959).

There is no indentation (fovea dentis of Williams,

1959) on the dorsal surface of the condyle, although a

trough for the brainstem is visible in dorsal view,

similar to the situation in Uropeltis, Rhinophis, and

P. perroteti (Rieppel & Zaher, 2002: character 17).

The sphenoid region of the otico-occipital complex,

consisting of fused parasphenoid and basisphenoidelements, makes up approximately one-half of the

complex in length, and is open dorsally. Its lateral

edges taper anteriorly in a stepwise fashion from the

anterior portion of the otic region. The first (most

posterior) stepwise reduction in lateral extent

happens at the position of the secondary anterior

opening of the vidian canal (secondary anterior

foramen of Underwood, 1967). The next lateral reduc-

tion happens at the position of the ossified base of the

crista trabecularis, which ends behind the (lateral)

frontoparietal suture (Rieppel & Zaher, 2002: char-

acter 6). Continuing anteriorly from this point, the

sphenoid region tapers smoothly to terminate in aventrally positioned, narrow interchoanal (or inter-

vomerine) process. That process extends to sit

between the posterior ends of the vomers in dorsal

view, and is positioned medially between the dorsal

surfaces of the choanal processes of the palatines

posteriorly. A low keel extends posteriorly from the

interchoanal process (Fig. 23B).

The junction between the otic and sphenoid regions

is roughly marked by the anterior opening of the

vidian canal, and the anterior opening of the sixth

cranial nerve (CN VI). The anterior opening for CN VI

probably also transmits the internal carotid artery

(Rieppel, 1979). The passage for CN VI follows aposterolateral course, merging with the vidian canal

and emptying posteriorly into the prootic canal.

The passage of the second (maxillary) branch of the

trigeminal nerve (V2) is visible in lateral view, where

the parietal meets the otico-occipital complex

(Fig. 3A). In the isolated otico-occipital element, this

opening is marked by a shallow notch along the

anterior surface of the otic region (Fig. 25). The

foramen for the third (mandibular) branch (V3) is

posterior to V2, and is separated from it by a fused,

broad laterosphenoid ossification (Rieppel, 1976;

Rieppel & Zaher, 2002). Posterior and slightly ventral

to the V3 foramen, the prootic canal is clearly aseparate opening. The facial foramen (transmitting

CN VII) opens within the prootic canal and traverses

its length. The posterior opening of the vidian canal

opens into the prootic canal internally along the

lateral edge of the anteroventral portion of the prootic

canal. The configuration of the foramina follows that

reported by Rieppel & Zaher (2002: character 12) for

Uropeltis, P. perroteti, and Rhinophis drummondhayi

Wall, 1921 (but not Rhinophis sanguineus Beddome,

1863).

Figure 22. Isolated parietal. A, anterior view. B, right-

lateral view, anterior to the right. Abbreviations: sag,

sagittal crest; sop, supraorbital process; V2f, fenestra for

the V2 branch of the trigeminal nerve; vlf, ventrolateral

flange. Scale bar: 1 mm.



7/29/2019 52949750

13/22

A conspicuous juxtastapedial recess is visible

immediately posterior to the prootic canal. In P.

aureus the recess is mostly open, although the ante-

rior half of the recess is somewhat restricted by a

dorsal incursion of the crista circumfenestralis.

Along the medial portion of the recess, the large

fenestra ovalis opens into the otic chamber. The sta-

pedial footplate fills the fenestra ovalis (Fig. 26). At

the posteroventral edge of the stapedial footplate,

the foramen pseudorotunda is visible in lateral view

(Fig. 25A). Just anterior and ventral to the foramen

pseudorotunda, the lateral aperture of the recessusscalae tympani opens beneath the stapedial foot-

plate. It is obscured in lateral view by the develop-

ment of the crista circumfenestralis. The recessus

scalae tympani traverses the otic region, and a

prominent medial aperture opens internally near

the floor of the braincase (Fig. 25B).

The vagus foramen (transmitting CN X and the

jugular) is located posterior to the fenestra ovalis, in

a shallow lateral pocket of bone formed by a lateral

extension of the crista circumfenestralis; it can be

interpreted to be inside the juxtastapedial recess,

because the recess is open posteriorly, with no distinct

posterior margin (Rieppel & Zaher, 2002: character

14). The vagus foramen is bifurcated internally,

a feature also exhibited by Uropeltis, Rhinophis, and

P. perroteti (Rieppel & Zaher, 2002: character 15).

A single, small hypoglossal foramen (transmitting

CN XII) perforates the otico-occipital complex lateral

to the base of the foramen magnum and ventral

to Rieppels canal.

The medial surface of the otic region is also pierced

by foramina (Fig. 25B). Just ventral to the trigeminalnerve branches, the vidian canal is directed posteri-

orly; medial to the vidian canal is an opening to

transmit CN VI and the internal carotid artery.

Beneath the ventral margin of the otic capsule, the

bone is excavated into the internal auditory meatus

(transmits CN VIII). A smaller opening for CN VII

passes through the otic chamber anterodorsal to the

internal auditory meatus. The endolymphatic

foramen pierces the medial wall of the otic chamber

(Fig. 27).

Figure 23. Isolated otico-occipital complex, anterior to the right. A, dorsal view. B, ventral view. Abbreviations: cer,

cerebral carotid; ct, anterior end of the ossified crista trabecularis; ef, endolymphatic foramen; ik, interchoanal keel; ip,

interchoanal process; jsr, juxtastapedial recess; oc, occipital condyle; pavc, primary anterior opening of the vidian canal;

Rc, Rieppels canal; savc, secondary anterior opening of the vidian canal; VI, cranial nerve six. Scale bar: 1 mm.



7/29/2019 52949750

14/22

Stapes

The stapes (Fig. 26) has a large stapedial footplate

that fills the fenestra ovalis within the juxtastapedial

recess of the otico-occipital complex. There is a short

but well-ossified stapedial shaft.

Statolithic mass

A large statolithic mass (Fig. 27) is located within

each of the otic chambers of the otico-occipital

complex. The statolith on the right side of the skull

has been digitally removed to better display the

inside of the bony vestibule. Each statolith is ovoid in

form and is composed of very dense material. In other

snakes, the statolithic mass is not ossified, but con-

sists of a densely packed clump of crystals (C. Bell,

pers. observ.).

Quadrate

The quadrate (Fig. 28) is suspended from the ventro-

lateral surface of the otic capsule. The suprastapedial

process (Lee, 2005) projects posteriorly, tapering into

a rounded tip that extends over the stapedial shaft.

The suprastapedial process is longer than the man-

dibular condyle, a feature that is characteristic of

uropeltids and Anomochilus (Rieppel & Zaher, 2002:

character 22). A shallow groove is located on the

anteroventral surface of the condyle where it articu-

lates with the compound bone. There is a low crest

curving laterally along the dorsal surface of thequadrate.

MANDIBLE

The mandible is delicately built and in the adult is

composed of five separate ossified elements (Fig. 29).

In the articulated cranium, there is a distinct gap

between the anteromedial tips of the mandibles, indi-

cating the presence of extensive soft tissues at the

symphyseal region. The mandible reaches its greatest

height at the level of the coronoid. A short retroar-

ticular process is formed posteriorly.

Dentary

A pronounced posterodorsal process of the dentary

(Fig. 30) contacts the compound and coronoid bones. A

posteroventral groove on the medial surface of the

dentary marks the articulation with the splenial and

angular. There is no posteroventral process of the

dentary, which differs from the situation reported for

P. perroteti by Rieppel & Zaher (2002: character 18).

There are eight tooth positions located on the dorsal

surface. A few of the teeth do not appear to be anky-

losed to the dentary. These are probably replacement

teeth, and are apparently less dense near their bases.

Meckels canal (Lee, 2005) is closed anteriorly (exceptfor a tiny ventromedial opening at the extreme ante-

rior end of the dentary), but is open posteriorly as a

prominent groove just dorsal to the splenial and

angular. A single mental foramen is present on the

lateral surface, at the level of the third tooth position.

Splenial

The splenial (Fig. 31) is a small, sharply triangular

bone positioned on the posteromedial surface of the

dentary. The broad, flat posterior end of the splenial

Figure 24. Isolated otico-occipital complex. A, anterior

view. B, posterior view. Abbreviations: cer, cerebral carotid;

ct, anterior end of the ossified crista trabecularis; ef,

endolymphatic foramen; fpsr, foramen pseudorotunda; ik,

interchoanal keel; ip, interchoanal process; jsr, juxtasta-

pedial recess; ls, laterosphenoid; oc, occipital condyle;

pavc, primary anterior opening of the vidian canal; Rc,

Rieppels canal; savc, secondary anterior opening of the

vidian canal; st, stapes; tfc, trigeminofacialis chamber; VI,

cranial nerve six; XII, hypoglossal foramen. Scale bar:

1 mm.



7/29/2019 52949750

15/22

meets the anterior end of the angular in a buttressing

vertical contact. The lateral side of the splenial lies

flat against the medial side of the dentary. A small

foramen is positioned dorsally, just posterior to the

position of the last tooth on the dentary. There is no

dorsal process (Cundall & Irish, 2008).

Figure 25. Isolated otico-occipital complex, anterior to the right; A, right-lateral view. B, left-medial view. Abbreviations:

apm, medial aperture for the recessus scalae tympani; cer, cerebral carotid; ct, anterior end of the ossified crista

trabecularis; ef, endolymphatic foramen; fpsr, foramen pseudorotundum; iam, internal auditory meatus; ik, interchoanal

keel; ip, interchoanal process; jug, jugular foramen; lsf, laterosphenoid foramen; oc, occipital condyle; pc, prootic canal

(containing the posterior opening of the vidian canal and cranial nerves VI and VII); savc, secondary anterior opening of

the vidian canal; st, stapes; V2, notch for trigeminal nerve branch; V3, opening for trigeminal nerve branch; vc, vidian

canal; VI, cranial nerve six; VII, cranial nerve seven; X, vagus nerve. Scale bar: 1 mm.

Figure 26. Close up of the right-lateral occipital region;

anterior to the right. The stapes is outlined, with the shaftlocated in the bottom left of the outline. Scale bar: 1 mm. Figure 27. Axial section through the otic region of the

otico-occipital complex in anterior view. Abbreviations: ef,

endolymphatic foramen; fm, foramen magnum; hcc, hori-

zontal semi-circular canal; jsr, juxtastapedial recess; otic,

otic capsule; pcc, posterior semi-circular canal; stat, sta-

tolithic mass. Scale bar: 1 mm.



7/29/2019 52949750

16/22

Angular

The angular (Fig. 31) is a small, triangular bone

situated along the anteroventral side of the compound

bone and the posteroventral edge of the dentary. It

contacts the splenial anteriorly in a flat, vertical,

buttressing contact. A small, finger-like splenial

process extends anteriorly between the splenial and

the dentary on the lateral side. The angular sits flatagainst the ventral side of the compound bone. There

are no foramina.

Coronoid

The anteromedial process of the coronoid (Fig. 31)

articulates along the medial side of the anterodorsal

portion of the coronoid process of the compound bone.

A slight groove on the lateral surface of the coronoid

facilitates additional articulation with the compound

bone. The anteromedial process is positioned medial

to the compound process of the dentary, although no

direct contact occurs between the two elements. There

are no foramina.

Compound boneA short retroarticular process forms the posterior

portion of the compound bone (Fig. 32), and immedi-

ately anterior to the process is a deep, crescentric

notch for the mandibular condyle of the quadrate. At

the anterior end of the medial surface of the retroar-

ticular process, a small foramen for the chorda

tympani nerve enters the compound bone.

A pronounced coronoid process rises on the lateral

side of the compound. The coronoid bone articulates

along the medial surface of this process, and extends

farther dorsally than the compound bone, so that

the coronoid is visible in lateral view. Anterior to

the coronoid process, the compound bone slopesanteroventrally, and forms an elongated slanting

contact with the posterodorsal process of the dentary.

Anteroventrally, a horizontal contact is formed with

the angular. The anterior end of the compound bone is

strongly bifurcated, with lateral and medial processes

positioned on either side of a central mandibular

canal. At about the midpoint of the element on the

medial side, an elongate mandibular fossa opens ven-

trally to the interior of the bone. A small foramen is

present anteriorly on the lateral surface.

Figure 28. Isolated quadrate in right-lateral view; ante-

rior to the right. Abbreviations: cr, crest; mc, mandibular

condyle; ssp, suprastapedial process. Scale bar: 1 mm.

Figure 29. Right lower jaw. A, lateral view. B, medial

view. Abbreviations: ang, angular; cb, compound bone; co,

coronoid; d, dentary; spl, splenial. Scale bar: 1 mm.

Figure 30. Isolated right dentary. A, right-lateral view,

anterior to the right. B, right-medial view, anterior to the

left. Abbreviations: antf, anterior opening of Meckels

canal; Mklc, Meckels canal; mtlf, mental foramen; pdp,

posterodorsal process. Scale bar: 1 mm.



7/29/2019 52949750

17/22

DISCUSSION

More complete knowledge of cranial morphology in P.

aureus allows for phylogenetically meaningful com-

parison with P. perroteti and other uropeltids included

by Rieppel & Zaher (2002). Their work provides a

framework of potentially useful morphologic charac-

ters that we used in our preliminarily assessment of

P. aureus. In order to summarize important features

and facilitate comparison, scores for P. aureus were

added to Rieppel & Zahers (2002) matrix, and this

information is presented in Table 1.

Significantly, P. aureus differs from P. perroteti in

regard to three of the characters described by Rieppel

& Zaher (2002). The first is character 6, which refersto the position of the ossified base of the crista tra-

becularis. In P. perroteti the base ends behind the

frontalparietal suture (Rieppel & Zaher, 2002),

whereas in P. aureus it terminates at the suture.

Plectrurus perroteti shares this condition with more

basal uropeltids (according to tree of Rieppel & Zaher,

2002), whereas P. aureus is most similar to

Brachyophidium rhodogaster Wall, 1921, and to some,

but not all, Rhinophis.

Secondly, P. aureus differs from P. perroteti in char-

acter 13 (Rieppel & Zaher, 2002), whether or not the

juxtastapedial recess is wide open laterally. This char-

acter was difficult to apply because the condition in P.aureus is intermediate between the states described

and illustrated by Rieppel & Zaher (2002). The incon-

gruity between P. aureus and P. perroteti may be

lessened if this character was coded in a different

fashion. As it stands, P. aureus shares a less open

recess with the more basal taxa included by Rieppel &

Zaher (2002), whereas P. perroteti more closely

resembles the derived Uropeltis and Rhinophis.

The third character that demonstrates variation

between P. aureus and P. perroteti is character 18 of

Rieppel & Zaher (2002), which describes the develop-

ment of the posteroventral process of the dentary. The

process was scored as reduced for P. perrotetiby Rieppel & Zaher (2002), but it is clearly absent in

P. aureus. Again, P. aureus is most similar to

B. rhodogaster and some Rhinophis species, as well as

Uropeltis and Pseudotyphlops Schlegel, 1839. Plectru-

rus perroteti exhibits a morphology like that of Platy-

plectrurus Gnther, 1868.

In several instances we had difficulty interpreting

the telegraphic character descriptions provided by

Rieppel & Zaher (2002) and applying them to our

description of P. aureus. This is not a problem unique

Figure 31. A, articulated splenial and angular in left-lateral view, anterior to the left. B, articulated splenial and angular

in right-medial view, anterior to the left. C, articulated splenial and angular in dorsomedial view, anterior to the right.

D, isolated coronoid in right-lateral view, anterior to the right. E, isolated coronoid in right-medial view, anterior to the

left. Abbreviations: ang, angular; amp, anteromedial process; spl, splenial; splf, splenial foramen; splp, splenial process

of the angular. Scale bar: 1 mm.

Figure 32. Isolated right compound bone. A, right-lateral

view, anterior to the right. B, right-medial view, anterior to

the left. Abbreviations: cop, coronoid process; ctf, foramen

for the chorda tympani; ldp, lateral dentary process; mdbc,

mandibular canal; mdbf, mandibular fossa; mdp, medial

dentary process; mj, mandibular joint; rap, retroarticular

process. Scale bar: 1 mm.



7/29/2019 52949750

18/22

Table1.

Datamatrixmo

difie

dfrom

Rieppe

l&

Za

her

(2002),w

ith

thea

dditiono

fPlectrurusaureus

(to

p)

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

202

1

22

23

24

25

26

27

28

29

30

31

32

33

Plectrurusaureus

1

1

1

0

1

1

0

1

1

1

1

2

0

1

0

1

1

2

2

2

1

1

1

1

1

1

1

1

2

1

1

1

1

Melanophidium

punctatum

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

2

2

1

1

1

1

1

1

1

1

1

1

1

1

1

Melanophidium

wynaudense

0

0

1

0

0

0

0 1

0

0

0

0

0

0

0

0

1

0

0

2

2

1

1

1

1

1

1

1

1

1

1

1

1

1

Platyplectrurus

0

1

1

0

0

0

0 1

1

1

1

1

2

0

1

1

1

0

1

2

2

1

1

1

1

1

1

1

1

2

1

1

1

1

Uropeltis

1

1

1

1

1

2

0

1

1

1

1

2

1

1

0

1

1

2

2

2

1

1

1

1

1

1

1

1

2

1

1

1

1

Teretrurus

1

0

1

0

0

1

1

1

1

0

1

2

0

0 1

1

0

2

2

2

1

1

1

1

1

1

1

1

2

1

1

1

1

Rhinophis

drummondhayi

1

1

1

1

1

1

1

1

1

1

1

2

1

1

0

1

1

2

2

2

1

1

1

1

1

1

1

1

2

1

1

1

1

Rhinophissanguineus

1

1

1

1

1

2

0

1

1

1

1

1

1

1

0

1

1

?

2

2

1

1

1

1

1

1

1

1

2

1

1

1

1

Plectrurusperroteti

1

1

1

0

1

0

0 1

1

1

1

1

2

1

1

0

1

1

1

2

2

1

1

1

1

1

1

1

1

2

1

1

1

1

Pseudotyphlops

1

1

1

0

0

0

1

1

1

1

0

1

1

0

1

1

1

2

2

2

1

1

1

1

1

1

1

1

2

1

1

1

1

Anomochilus

0

0

1

0

0

0

1

0

0

0

0

0

0

0

?

0

0

0

1

1

1

1

1

1

1

1

1

1

2

0

0

0

0

Cylindrophis

0

0

0

0

0

0

0 1

0

0

0

0 1

0

0

0

0

0

0

0

1

0

0

0

0

0

1

1

1

1

1

0

0

0

0

Anilius

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0



7/29/2019 52949750

19/22

to their paper, but rather is one that impacts many

morphological studies. Space constraints in journals

and the additional expense incurred from publishing

numerous illustrations appear to be the primary

forces contributing to a persistent problem with

adequate character descriptions. Careful attention to

the description, and especially illustration, of morpho-logical character states will greatly reduce the poten-

tial for confusion or misunderstanding in subsequent

analyses (Joyce & Bell, 2004).

For example, character 4 of Rieppel & Zaher (2002)

is problematic because of unclear wording and a lack

of visual representation. The buttressing contact

between the anteromedial process of the maxilla and

the anterolateral process of the vomer described by

those authors for many uropeltids appears to be

somewhat misleading, because this region of contact

includes contributions from the maxilla, vomer, and

premaxilla; in other taxa (e.g. some Rhinophis and

Uropeltis species), there may also be a ventral pro-jection from the septomaxilla that is visible in palatal

view (J. Olori, C. Bell, pers. observ.), although this is

not the case in our specimen of P. aureus.

A second major issue affecting character description

and interpretation stems from incomplete taxon sam-

pling, which can result in the discovery of new char-

acter states that are not encompassed by the original

character description. This situation can present

obstacles for the inclusion of new taxa in subsequent

analyses. When intermediate or previously unknown

states are identified, characters must be redescribed,

which necessitates rescoring the taxa used in the

original analysis. Demonstrating this point, our studyof P. aureus reveals multiple states intermediate to

those documented by Rieppel & Zaher (2002).

Character 1, for example, does not appear to

adequately represent the total range of variation

present among uropeltid snakes. In P. aureus, a clear

contact between the maxilla and premaxilla is formed

(unlike the configuration in Melanophidium Gnther,

1864, depicted by Rieppel & Zaher 2002), but the

bones do not form a fully straight and tightly but-

tressing articulation, as was depicted for R. san-

guineus by Rieppel & Zaher (2002).

Likewise, Rieppel & Zaher (2002) reported that all

of the uropeltids that they surveyed exhibited contactbetween the premaxilla and vomer within a well-

defined recess, rather than as an overlapping articu-

lation (their character 26). This description is unclear

and deficient because we find that the contact is

complex and intermediate between these two states

in P. aureus. Character 13 of Rieppel & Zaher (2002),

which refers to the openness of the juxtastapedial

recess, is also subjective because the character defi-

nitions are ambiguous. We found that the morphology

in P. aureus is again intermediate to the states pro-

posed by Rieppel & Zaher (2002), signifying a need to

better understand variation in uropeltid snakes.

Overall it is unclear where P. aureus fits among the

hypotheses of uropeltid relationships. The majority of

the features shared by P. aureus and P. perroteti are

common to all other uropeltids found by Rieppel &

Zaher (2002) to be more derived than Melanophidium.In other words, at the level of our current understand-

ing of uropeltid anatomy these features are not phylo-

genetically informative for many taxa. Looking closely

at the characters in which P. aureus and P. perroteti

differ, both species exhibit a combination of states

shared with both the derived and basal taxa hypoth-

esized by Rieppel & Zaher (2002). It is apparent that P.

aureus and P. perroteti may not necessarily be sister

taxa, and that broader taxonomic sampling of uro-

peltid species will in all likelihood result in new

phylogenetic hypotheses not predicted by existing

analyses. This conclusion may not be surprising con-

sidering that monophyly has not been established forany non-monotypic genus included in any previous

uropeltid phylogenetic analyses.

CONCLUSION

Existing morphological and molecular data do not

yield a pretty picture of the probable stability of

current uropeltid taxonomy. Our study demonstrates

a need for the clarification of existing characters, as

well as a need for improved taxon sampling in ana-

tomical studies and phylogenetic analyses. Further-

more, the acquisition of new material, the discovery of

additional characters for analysis, and an improvedunderstanding of patterns of variation in all charac-

ters will play an important role in helping to recover

a more thorough understanding of the evolutionary

dynamics of this interesting and enigmatic group of

snakes. High-resolution X-ray CT provides a ready

means of gathering additional data on skeletal mor-

phology, and permits the nondestructive utilization of

existing large collections of preserved uropeltids for

that purpose. The small size of most uropeltid skulls

appears to place them near the limit of traditional CT

scanning protocols, but a new generation of micro-CT

scanners provides a promising technological advance

that could yield higher-resolution anatomical data forthese tiny snakes. Our efforts to gather and interpret

such additional data are now underway.

ACKNOWLEDGEMENTS

We offer special thanks to J. Vindum of the California

Academy of Sciences for facilitating the loan of the

specimen used in this study. R. Ketcham and M.

Colbert conducted the CT scanning and digital data

acquisition. We benefited greatly from discussions



7/29/2019 52949750

20/22

with D. Cundall, C. Gans, D. Gower, and O. Rieppel,

all of whom freely shared their thoughts and opinions

about uropeltids. Special thanks also go to D. Cundall

and O. Rieppel for providing advance copies of their

unpublished works. B.-A. S. Bhullar, K. Claeson, E.

Ekdale, T. LaDuc, M. Maga, J. Rodgers, and T. Rowe

provided encouragement, advice, and comments onvarious aspects of this project. Funding for this

research was provided by a University COOP fellow-

ship award from The University of Texas at Austin to

RC, and from the Jackson School of Geosciences at

The University of Texas at Austin.

REFERENCES

Baumeister L. 1908. Beitrge zur Anatomie und Physiologie

der Rhinophiden. Integument, Drsen der Mundhhle,

Augen und Skeletsystem. Zoologische jahrbcher, Abteilung

fr Anatomie und Ontogenie der Tiere 26: 423526, Plates

2326.Bellairs A dA. 1949. The Journal of the Linnean Society of

London 41: 482512. The anterior brain-case and interor-

bital septum of Sauropsida, with a consideration of the

origin of snakes. Zoology Plates 9-11.

Bellairs A dA, Kamal AM. 1981. The chondrocranium and

the development of the skull in recent reptiles. In: Gans C,

Parsons TS, eds. Biology of the Reptilia, Volume 11, Mor-

phology F. New York: Academic Press, 1263.

Bellairs A dA, Underwood G. 1951. The origin of snakes.

Biological Reviews 26: 193237.

Boulenger GA. 1893. Catalogue of snakes in the British

Museum (Natural History). Volume I., containing the fami-

lies Typhlopid, Glauconiid, Boid, Ilysiid, Uropeltid,

Xenopeltid, and Colubrid Aglyph, Part. London: Taylor

and Francis.

Cadle JE, Dessauer HC, Gans C, Gartside DF. 1990.

Phylogenetic relationships and molecular evolution in uro-

peltid snakes (Serpentes: Uropeltidae): allozymes and

albumin immunology. Biological Journal of the Linnean

Society 40: 293320.

Caldwell MW. 2007. Snake phylogeny, origins, and evolution:

the role, impact, and importance of fossils (1869-2006). In:

Anderson JS, Sues H-D, eds. Major transitions in vertebrate

evolution. Bloomington, IN: Indiana University Press, 253

302.

Cundall D, Irish FJ. 2008. The snake skull. In: Gans C,

Gaunt AS, Adler K, eds. Biology of the Reptilia, volume 20,Morphology H. The skull of Lepidosauria. Ithaca: Society for

the Study of Amphibians and Reptiles, 349692.

Cundall D, Rossman DA. 1993. Cephalic anatomy of the

rare Indonesian snake Anomochilus weberi. Zoological

Journal of the Linnean Society 109: 235273.

Cundall D, Wallach V, Rossman DA. 1993. The systematic

relationships of the snake genus Anomochilus. Zoological

Journal of the Linnean Society 109: 275299.

Dessauer HC, Cadle JE, Gans C. 1987. Immunological

relationships of the Uropeltidae. In: Program and Abstracts,

67th Annual Meeting, American Society of Ichthyologists

and Herpetologists, 3rd Annual Meeting, American Elasmo-

branch Society, 2126 June, 1987, The University of the

State of New York, New York State Museum, The State

Education Department, Albany, New York, 43.

Dessauer HC, Gartside DF, Gans C. 1976. Protein evidence

on the genetic diversity and affinities of uropeltid snakes.

American Zoologist 16: 268.

Dumril AMC. 1853. Prodome de la classification des Rep-

tiles Ophidiens. Mmoires de lAcadmie des Sciences de

lInstitut. de France 23: 399536, Plates 12.

Gans C. 1966. Liste der rezenten Amphibien und Reptilien.

Uropeltidae. In: Mertens R, Hennig W, eds. Das Tierreich.

Lieferung 84: 129.

Gans C. 1973. Uropeltid snakes survivors in a changing

world. Endeavour 32: 6065.

Gans C. 1976. Aspects of the biology of uropeltid snakes.

In: Bellairs A dA, Cox CB, eds. Morphology and biology

of reptiles. Linnean Society Symposium Series Number 3.

New York: Academic Press, 191204, Plates 14.

Gower DJ, Vidal N, Spinks JN, McCarthy CJ. 2005. Thephylogenetic position of Anomochilidae (Reptilia: Ser-

pentes): first evidence from DNA sequences. Journal of

Zoological Systemtics & Evolutionary Research 43: 315320.

Groombridge B. 1979a. On the vomer in Acrochordidae

(Reptilia: Serpentes), and its cladistic significance. Journal

of Zoology (London) 189: 559567.

Groombridge BC. 1979b. Variations in morphology of the

superficial palate of henophidian snakes and some possible

systematic implications. Journal of Natural History 13:

447475.

Groombridge BC. 1979c. Comments on the intermandibular

muscles of snakes. Journal of Natural History 13: 477498.

Heise PJ, Maxson LR, Dowling HG, Hedges SB. 1995.

Higher-level snake phylogeny inferred from mitochondrial

DNA sequences of 12S rRNA and 16S rRNA genes. Mole-

cular Biology and Evolution 12: 259265.

Hoffstetter R. 1939. Sur larticulation occipito-vertbrale des

Uropeltid (Ophidiens fouisseurs). Bulletin du Musum

National dHistoire Naturelle, Srie 2 11: 426433.

Jan G, Sordelli F. 1865(1961). Iconographie Gnrale des

Ophidiens. Tome Primier, Livraison 9, Plate 2. [Reprint by J.

Cramer, Weinheim: Wheldon & Wesley, LTD, and New York:

Hafner Publishing Co.].

Joyce WG, Bell CJ. 2004. A review of the comparative

morphology of extant testudinoid turtles (Reptilia:

Testudines). Asiatic Herpetological Research 10: 53109.

Lee MSY. 2005. List of morphological characters. OrganismsDiversity & Evolution 5, Electronic Supplement 4: 1

28.

Lee MSY, Hugall AF, Lawson R, Scanlon JD. 2007. Phy-

logeny of snakes (Serpentes): combining morphological and

molecular data in likelihood, Bayesian and parsimony

analyses. Systematics and Biodiversity 5: 371389.

Lee MSY, Scanlon JD. 2002. Snake phylogeny based on

osteology, soft anatomy and ecology. Biological Reviews 77:

333401.

McDiarmid RW, Campbell JA, Tour TA. 1999. Snake



7/29/2019 52949750

21/22

species of the world: a taxonomic and geographic reference.

Volume 1. Washington, DC: The Herpetologists League.

Mahendra BC. 1938. Some remarks on the phylogeny of the

Ophidia. Anatomischer Anzeiger 86: 347356.

Peters GCH. 1861. De serpentum familia Uropeltaceorum.

Berolini: G. Reimer.

Radovanovic M. 1937. Osteologie des Schlangenkopfes.

Jenaische Zeitschrift fr Naturwissenschaft 71: 179312,

Plate 4.

Rajendran MV. 1970. Notes on uropeltid snakes of

Tirunelveli and Kanyakumari Districts. St. Xaviers College

Palayamkottai Annual April 1970: 5560.

Rajendran MV. 1977. A survey of uropeltid snakes. Journal

of the Madurai University 6: 6873.

Rajendran MV. 1985. Studies in uropeltid snakes. Madurai

Kamaraj University Publication 80. Madurai, India:

Madurai Kamaraj University.

Rajendran MV. 1990. Ecological imbalance and endemism in

uropeltid snakes. Cobra 2: 45.

Rieppel O. 1976. The homology of the laterosphenoid bone in

snakes. Herpetologica 32: 426429.Rieppel O. 1977. Studies on the skull of the Henophidia

(Reptilia: Serpentes). Journal of Zoology (London) 181: 145

173.

Rieppel O. 1978. The evolution of the naso-frontal joint in

snakes and its bearing on snake origins. Zeitschrift fr

zoologische Systematik und Evolutionsforschung 16: 14

27.

Rieppel O. 1979. The evolution of the basicranium in the

Henophidia (Reptilia: Serpentes). Zoological Journal of the

Linnean Society 66: 411431.

Rieppel O. 1980a. The sound-transmitting apparatus in

primitive snakes and its phylogenetic significance. Zoomor-

phology 96: 4562.

Rieppel O. 1980b. The evolution of the ophidian feeding

system. Zoologischer Jahrbucher, Abteilung fr Anatomie

103: 551564.

Rieppel O. 1983. A comparison of the skull of Lanthanotus

borneensis (Reptilia: Varanoidea) with the skull of primitive

snakes. Zeitschrift fr zoologische Systematik und Evolu-

tionsforschung 21: 142153.

Rieppel O, Zaher H. 2000. The intramandibular joint in

squamates, and the phylogenetic relationships of the fossil

snake Pachyrachis problematicus Haas. Fieldiana: Geology

n.s. 43: 169.

Rieppel O, Zaher H. 2002. The skull of the Uropeltinae

(Reptilia, Serpentes), with special reference to the otico-

occipital region. Bulletin of The Natural History Museum,

Zoology Series 68: 123130.

Scanlon JD, Lee MSY. 2000. The Pleistocene serpent

Wonambi and the early evolution of snakes. Nature 403:

416420.

Slowinski JB, Lawson R. 2002. Snake phylogeny: evidence

from nuclear and mitochondrial genes. Molecular Phyloge-

netics and Evolution 24: 194202.

Smith MA. 1943 (1981). The fauna of British India Ceylon

and Burma, Including the Whole of the Indo-Chinese Sub-

region. Reptilia and Amphibia. Vol. III. Serpentes. New

Delhi: Today and Tomorrows Printers and Publishers (origi-

nally published by Taylor and Francis, London).

Tchernov E, Rieppel O, Zaher H, Polcyn MJ, Jacobs LL.

2000. A fossil snake with limbs. Science 287: 2010

2012.

Underwood G. 1967. A contribution to the classification of

snakes. London: Trustees of the British Museum (Natural

History).Vidal N, Hedges SB. 2002. Higher-level relationships of

snakes inferred from four nuclear and mitochondrial genes.

Comptes Rendus Biologies 325: 977985.

Vidal N, Hedges SB. 2004. Molecular evidence for a terres-

trial origin of snakes. Proceedings of the Royal Society of

London B (Supplement) 271: S226S229.

White ME, Kelly-Smith M, Crother BI. 2005. Higher-level

snake phylogeny as inferred from 28S ribosomal DNA and

morphology. In: Donnelly MA, Crother BI, Guyer C, Wake

MH, White ME, eds. Ecology & evolution in the tropics.

A herpetological perspective. Chicago, IL: The University

of Chicago Press, 156173.

Wilcox TP, Zwickl DJ, Heath TA, Hillis DM. 2002. Phy-

logenetic relationships of the dwarf boas and a comparison

of Bayesian and bootstrap measures of phylogenetic

support. Molecular Phylogenetics and Evolution 25: 361

371.

Williams EE. 1959. The occipito-vertebral joint in the bur-

rowing snakes of the family Uropeltidae. Breviora 106:

110.

Zaher H, Rieppel O. 1999. Tooth implantation and replace-

ment in squamates, with special reference to mosasaur

lizards and snakes. American Museum Novitates 3271:

119.



7/29/2019 52949750

22/22

Copyright of Zoological Journal of the Linnean Society is the property of Wiley-Blackwell and its content may

not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.

52949750

Documents