A second species of Eucnemesaurus Van Hoepen, 1920 (Dinosauria, Sauropodomorpha): New information on the diversity and evolution of the sauropodomorph fauna of South Africa's lower Elliot Formation (latest Triassic)

Abstract

The Late Triassic-Early Jurassic Elliot Formation of South Africa is one of the most important geological formations worldwide for understanding the early evolution of sauropodomorph dinosaurs. However, many of the taxa currently recognized as valid within its lower strata remain either poorly understood, vaguely diagnosed, or both. The recent discovery of an articulated partial skeleton of a single individual of the enigmatic lower Elliot genus Eucnemesaurus provides an important opportunity to expand our understanding of the anatomy and phylogeny of this poorly known taxon. A comprehensive investigation of the morphological relationships of this new specimen identified key features, pertaining primarily to the femoral shaft and distal tibia, which distinguish it from the only other previously named species of Eucnemesaurus-E. fortis. A new species, E. entaxonis, is erected within which to accommodate it. A cladistic analysis confirms the monophyly of Eucnemesaurus, as well as its continued inclusion within the low-diversity Riojasauridae. Nonetheless, this result highlights continued uncertainties regarding the constituency of the Riojasaurus hypodigm. The relatively robust pedal architecture of E. entaxonis suggests an unexpectedly early experiment in a slower, subgraviportal form of locomotion within Late Triassic basal Massopoda, whereas the intriguing mosaic of plesiomorphic and derived characters evident in E. entaxonis raises questions regarding the hypothesized population dynamics of the basal-most sauropodomorph taxa of the lower Elliot Formation. This latter concern has particular bearing on newly observed inconsistencies in the prevailing hypodigms of other lower Elliot basal sauropodomorph taxa such as Melanorosaurus.

Full text

This article was downloaded by: [Jonah Choiniere]
On: 07 August 2015, At: 21:15
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick Place,
London, SW1P 1WG
Click for updates
Journal of Vertebrate Paleontology
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/ujvp20
A second species of Eucnemesaurus Van Hoepen, 1920
(Dinosauria, Sauropodomorpha): new information on
the diversity and evolution of the sauropodomorph
fauna of South Africa's lower Elliot Formation (latest
Triassic)
Blair W. McPheeabc, Jonah N. Choiniereab, Adam M. Yatesad & Pia A. Vigliettiabc
a Evolutionary Studies Institute, University of the Witwatersrand, Private Bag 3. 2050,
Johannesburg, Gauteng, South Africa. Email:
b DST/NRF Centre of Excellence in Palaeosciences, University of the Witwatersrand, Private
Bag 3. 2050, Johannesburg, Gauteng, South Africa
c School of Geosciences, University of the Witwatersrand, Private Bag 3. 2050,
Johannesburg, Gauteng, South Africa
d Museum of Central Australia, Araluen Cultural Precinct, P.O. Box 3521. 0871, Alice Springs,
Northern Territory, Australia
Published online: 07 Aug 2015.
To cite this article: Blair W. McPhee, Jonah N. Choiniere, Adam M. Yates & Pia A. Viglietti (2015): A second species of
Eucnemesaurus Van Hoepen, 1920 (Dinosauria, Sauropodomorpha): new information on the diversity and evolution of the
sauropodomorph fauna of South Africa's lower Elliot Formation (latest Triassic), Journal of Vertebrate Paleontology, DOI:
10.1080/02724634.2015.980504
To link to this article: http://dx.doi.org/10.1080/02724634.2015.980504
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions

ARTICLE
A SECOND SPECIES OF EUCNEMESAURUS VAN HOEPEN, 1920 (DINOSAURIA,
SAUROPODOMORPHA): NEW INFORMATION ON THE DIVERSITY AND EVOLUTION
OF THE SAUROPODOMORPH FAUNA OF SOUTH AFRICA’S LOWER ELLIOT
FORMATION (LATEST TRIASSIC)
BLAIR W. McPHEE,*
,1,2,3
JONAH N. CHOINIERE,
1,2
ADAM M. YATES,
1,4
and PIA A. VIGLIETTI
1,2,3
1
Evolutionary Studies Institute, University of the Witwatersrand, Private Bag 3, Johannesburg, Gauteng, 2050, South Africa,
blair.mcphee@gmail.com;
2
DST/NRF Centre of Excellence in Palaeosciences, University of the Witwatersrand, Private Bag 3, Johannesburg, Gauteng, 2050,
South Africa;
3
School of Geosciences, University of the Witwatersrand, Private Bag 3, Johannesburg, Gauteng, 2050, South Africa
4
Museum of Central Australia, Araluen Cultural Precinct, P.O. Box 3521, Alice Springs, Northern Territory, 0871, Australia
ABSTRACT—The Late Triassic–Early Jurassic Elliot Formation of South Africa is one of the most important geological
formations worldwide for understanding the early evolution of sauropodomorph dinosaurs. However, many of the taxa
currently recognized as valid within its lower strata remain either poorly understood, vaguely diagnosed, or both. The recent
discovery of an articulated partial skeleton of a single individual of the enigmatic lower Elliot genus Eucnemesaurus provides
an important opportunity to expand our understanding of the anatomy and phylogeny of this poorly known taxon. A
comprehensive investigation of the morphological relationships of this new specimen identified key features, pertaining
primarily to the femoral shaft and distal tibia, which distinguish it from the only other previously named species of
Eucnemesaurus—E. fortis. A new species, E. entaxonis, is erected within which to accommodate it. A cladistic analysis
confirms the monophyly of Eucnemesaurus, as well as its continued inclusion within the low-diversity ‘Riojasauridae.’
Nonetheless, this result highlights continued uncertainties regarding the constituency of the Riojasaurus hypodigm. The
relatively robust pedal architecture of E. entaxonis suggests an unexpectedly early experiment in a slower, subgraviportal
form of locomotion within Late Triassic basal Massopoda, whereas the intriguing mosaic of plesiomorphic and derived
characters evident in E. entaxonis raises questions regarding the hypothesized population dynamics of the basal-most
sauropodomorph taxa of the lower Elliot Formation. This latter concern has particular bearing on newly observed
inconsistencies in the prevailing hypodigms of other lower Elliot basal sauropodomorph taxa such as Melanorosaurus.
http://zoobank.org/urn:lsid:zoobank.org:pub:68A7F233-2424-469E-A008-442C4E04B02F
SUPPLEMENTAL DATA—Supplemental materials are available for this article for free at www.tandfonline.com/UJVP
INTRODUCTION
The morphological and phylogenetic significance of the sauro-
podomorph dinosaurs of the lower Elliot Formation (hereafter
LEF; Norian/Rhaetian) of South Africa has been successively
reiterated in recent years (Yates, 2003a, 2007a, 2007b; Yates and
Kitching, 2003; Yates et al., 2004, 2010, 2011, 2012; McPhee
et al., 2014). This fauna is composed of an anatomically and tax-
onomically diverse set of animals that ranges from ‘primitive’
taxa occupying basal positions within Sauropodomorpha to rela-
tively derived forms that share some anatomical features with
the specialized bauplan of Sauropoda. The morphological
breadth of this community not only furthers our understanding
of basal sauropodomorph diversity immediately prior to the end
Triassic extinction event, but also is of particular significance in
tracking the nature and tempo of the character changes that led
to the graviportal quadrupedal gait of the gigantic sauropods.
Although our anatomical and systematic knowledge of these
lower Elliot taxa has increased dramatically since the late 20th
century (when every large-bodied form tended to be lumped
within the catch-all taxon of ‘Euskelesaurus browni’ [e.g., Van
Heerden, 1979]), many genera currently recognized as valid
within the LEF remain very poorly understood. Of these, the
most enigmatic is undoubtedly Eucnemesaurus. Although
recently the subject of an updated diagnosis and description
(Yates, 2007a), the only previously recognized species within
the genus—Eucnemesaurus fortis—is only known from sparse,
incomplete material. These remains are restricted to a few iso-
lated vertebrae, a coracoid, a fragmentary pubis, three partial
femora, and two tibiae (of which only the proximal surface of
one is preserved). Unsurprisingly, the majority of diagnostic
traits attributed to Eucnemesaurus are exclusive to the femur,
rendering the taxonomic validity of this genus tentative at
best.
The discovery of a largely complete pelvis, hind limb, and par-
tial vertebral column of a Eucnemesaurus-like sauropodomorph
(BP/1/6234) from the LEF of the Aliwal North district in the
Eastern Cape Province of South Africa provides an opportunity
to advance our understanding of this rare and poorly defined
taxon. We present and describe this material as belonging to a
new species in the genus Eucnemesaurus. We analyze the phylo-
genetic relationships of this new species and examine its implica-
tions for basal sauropodomorph evolution, especially in regards
*Corresponding author.
Color versions of one or more of the figures in this article can be found
online at www.tandfonline.com/ujvp.
Journal of Vertebrate Paleontology e980504 (24 pages)
Óby the Society of Vertebrate Paleontology
DOI: 10.1080/02724634.2015.980504
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

to the interrelationships of other poorly known LEF genera (i.e.,
Plateosauravus, Melanorosaurus).
Additionally, the phylogenetic position of Eucnemesaurus is of
particular interest because it represents one of the two genera
within the taxonomically depauperate ‘Riojasauridae’ (Yates,
2007a, 2007b, 2010) and thus suggests a close biogeographic rela-
tionship between South Africa and South America during the
Late Triassic (Bonaparte, 1971). The additional data afforded by
the new Eucnemesaurus specimen described here represent an
opportunity to more stringently test that phylogenetic relation-
ship and to address questions pertaining to the dispersal and bio-
geography of basal members of Sauropodomorpha within the
Gondwana in the Late Triassic.
History of Eucnemesaurus
During the 1860s, Alfred Brown sent several shipments of
large sauropodomorph bones collected from the Aliwal region
(Eastern Cape) of the lower Elliot Formation to several institu-
tions in Europe. Amongst material that ultimately made its way
to the Naturhistorisches Museum in Vienna was a distinctive
femur (NMW 1889-XV-39) that Huene (1906; see also Cooper,
1980) allocated to ‘Euskelosaurus browni.’ This femur experi-
enced a rather confused taxonomic history for the remainder of
the century, being ultimately reclassified as the herrerasaurid
theropod Aliwalia rex (Galton, 1985) based on the assumed
interrelatedness of this bone to a carnivorous archosaurian max-
illa that Brown had sent in an earlier shipment to the Natural
History Museum in London.
In 2003, a femur (BP/1/6111) that closely resembled that of the
Viennese ‘Euskelosaurus/Aliwalia’ was found in association with
vertebrae clearly diagnostic of Sauropodomorpha, leading Yates
(2007a) to conclude that this was not the first time an Aliwalia-
type femur had been recovered from the lower Elliot alongside
material referable to Sauropodomorpha. Specifically, the long-
forgotten Eucnemesaurus fortis Van Hoepen, 1920, was recog-
nized as displaying a similar morphology and subsequently resur-
rected from its status as one of the many taxa synonymized with
Euskelosaurus browni by Van Heerden in 1979. Based on the
revised understanding of the Eucnemesaurus hypodigm afforded
by this new material, Yates (2007a) also erected the novel taxon
Riojasauridae as a low-diversity (Eucnemesaurus CRiojasau-
rus), pan-Gondwanan clade at the base of Massopoda (the his-
tory of Eucnemesaurus fortis and its associated material is
covered in greater detail in Yates, 2007a). The new specimen
(BP/1/6234) that we hereby ascribe to E. entaxonis was not
included in Yates’ (2007a) resurrection of Eucnemesaurus (due
to being both incompletely prepared and excavated at the time).
Institutional Abbreviations—BPI, Evolutionary Studies insti-
tute, Johannesburg, South Africa (formerly Bernard Price Insti-
tute); GPIT, Institute for Geosciences, Eberhard-Karls-
Universit€
at T€
ubingen, T€
ubingen, Germany (Formerly Geolo-
gisch-Pal€
aontologisches Institut T€
ubingen); IVPP, Institute of
Vertebrate Paleontology and Paleoanthropology, Beijing, Peo-
ple’s Republic of China; MB.R., Museum f€
ur Naturkunde–
Leibniz-Institut f€
ur Evolutions und Biodiversit€
atsforschung an
der Humboldt-Universit€
at zu Berlin; NM QR, National
Museum, Bloemfontein, South Africa; PULR, Museo de Cien-
cias Naturales, Universidad Nacional de La Rioja, La Rioja,
Argentina; PVL, Paleontolog
ıa de Vertebrados, Instituto
‘Miguel Lillo,’ San Miguel de Tucum
an, Argentina; SAM-PK,
Iziko-South African Museum, Cape Town, South Africa; SMNS,
Staatliches Museum fur Naturkunde Stuttgart, Stuttgart, Ger-
many; TM, Ditsong Museum of Natural History, Pretoria, South
Africa (previously Transvaal Museum).
SYSTEMATIC PALEONTOLOGY
DINOSAURIA Owen, 1842
SAURISCHIA Seeley, 1888
SAUROPODOMORPHA Huene, 1932
MASSOPODA Yates, 2007a
‘RIOJASAURIDAE’ Yates, 2007a
EUCNEMESAURUS Van Hoepen, 1920
Diagnosis—Yates (2007a) included a diagnosis for both the
Riojasauridae and E. fortis that centered almost exclusively on
features of the femur. Although the issues surrounding the diag-
nostic validity of Riojasaurus are discussed below, many of these
features remain relevant to the generic diagnosis of Eucnemesau-
rus. These include: presence of a large posterior proximal tuber-
cle on the femoral head (although the phylogenetic expression of
this character has recently come in doubt; see below); proximally
abrupt lesser trochanter that is taller than its basal width; fourth
trochanter rounded in profile and obliquely oriented so that the
proximal half is more medially set than the distal half.
EUCNEMESAURUS ENTAXONIS, sp. nov.
Holotype—BP/1/6234, the articulated remains of most of the
hindquarters of a medium-sized sauropodomorph dinosaur. It
consists of an articulated vertebral column composed of the pos-
terior-most dorsal vertebrae, sacral vertebrae, and the anterior
portion of the tail; partial right ilium; partial pubic apron; left
ischium; right femur and fragments of the left; distal epipodium
(crus); and almost complete right pes (Fig. 1).
FIGURE 1. Reconstruction of Eucnemesaurus
entaxonis. Illustrated bones are those preserved in
the holotype BP/1/6234. Scale bar equals 50 cm.
McPhee et al.—A second species of Eucnemesaurus (e980504-2)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

FIGURE 2. Stratigraphic section of the
Cannon Rock site. Arrow on map indicates
excavation site of BP/1/6234. The large-bod-
ied sauropodomorph is of indeterminate tax-
onomy and cataloged within the ESI
collections as BP/1/7408. The section spans
most of lower Elliot and may include the
basal-most rocks of the upper Elliot
Formation.
McPhee et al.—A second species of Eucnemesaurus (e980504-3)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

Locality and Horizon—The material was unearthed in 2003
during a survey of Elliot rocks on Cannon Rock Farm (owned by
Piet Prinsloo) in the Aliwal North area of the Eastern Cape
Province, Republic of South Africa (RSA). The specimen was
found in grayish-brown mudstone (Munsell color 5YR 3/2)
immediately above a thin crevasse-splay sandstone approxi-
mately 10–15 m above the Elliot-Molteno contact at the base of
the lower Elliot Formation (Fig. 2). The Elliot Formation, along
with the underlying Molteno and overlying Clarens formations,
forms part of the Stormberg group, the youngest depositional
sequence of the Karoo Supergroup (Catuneanu et al., 1998).
Currently considered mid-Norian to Rhaetian in age (Olsen and
Galton, 1984; Lucas and Hancox, 2001; Bordy et al., 2005; Irmis,
2010; McPhee et al., 2014), the lower Elliot can be distinguished
from the Early Jurassic upper Elliot by its thicker, more laterally
persistent mudstone intervals, considerably thicker sandstone
bodies, and comparative lack of pedogenic artifacts. These
geometries are characteristic of a palaeoenvironment dominated
by perennial meandering river systems with extensive floodplains
and overbank areas, evincing a humid to semiarid climate domi-
nated by riparian forests that were able to support large-bodied
animals (Anderson et al., 1998; Bordy et al., 2004, 2006).
Preservation—Many of the periosteal bone surfaces through-
out the assemblage appear to have experienced decay prior to
fossilization, and temerarious preparation has in some rare
instances erroneously removed cortical bone. Preservational
issues stemming from these caveats will be addressed when
applicable throughout the following description.
Diagnosis—Two potential autapomorphies diagnose E. entax-
onis: a deep brevis fossa with relatively thin lateral and medial
walls on the ventral surface of the postacetabular process of the
ilium (a pronounced brevis fossa is also present in the holotype
of Riojasaurus, but this can be distinguished from E. entaxonis
with respect to its expansive mediolateral width and considerably
thicker medial and lateral margins); and a sharp ventral keel on
the centra of the proximal caudal vertebrae (although this region
is poorly preserved).
In addition to these features, E. entaxonis can be further
diagnosed with respect to a unique suite of local autapomor-
phies (given its current position within our phylogeny): a small,
circular pit (‘non-articulating gap’) that excavates the sacral rib
of the first primordial sacral at midheight (present also in Mela-
norosaurus); femoral shaft beneath the fourth trochanter trans-
versely elliptical in cross-section (present in taxa from
Melanorosaurus crownwards); posterior descending process of
the distal tibia does not extend as far laterally as the anterior
ascending process, rendering the latter visible in posterior view
(present in Aardonyx crownwards; convergently acquired in
Anchisaurus); and a semi-stout pes in which the maximum prox-
imal breadth of the first metatarsal is approximately 0.6 times
its proximodistal length (present in several derived basal sauro-
podiform taxa).
E. entaxonis differs from E. fortis in the following features: (1)
E. fortis presents a femur that is subcircular in cross-section
rather than elliptical, and (2) the holotype of E. fortis (TM 119)
displays a distal tibia in which the mediolaterally extensive pos-
terior surface extends as far laterally as the anterior ascending
process (for articulation with the astragalus). It is possible that
the distal embayment present on the fourth trochanter in E. for-
tis is absent in E. entaxonis, but poor preservation in the latter
taxon precludes confirmation of this. Furthermore, a dorsal neu-
ral arch of E. fortis possesses a unique accessory lamina within
the centrodiapophyseal fossa—a feature that cannot be con-
firmed in the preserved material of E. entaxonis. This feature
therefore remains a valid diagnostic character of E. fortis for the
time being.
Etymology—The novel species name alludes to the surpris-
ingly robust foot architecture of this species. An entaxonic pes
(or manus) is one in which the medial digits bear the main
weight-resisting forces and are thus the most developed—a trait
later developed in the extreme within Sauropoda (see e.g., Car-
rano, 2005).
Preparation Methods—Exposed in situ bone was consolidated
using a dilute solution of Paraloid B-72 in acetone solvent. Once
consolidated, the specimen was excavated using hand tools,
including rock hammers, chisels, and shovels. It was removed
from the ground in several major blocks containing the sacrum
and vertebral column, the distal epipodium, and the pes, respec-
tively. These were protected during removal by a layer of news-
paper dampened in water, followed by jackets composed of
layers of burlap and plaster of Paris. Rock matrix was removed
from the specimen in the laboratory primarily with handheld
pneumatic air scribes. Fossilized bone was consolidated using an
approximately 10% solution of Paraloid B-72 solid-grade ther-
moplastic acrylic resin (Rohm and Haas Company, 2007) in
100% acetone solvent. Individual pieces of bone (e.g., neural
spines of proximal vertebrae) were glued together using either
cyanoacrylate (various brands) or a highly concentrated (»30%)
solution of Paraloid B-72 in 100% acetone solvent (remainder of
specimen).
PHYLOGENETIC METHODS
The data matrix for this analysis was drawn from the data
matrix originally introduced by Yates (2007b) and subsequently
employed (with various alterations) by a number of other sauro-
podomorph workers (Smith and Pol, 2007; Ezcurra, 2010; Yates
et al., 2010; Apaldetti et al., 2011; Otero and Pol, 2013; Mart
ınez
et al., 2012; McPhee et al., 2014). It was further modified in the
current study via the alteration of several preexisting characters
and the addition of two novel ones (see below). The phylogenetic
matrix, comprising 53 taxa and 362 characters, was analyzed with
TNT 1.1 (Goloboff et al., 2008) using a heuristic search of 1000
replicates of Wagner trees followed by TBR branch swapping
with 10 trees saved per replication. Characters were equally
weighted. The following 40 multistate characters were treated as
ordered: 8, 13, 19, 23, 40, 57, 69, 92, 102, 117, 121, 131, 144, 147,
149, 150, 157, 162, 167, 170, 177, 205, 207, 222, 227, 234, 242, 251,
254, 267, 280, 301, 307, 315, 335, 348, 351, 353, 358, 362. Addi-
tional analyses were performed on the data set in order to assess
the robustness of the position of Eucnemesaurus and other
related taxa. These included implied weighting, constraint analy-
ses, and, in the case of Riojasaurus, the a priori removal of a
taxon.
The novel character additions are as follows:
Character 180: ‘Deep, medially directed pit excavating the sur-
face of the non-articulating gap of the first primordial sacral rib:
absent (0); present (1).’ This addition is intended to capture a
morphology that has now been observed in both Eucnemesaurus
and Melanorosaurus.
Character 335: ‘Size of first metatarsal: maximum proximal
breadth less than 0.4 times its proximodistal length (0); maxi-
mum proximal breadth between 0.4 and 0.7 times its proximodis-
tal length (1); maximum proximal breadth greater than 0.7 times
its proximodistal length (2).’ There is a crownward tendency
within Sauropodomorpha towards a more robust, stout foot
architecture in which the transverse width of the first metatarsal
increases relative to its proximodistal length (curiously reversed
in the basal sauropod genera Vulcanodon and Tazoudasaurus), a
trend that has not been adequately expressed in previous
matrices.
Four preexisting characters were altered in order to more fully
capture our improved understanding of sauropodomorph evolu-
tion and variation. The character ‘pneumatic excavation of the
dorsal neural arches’ (char. 162; Yates, 2007b) was altered to
more clearly express the varied nature of invasive pneumaticity
McPhee et al.—A second species of Eucnemesaurus (e980504-4)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

within both basal and derived sauropodomorphs (see Yates
et al., 2012). An additional state was added to the character
‘shape of the proximal articular surface of the tibia’ (char. 307
[Yates, 2007b:char. 305]) in order to homologize a subset of this
ratio feature and better differentiate between the considerable
variation of the archosaur proximal tibia, which ranges from
being as transversely wide as it is anteroposteriorly long to over
twice as anteroposteriorly long as it is wide. An additional state
was added to the character ‘projection of ventral flange on proxi-
mal surface of second metatarsal’ (char. 341 [Smith and Pol,
2007:char. 354]). As originally defined, this character only codi-
fied for the presence of a laterally directed flange on the ventral
surface of the proximal metatarsal II—a feature seen almost
exclusively in massospondylids. The additional character state is
intended to homologize a ventromedial flange that is significantly
more developed than the ventrolateral flange, as is observed in
other basal sauropodomorph taxa. Finally, an additional state
was added to ‘length of the ungual of pedal digit one’ (char. 353
[Yates, 2007b:char. 344]). The relative length of this element
varies widely across Sauropodomorpha, from shorter than some
non-terminal phalanges (e.g., Anchisaurus), to longer than the
first metatarsal (e.g., Blikanasaurus). The additional state homo-
logizes this latter morphology in a range of taxa. See Supplemen-
tary Data for the full list of character descriptions and their
respective states.
DESCRIPTION OF EUCNEMESAURUS ENTAXONIS
Sources of comparative information employed throughout the
description are referenced in Table 1. See Table 2 for selected
measurements of BP/1/6234.
Axial Skeleton—The axial skeleton of E. entaxonis is repre-
sented by an articulated series of vertebrae consisting of the two
posterior-most dorsal vertebrae (probably D13 and D14), a dor-
sosacral, two primordial sacrals, and the first five anterior caudal
vertebrae (Figs. 3 and 4). An additional, isolated, more posterior
caudal (from approximately the middle of the caudal series) is
also present. The proximal ends of the ribs associated with the
posterior dorsals are preserved in close proximity to the dia-
pophyses on the left side of each vertebra. The distal ends have
been displaced ventrally, approximating the lateral side of their
respective centra.
Dorsal Vertebrae—Although relatively complete, some
aspects of the anatomy of the dorsal vertebrae are obscured
due to a combination of postmortem deterioration and the
articulated manner of preservation. This applies especially to
the morphology of the articular facets of the zygapophyses,
the proportions of the hyposphene, and the morphology
of the anterior and posterior faces of the neural arches
generally.
The amphicoelous centra of the two posterior-most dorsals are
both subequal in their relative dorsoventral height and antero-
posterior length, the plesiomorphic morphology for non-
sauropodan Sauropodomorpha with the exception of the Masso-
spondylidae, which tend to have dorsal centra that are longer
than they are tall (Apaldetti et al., 2013). The transverse width
of the anterior and posterior articular surfaces is approximately
0.68 times their dorsoventral height, a metric also typical of most
basal sauropodomorphs, especially in dorsals of the posterior
series (Apaldetti et al., 2013). Although the internal structure of
the centra cannot be discerned, it is clear from the absence of
any pleurocoelous fossae or foramina on the lateral side of the
elements that they are acamerate.
TABLE 1. Sources of comparative data used in this study.
Taxon Source(s)
Saturnalia tupiniquim Langer, 2003
Panphagia protos Mart
ınez and Alcober, 2009
Thecodontosaurus antiquus Benton et al., 2000
Pantydraco caducus Yates, 2003b
Plateosauravus cullingworthi SAM- PK 3341–3356, 36 02–3603, 3607–3609
Plateosaurus engelhardti Huene, 1926; Yates, 2003c; Mallison,
2010a, 2010b
Eucnemesaurus fortis BP/1/6107, 6110–6115, 6220; TM 119
Riojasaurus incertus PVL 3808; Bonaparte, 1971
Leonerasaurus taquetrensis Pol et al., 2011
Massospondylus carinatus BP/1/4377, 4693, 4924, 4934, 4998, 5000,
5241; Cooper, 1981
Coloradisaurus brevis PVL 5904 (field no. 6); Apaldetti et al.,
2013
Lufengosaurus hueni Young, 1941
Adeopapposaurus mognai Mart
ınez, 2009
Glacialisaurus hammeri Smith and Pol, 2007
Yunnanosaurus huangi Young, 1942
Jingshanosaurus xinwaensis Zhang and Yang, 1994
Mussaurus patagonicus Otero and Pol, 2013
Aardonyx celestae BP/1/386; various elements cataloged BP/
1/5379–6893
Anchisaurus polyzelus Galton, 1976; Yates, 2004, 2010
Melanorosaurus readi NM QR1551, 3314; SAM 3449, 3450, 3532
Blikanasaurus cromptoni SAM-PK 403
Lessemsaurus sauropoides PVL 4822; Pol and Powell, 2007
Antetonitrus ingenipes BP/1/4952a, b, c
Vulcanodon karibaensis Cooper, 1984
Tazoudasaurus naimi Allain and Aquesbi, 2008
TABLE 2. Selected measurements of BP/1/6234.
Dorsal vertebra ?13
Dorsoventral height of centrum »95
Transverse width of centrum (may be taphonomically
compressed)
»68
Anteroposterior length of centrum »91
Anteroposterior length of base of neural spine »65
Dorsoventral height of neural spine »76
Total height of vertebra 230
Dorsosacral vertebra
Length of base of neural spine »48
Height of neural spine »96
First primordial sacral vertebra
Length of base of neural spine »69
Height of neural spine »100
Second primordial sacral vertebra
Length of base of neural spine 77
Height of neural spine 106
First caudal vertebra
Height of centrum »100
Width of centrum (may be taphonomically compressed) »80
Length of centrum »75
Ischium
Total length »400
Dorsoventral depth distal end 115
Transverse width distal end 75
Femur
Total length 535
Transverse width femoral shaft 86
Anteroposterior depth femoral shaft 64
Tibia
Transverse with distal tibia 117
Anteroposterior depth medial side distal tibia 70
Anteroposterior depth lateral side distal tibia »45
Pes
Maximum proximal breadth metatarsal 1 60
Total length metatarsal 1 92
Proximal dorsoventral depth metatarsal 2 73
Total length metatarsal 2 149
Transverse width of midshaft metatarsal 2 37
Total length of metatarsal 3 »170
Transverse width of proximal surface of metatarsal 5 36
Total length metatarsal 5 74
All measurements in mm.
McPhee et al.—A second species of Eucnemesaurus (e980504-5)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

The dorsal neural spines of E. entaxonis are mediolaterally
thin, lacking the transverse dorsal expansion seen in the derived
sauropodiform taxa Antetonitrus (BP/1/4952) and Lessemsaurus
(Pol and Powell, 2007). The spines are directed primarily
dorsally, and whereas the anterior margin displays a subtle ante-
roventral inclination, the posterior margin is essentially dorso-
ventrally straight in lateral view. This distinguishes E. entaxonis
from a number of basal sauropodomorph taxa that possess dorsal
neural spines with a concave posterior margin in lateral view and
a projecting posterodorsal corner (e.g., Riojasaurus;Plateosaurus
engelhardti [see Yates, 2003a]). The posterior dorsal neural
spines are approximately as long at the base (measured antero-
posteriorly) as they are dorsoventrally tall. This contrasts with
the dorsal neural spines of the slightly more derived
FIGURE 3. Axial column and pelvic elements of Eucnemesaurus entaxonis (BP/1/6234) in left lateral view. Abbreviations: c1, first caudal vertebra;
c2, second caudal vertebra; c3, third caudal vertebra; c4, fourth caudal vertebra; c5, fifth caudal vertebra; ch, chevron; ?d13, probable thirteenth dorsal
vertebra; ?d14, probable fourteenth dorsal vertebra; dr, dorsal rib; ds, dorsosacral vertebra; il, right ilium; ipi, left ischial peduncle of the ilium; nag,
non-articulating gap; ps1, first primordial sacral vertebra; ps2, second primordial sacral vertebra; sr1, first sacral rib; sr2, second sacral rib. Dashed lines
represent uncertainty in natural bone surface. Scale bar equals 10 cm.
McPhee et al.—A second species of Eucnemesaurus (e980504-6)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

Melanorosaurus, which exhibits a neural spine base/height ratio
of just under 1.5 (Galton et al., 2005; seen also in NM QR3314).
However, it is still higher than in other basal sauropodomorph
taxa such as Riojasaurus, Massospondylus, and the unnamed pro-
sauropod from the LEF (BP/1/4953; see Yates, 2003a), which
tend to have dorsal neural spines that are longer at the base than
dorsoventrally tall. Poor preservation means that it cannot be
confirmed if E. entaxonis possesses the incipient keel on the
anterior margin of the neural spines observed in the posterior
dorsals of Antetonitrus and some specimens of Plateosauravus
(van Heerden, 1979).
The main body of the posterior dorsal neural arches of E.
entaxonis lacks the pronounced anteroposterior constriction
seen in Aardonyx, Antetonitrus, and more derived taxa. The pre-
and postzygapophyses are preserved in articulation with one
another, which obscures many of their morphological attributes.
FIGURE 4. Axial column and pelvic elements of Eucnemesaurus entaxonis (BP/1/6234) in right lateral view. Abbreviations: ac, acetabulum; bf, bre-
vis fossa; c1, first caudal vertebra; c2, second caudal vertebra; c3, third caudal vertebra; c4, fourth caudal vertebra; c5, fifth caudal vertebra; ch, chevron;
?d13, probable 13th dorsal vertebra; il, right ilium; opi, right ischium obturator plate; pcdl, posterior centrodiapophyseal lamina; ps2, second primor-
dial sacral vertebra; sr2, second sacral rib; tp, transverse process. Dashed lines represent uncertainty in natural bone surface. Wavy lines represent
exposed, internal bone. Scale bar equals 10 cm.
McPhee et al.—A second species of Eucnemesaurus (e980504-7)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

The anterior-most projection of the prezygapophyses appears to
have been relatively flush with the level of the anterior margin of
the centrum in both posterior dorsals. As is typical of sauropodo-
morph posterior dorsal vertebrae, the parapophyses are not
located on the centrum and are presumed to be located directly
ventrally to the diapophyses, although this is obscured by the
close proximity of the dorsal ribs. The diapophyses of the left
side of both vertebrae are poorly preserved but clearly visible
slightly below the level of the zygapophyses. The right side of
the anterior-most preserved dorsal vertebra bears a centro-infra-
diapophyseal fossa, which is bounded posteriorly by the poste-
rior centrodiapophyseal lamina, although both have been
displaced ventrally during preservation (Wilson, 1999; Yates
et al., 2012). A middle dorsal vertebra of the holotype of E. fortis
was described as possessing a small accessory lamina that
branched off of the paradiapophyseal lamina, partially dividing
the centro-infradiapophyseal fossa (Yates, 2007a). This was
treated as a diagnostic feature of E. fortis (another accessory
lamina located within the posterior infradiapophyseal fossa was
later reinterpreted as the external margin of a putatively pneu-
matic subfossa; Yates et al., 2012). Unfortunately, the level of
preservation, along with the adherent orientation of the dorsal
ribs, precludes confirmation of a similar structure in E. entaxonis.
The neural canal is low and subcircular in cross-section, and
the flanking pedicles are low. In Aardonyx and more derived sau-
ropodomorphs, the neural canals are more slot-like in cross-sec-
tion and are generally accompanied by a concomitant increase of
the distance between the diapophyses and the base of the arch.
Overall, the neural arches of E. entaxonis are approximately
1.4 times the height of their accompanying centra. This ratio
approaches the condition observed in Melanorosaurus but falls
radically short of the ratio displayed by Antetonitrus, which has
neural arches that are over twice as tall as the associated dorsal
centra. Most other sauropodomorph taxa (including Riojasau-
rus) have dorsal neural arches that are roughly subequal to the
height of their centra. Two posterior dorsal vertebrae located
amongst the Plateosauravus assemblage (SAM-PK-3607) are
remarkable for displaying neural arches that are approximately
1.57 times the height of their centra. This is a relatively high met-
ric for basal Sauropodomorpha and suggests either the conver-
gent acquisition of a derived vertebral morphology in a taxon for
which the majority of remaining character information is rather
plesiomorphic, or the presence of more than one taxon within
that assemblage (see Yates, 2003a).
The morphology of the head of the ribs (capitulum and tuberc-
ulum) cannot be discerned because of their close articulation
with the dorsal vertebrae. The shaft of these posterior elements
is flat and dorsoventrally compressed, being considerably wider
than high. It appears that none of the dorsal vertebrae of E.
entaxonis were ribless, as has been suggested for the posterior-
most dorsals of some taxa (e.g., Plateosaurus [Huene, 1926];
Yates, 2007b).
Sacral Vertebrae—Portions of a dorsosacral and two primor-
dial sacral vertebrae are preserved. Although the dorsosacral is
missing its centrum and portions of the neural arch (the former
is possibly represented as an amorphous block of spongy bone),
the position of the dorsosacral in relation to the adherent right
ilium argues strongly for its participation in the sacrum. Thus, E.
entaxonis likely presents the ‘typical’ basal sauropodomorph
arrangement of a dorsosacral plus two primordial sacrals (Pol
et al., 2011). The two more posterior sacral vertebrae are rela-
tively more complete, preserving the general morphology of
their respective sacral ribs.
The neural spines of the sacral series are markedly taller than
the neural spines of the dorsal series, as generally occurs at the
dorsosacral transition in sauropodomorph dinosaurs. The height
of the dorsosacral neural spine of E. entaxonis is approximately
1.60 times the anteroposterior length of its base. Ventral to the
neural spine the prezygapophyses can be seen in tight articula-
tion with the postzygapophyses of the posterior-most dorsal, but
the deterioration of the transverse process and sacral rib has
erased all information pertaining to the remainder of the neural
arch and centrum.
Most of the lateral components of the sacral ribs of the first
primordial vertebra have been eroded away, and only the base is
well preserved. The base extends dorsoventrally as an hourglass-
shaped (in lateral view) plateau that is bordered both anteriorly
and posteriorly by deep concavities. The anterior concavity is
clearly a natural feature of the skeleton, but the posterior con-
cavity may have been produced by over-preparation. This sug-
gests that, when complete, the sacral rib would have housed an
anteroposteriorly expanded dorsal process (Dtransverse pro-
cess) that likely roofed the intercostal cavity both anteriorly and
posteriorly, as seen in a number of basal sauropodomorph taxa
(e.g., Riojasaurus;Leonerasaurus).
A deep circular pit is positioned directly at midheight of this
plateau, dividing it into dorsal and ventral halves. A distinct rim
of bone bordering this pit suggests that this feature is real, and
that it is thus probably homologous to the ‘non-articulating gap’
observed by Yates (2007a) in referred specimens of Melanoro-
saurus (NM QR1551 and NM QR3314). Although a great many
sauropodomorph taxa (e.g., Riojasaurus; Yates, 2007b) display
some form of intercostal fenestra (sensu Wilson, 2011) between
the dorsal and ventral sections of the first primordial sacral rib,
this medially directed excavation of the rib itself appears to be a
much rarer occurrence (see Discussion below).
Presaging the condition of the neural spines of the caudal
series, the neural spine of the second primordial sacral vertebra
is more posteriorly inclined than the first, as seen in most non-
sauropodan sauropodomorphs (e.g., Plateosaurus [Huene, 1926];
Lufengosaurus;Melanorosaurus [NM QR1551]). The neural
arch of this element appears to have been better preserved than
in the two anterior sacrals, and although either temerarious prep-
aration or poor preservation has removed the majority of the
cortical bone surface, the general morphology nonetheless
remains clearly distinguishable. The most distinctive feature of
the second primordial neural arch is the strong posterolateral
projection of the transverse processes, a geometry seen in the
final non-caudal sacral of most basal sauropodomorphs (e.g.,
Lufengosaurus, Melanorosaurus, Plateosaurus). This process
partially roofs the same deep, expansive concavity that borders
the posterior margin of the first primordial sacral rib. Although
most sauropodomorph genera present a large intercostal aper-
ture in the region between the two primordial sacral ribs, it is
possible that this aperture in BP/1/6234 has been artificially exag-
gerated due to poor preservation. The intercostal concavity is
bordered ventrally by the second primordial sacral rib, which
rises posterodorsally at an oblique angle towards the transverse
process, with which it is fused. This ‘en echelon’ morphology of
the second primordial sacral rib is common to basal Sauropodo-
morpha, also being seen in forms such as Riojasaurus, Lufengo-
saurus, and Yunnanosaurus. The dorsal displacement of the
ilium has exposed the right sacral rib of the second primordial
sacral vertebra, although this same displacement has removed
the articulation of the latter with the medial surface of the for-
mer. Nonetheless, general agreement in contours between the
lateral articular surface of the rib and the posteroventral corner
of the ilium suggests that the rib articulated along most of the
ventral extent of the postacetabular process (Fig. 4). This would
have left only a minimal amount of space in which to accommo-
date a potential caudosacral rib—an arrangement that we con-
sider unlikely (see below).
Caudal Vertebrae—The first five caudals are present as an
articulated series. The anterior-most caudal is in poor condition
and most of its cortical surface is missing. The remaining four cau-
dals appear to retain the majority of their cortical bone surface.
McPhee et al.—A second species of Eucnemesaurus (e980504-8)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

The shallowly amphicoelous centrum of the second anterior
caudal vertebra is approximately 1.4 times dorsoventrally taller
than wide at the anterior articular facet. It is also taller than it is
anteroposteriorly long, with a height/length ratio of approxi-
mately 1.6. These ratios are fairly typical for anterior-most cau-
dals in basal Sauropodomorpha, although the mediolateral
constriction of the centra is comparatively high. Although it is
possible that this is an effect of taphonomic distortion, a mid-
anterior caudal from the Spioenkop Eucnemesaurus (BP/1/6220;
Yates, 2007a) displays a similar height/width ratio of 1.3. The
ventral surface of the centra of C2, C3, and C4 (surface not pre-
served in C1) is sharply delineated into an acute median ridge.
Although taphonomic compression cannot be ruled out, if natu-
ral, this keel would appear to be the opposite condition of certain
derived sauropodomorphs (e.g., Melanorosaurus [SAM-PK-
3449]; Vulcanodon) that exhibit a ventral furrow on the under-
side of the anterior centra (Yates, 2004). This feature represents
a possible autapomorphy of E. entaxonis.
The mediolaterally thin neural spines of C1, C2, C3, and C4
are high and posterodorsally sloped, with their posterior projec-
tion far exceeding the level of the posterior margin of the cen-
trum. A sharp keel is visible along the anterior margin of the
neural spines, but it is possible that this is an artefact of mediolat-
eral compression during diagenesis. If it is indeed present, this
would be homologous with the same feature observed in BP/1/
6220 (Yates, 2007a).
The neural arches occupy the great majority of the dorsal sur-
face of their respective centra, with the prezygapophyses projec-
ting slightly anterior to the anterior margin of the centrum.
However, as with the dorsal and sacral series, the finer details of
the zygapophyses of the caudal vertebrae are obscured due to
being in articulation with one another. The manner of preserva-
tion of the transverse processes in the holotype of E. entaxonis is
curious. On the left side of C2 and C3, the transverse processes
have been displaced ventrally and compressed to the neurocen-
tral suture in such a fashion as to superficially resemble para-
pophyses (recalling Hatcher’s [1903] suggestion that the
transverse processes of anterior caudal vertebrae are composed
of the coalesced diapophyses and parapophyses; see Wilson,
2011). In the anterior-most caudal vertebra, this same compac-
tion gives the transverse process the appearance of having possi-
bly contacted the ilium as a caudosacral rib. However, on the
right side of C1, the transverse process can be seen as a dorsally
flattened arm that, in life, most likely exhibited the laterally
flared, tapered morphology typical of caudal transverse pro-
cesses. This dorsally oriented dislocation of the transverse pro-
cesses is repeated throughout the right side of the preserved
caudal series and gives the transverse processes the appearance
of not having been fully fused to the neural arch at death.
Because the transverse processes of the caudal vertebrae of prac-
tically all vertebrates ossify as a single unit with the neural arch,
this peculiar arrangement is interpreted here as simply repre-
senting a unique breakage pattern.
The fifth caudal in the series most closely resembles the ele-
ment referred to Eucnemesaurus and figured and described in
Yates’ previous study (2007a; BP/1/6220). Besides sharing a gen-
eral equivalence in proportions, these elements both have an
enlarged chevron facet on the posteroventral corner of the cen-
trum that hangs well below the level of the anteroventral corner
of the centrum. Based on observed morphology, it is also possi-
ble that, as in BP/1/6220, C5 in E. entaxonis exhibited the same
shallow, paramedian fossae and associate flanking ridges either
side of the base of the anterior margin of the neural spine. How-
ever, poor preservation renders this difficult to quantify. It is also
possible that this morphology became more prevalent in poste-
rior elements, and because it appears that the Spioenkop Eucne-
mesaurus was a relatively larger individual than E. entaxonis
(based on the comparative sizes of the distal femora), BP/1/6220
likely derives from a more posterior position in the series—likely
»C6–C8.
Chevrons are present between all the preserved caudal verte-
brae with the exception C1–C2. This is consistent with all known
basal sauropodomorphs (at least those for which the anterior tail
is complete) with the exception of Plateosaurus, which, due to
the anterior-most caudal vertebra being incorporated into the
sacral unit as a caudosacral, possesses a true chevron (i.e., not
just a lenticular oval intercentrum between the first and second
caudal vertebrae; Huene, 1926; Galton, 1999; Mart
ınez, 2009).
The chevrons are preserved articulated to their respective centra,
and although obscured distally by adherent matrix, it is clear that
they displayed the proximodistally elongate, transversely com-
pressed morphology typical to all basal Sauropodomorpha.
FIGURE 5. Detail of right ilium of Eucnemesaurus entaxonis (BP/1/
6234) in A, lateral view and B, ventral view. Abbreviations: ac, acetabu-
lum; bf, brevis fossa; pap, postacetabular process. Scale bar equals 5 cm.
McPhee et al.—A second species of Eucnemesaurus (e980504-9)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

Ilium—The main body of the right ilium, although preserved
in semi-articulation with the first and second primordial sacral
vertebrae, has been dislocated dorsally and thus sits at an oblique
angle above these elements (Figs. 4, 5). The anterior portion of
the preacetabular process and both the pubic and ischial
peduncles of the right ilium are missing. Only small portions of
the left ilium are preserved—the isolated pubic peduncle was
located amongst loose material accessioned under BP/1/6234,
and the ischial peduncle is adhered to the opposite side of the
block to the right ilium (albeit in a much more ventral position;
Fig. 3). Unfortunately, the medial surface of the ilium is too
poorly preserved to distinguish sacral attachment scars or other
related morphologies.
The ilium of E. entaxonis is dorsoventrally low with a straight,
non-convex dorsal margin, as in the majority of basal sauropodo-
morphs. It is possible that the anterior portion of the dorsal mar-
gin may have displayed a distinct ‘step’ similar to that seen in
Riojasaurus and some specimens of Massospondylus; however,
its incompleteness makes this difficult to confirm.
Dorsal to the acetabular region the lateral surface of the iliac
blade is deeply concave, typical of the condition in most non-sau-
ropodan sauropodomorphs. The posterior margin of the antero-
posteriorly elongate postacetabular process is semicircular in
lateral view. This contrasts with the morphology of forms such as
Plateosauravus (SAM-PK-3609) and Plateosaurus (Huene, 1926)
where the postacetabular process is posteriorly squared in lateral
view and bears a sharp posteroventral corner. A subtle protuber-
ance can be discerned in the center of the lateral surface of the
postacetabular process, likely related to the origin point of the
flexor tibialis musculature (Carrano and Hutchinson, 2002;
Langer, 2003).
Perhaps the most notable aspect of the ilium in E. entaxonis is
the deep, strongly emarginated brevis fossa that extends along
the ventral surface of the postacetabular process for the entirety
of its length. The lateral wall of this fossa (homologous to the
‘brevis shelf’; e.g., Novas, 1996; Mart
ınez et al., 2012) appears to
have been more strongly developed than the medial, and at its
tallest point measures approximately 2 cm from its ventral mar-
gin to the base of the fossa. The morphology of the brevis fossa
in E. entaxonis differs from the same structure in the holotype of
Riojasaurus (PVL 3808), which, although of proportionally simi-
lar dorsoventral depth, is mediolaterally expanded to an almost
autapomorphic extent for sauropodomorph dinosaurs (see
Discussion).
The acetabulum is laterally expanded with a well-developed
supracetabular crest along its anterior margin. As in a number of
basal sauropodomorphs, this crest (based on the morphology of
the dislocated left pubic peduncle) appears to have extended
anteriorly along the posterolateral margin of the pubic peduncle
until a point just short of its ventral termination. The pubic
peduncle is ovoid to teardrop-shaped in cross-section. As in Col-
oradisaurus (Apaldetti et al., 2013), it is possible that a mediolat-
erally narrow flange of bone may have been present on the
posteromedial margin of this peduncle, although the possibility
remains that this is simply a product of taphonomic excavation
of the posterior surface.
The ischial peduncle is dorsoventrally extensive, as is plesio-
morphic for Sauropodomorpha. Although its distal end is poorly
preserved, it appears to lack the distinct posteroventral ‘heel’
seen in forms such as Plateosauravus (SAM-PK-2609) and
Plateosaurus.
Pubis—A poorly preserved proximal portion (the iliac pedun-
cle) of one pubis and scrappy sections of the pubic apron of both
pubic blades were recovered with the specimen (Fig. 6).
Although poorly preserved, the pubic blades clearly show that E.
entaxonis retained the mediolaterally wide, transversely oriented
apron typical of all non-eusauropodan sauropodomorphs. The
lateral margins of the pubic apron appear to have been relatively
straight in anterior or posterior views, lacking the concavity pres-
ent in Massospondylus (BP/1/4924) and Mussaurus (Apaldetti
et al., 2013; Otero and Pol, 2013).
Ischium—The left ischium is relatively complete except for
part of the proximal end (Dobturator plate), which, aside from
missing most of its proximal articular processes, has also experi-
enced the complete loss of cortical bone (Fig. 7). However, the
head of the right ischium is preserved adhered to the main sacral
block (Fig. 4).
Although generally better preserved than the same region of
the left ischium, the pubic and iliac peduncles of the right proxi-
mal end are poorly preserved and the acetabular margin is
merely hinted at. Nonetheless, the preserved morphology of the
ventral margin of the obturator plate suggests that the distal end
was smoothly confluent with respect to the proximal ischial shaft,
therefore lacking the distinct notch that ventrally separates the
two primary ischial bodies seen in Plateosaurus and Riojasaurus
(Yates, 2003c; Apaldetti et al., 2013; Otero and Pol, 2013). As in
the majority of non-eusauropodan sauropodomorphs, the shaft is
subtriangular in cross-section with a broad and flat medial sym-
physis and an acutely convex lateral margin (becoming more
obtuse towards the distal end). In contrast, derived members of
Sauropoda develop flattened, blade-like ischial shafts that are
transversely wider than dorsoventrally deep throughout their
length (observed also in Anchisaurus; Yates, 2004). An irregular
sulcus excavates the dorsal surface of the proximal third of the
ischial shaft. This is interpreted as the remains of the proximal
ischial groove that is present in the majority of saurischian dino-
saurs (Langer and Benton, 2006; Otero and Pol, 2013).
FIGURE 6. Isolated fragments of the pubic apron of Eucnemesaurus
entaxonis (BP/1/6234). A, proximal portion of left pubis in ventral view;
B, section of the pubic blade of the ?right pubis in dorsal view. Scale bar
equals 5 cm.
McPhee et al.—A second species of Eucnemesaurus (e980504-10)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

Immediately distal to its connection with the obturator plate
the dorsoventral depth of the proximal ischial shaft is approxi-
mately 0.74 times its mediolateral width. This value increases
throughout the proximodistal length of the shaft, with the dorso-
ventral height of the subovoid distal end being ultimately
1.4 times its transverse width. However, a number of basal taxa
(e.g., Plateosaurus;Massospondylus) display proportionately
narrower distal ischial shafts in which the dorsoventral depth is
over twice the mediolateral width.
This expansion of the ischial shaft appears to be conducted pri-
marily along the ventral surface. This imparts a subtle, ventrally
oriented deflection to the distal end of the shaft and thus a con-
cave ventral margin of the shaft when viewed laterally. A similar
ventral deflection can be seen in the ischia of Coloradisaurus
(Apaldetti et al., 2013). In contrast, other basal sauropodomorph
taxa (e.g., Massospondylus [BP/1/4693]; Mussaurus) tend to
exhibit ischial shafts with a dorsally expanded distal end and
thus a concave dorsal margin in lateral view. The dorsal margin
of the distal end extends further distally than the ven-
tral—similar to that of a number of basal sauropodomorph forms
(e.g., Plateosaurus;Coloradisaurus;Massospondylus).
Femur—A nearly complete right femur and fragments of the
left femur (notably including the fourth trochanter) are pre-
served (Figs. 8, 9). Because femora are known from the holotype
of E. fortis (TM 119), as well as the referred E. fortis specimens
NMW 1889-XV-39 and BP/I/6111, this element represents our
best means of assessing the degree to which BP/I/6234 can be
confidently integrated within the Eucnemesaurus hypodigm.
The preservation of the femoral head in E. entaxonis is curi-
ous, being either heavily eroded or strongly anteroposteriorly
compressed (or both). An accreted nodule (possibly left as a pre-
paratory ‘control’) is adhered to the anterolateral corner of the
femoral head. Because this sedimentary mass bears no marked
textural difference to the anterior ‘surface’ of the head, it is likely
that this portion of the femur decomposed prior to fossilization.
The posterior surface of the femoral head is slightly better pre-
served, showing some periosteal bone surface. There is a small,
posteriorly directed bulge (in either posterior or proximal view)
located towards the middle of the head, roughly congruent with
the medial margin of the femoral shaft (Fig. 9A). This corre-
sponds to the ‘posterior tubercle’ of Yates (2007a), a feature that
was posited as representing “a reversal to the non-dinosaurian
condition” (Yates, 2007a:94; see Discussion below).
Given the poor preservation of the femoral head, it cannot be
said with confidence if E. entaxonis displayed the same mediolat-
erally elongate morphology described for specimens of E. fortis
(Yates, 2007a). Additionally, the sharp mediodistal edge of the
femoral head—where the ligaments of the caput femoris would
have inserted—appears to be composed primarily of matrix,
meaning that caution should be used before attributing to E.
entaxonis the opposite condition to the rounded, indistinct medi-
odistal corner of the femoral head seen in E. fortis (e.g., BP/1/
6111).
Distal to the femoral head there is a ca. 40 mm section of the
anterior shaft missing, although the proximal portion of the
lesser trochanter is well preserved. The lesser trochanter rises
abruptly from the anterior surface of the femoral shaft from a
position lateral to the midline, with its proximal termination dis-
tal to the distal edge of the femoral head. Both of these features
are congruent with E. fortis. Additionally, as in E. fortis (BP/1/
6111), the lateral surface of the lesser trochanter in E. entaxonis
is considerably deeper (i.e., more anteroposteriorly extensive)
than the medial surface (Fig. 9B). It is probable that the trochan-
ter as a whole was much higher than mediolaterally wide, a
feature consistent with Yates’ (2007a) diagnosis of the Riojasaur-
idae. In contrast, many other basal sauropodomorphs exhibit a
lesser trochanter in which the proximal end is more gradually
sloped towards the femoral shaft and the main body of the pro-
cess is wider than it is tall (e.g., Plateosauravus [SAM-PK-3602,
3603]; Aardonyx [BP/1/6510]; Melanorosaurus [SAM-PK-3450]).
However, this latter morphology is often observed primarily in
the distal half of the trochanter, a section that is poorly preserved
in the lesser trochanter of E. entaxonis. A small yet distinct notch
is located at the proximal-most tip of the lesser trochanter and
may be homologous to the trio of small, rounded spurs men-
tioned by Yates (2007a) in BP/1/6111, although the possibility
remains that this is merely an artefact of preservation.
All of the fourth trochanter of the right femur is preserved, but
it is slightly eroded and its finer anatomical details are hard to
make out. An isolated section of the left femoral shaft presents a
very well preserved fourth trochanter (Fig. 9C, D) of equal pro-
portions, and we base our description largely on that element.
Consistent with the primitive sauropodomorph condition, the
fourth trochanter in E. entaxonis is located almost entirely within
the proximal half of the femoral shaft. Its proximal end is posi-
tioned on the medial margin of the femoral shaft, whereupon it
extends distolaterally so that the distal end terminates slightly
medial to the midline of the posterior surface of the shaft. This is
accompanied by a slight curvature of the long axis of the fourth
FIGURE 7. Left ischium of Eucnemesaurus entaxonis (BP/1/6234) in
lateral view. Abbreviations: de, distal expansion; ip, iliac peduncle. Scale
bar equals 5 cm.
McPhee et al.—A second species of Eucnemesaurus (e980504-11)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

FIGURE 8. Right femur of Eucnemesaurus entaxonis (BP/1/6234) in A, anterior view; B, medial view; C, posterior view; and D, lateral view. Abbreviations:
4t, fourth trochanter; lt, lesser trochanter; pt, posterior tubercle. Dashed lines represent uncertainty in natural bone surface. Scale bar equals 10 cm.
McPhee et al.—A second species of Eucnemesaurus (e980504-12)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

trochanter such that the proximal half is oriented more obliquely
than the distal half, whereas the latter is set subparallel to the
proximodistal axis of the femoral shaft. Both the oblique orienta-
tion of the fourth trochanter and the curvature of its long axis are
thus considered diagnostic of Eucnemesaurus at the genus level.
The profile of the fourth trochanter of the right femur is
obscured on account of its outer margin having been eroded;
however, the dislocated fourth trochanter of the fragmentary left
element clearly displays (when viewed anteriorly) the Eucneme-
saurus-like rounded profile that distinguishes this taxon from
other basal sauropodomorphs, which display a more rectangular-
shaped fourth trochanter. Unfortunately, preservation is insuffi-
cient to determine if E. entaxonis displayed the same distinctly
notched, pendant-shaped distal termination of the fourth tro-
chanter to that seen in BP/1/6111.
Distal to the fourth trochanter the anteroposterior depth of
the femoral shaft is approximately 0.75 times its mediolateral
width. This degree of transverse eccentricity approaches the
condition of derived taxa such as Melanorosaurus and Anteto-
nitrus. In contrast, the well preserved femoral midshaft of a
specimen of E. fortis (BP/1/6111) displays a depth/width ratio
of 0.89, almost identical to that of Plateosauravus (SAM-PK-
3602). Although repairs to the midshaft of the femur are evi-
dent due to the presence of large quantities of glue (poten-
tially resulting in a superficially heightened degree of
eccentricity), well-preserved sections of femoral shaft adja-
cent to the repaired section strongly suggest that the midshaft
was moderately eccentric in cross-section, and that this mor-
phology is therefore more variable amongst non-sauropodan
sauropodomorphs than previously appreciated.
The posterior components of the distal femoral condyles are
not preserved, and although they appear to have been mediolat-
erally extensive, this is likely an exaggeration due to anteropos-
terior crushing during diagenesis. The attachments for the
extensor musculature can be seen in the very shallow depression
positioned centrally on the distal end of the anterior surface of
the femur.
Tibia—The tibia of E. entaxonis is incompletely represented.
The distal end of the right tibia is preserved in articulation with
distal limb elements (Fig. 10). A partial proximal end is pre-
served, but because the anterior portion is missing (including the
cnemial crest), it is impossible to say with confidence whether it
is of the right or left side. However, one side of the proximal sur-
face is expanded in such a manner that it roofs a shallow depres-
sion on the proximal surface of the shaft. It is possible that this
morphology is consistent with the large fibular condyle observed
in the tibia of TM 119, suggesting that the preserved element in
E. entaxonis is of the left side. As in the corresponding element
in the holotype of E. fortis, this condyle is centrally located and
does not extend to the posterior margin of the proximal surface.
The posterior expansion of the posterior margin of the proximal
surface is consistent with a number of derived non-sauropodan
sauropodomorphs (Melanorosaurus;Antetonitrus), whereas
basal forms such as Saturnalia and Panphagia (but also seen in
FIGURE 9. Details of diagnostic features of
the femur of Eucnemesaurus entaxonis (BP/
1/6234). Proximal portion of right femur in
A, proximal view and B, posterior view. Iso-
lated fourth trochanter of left femur in C,
posterolateral view and D, medial view.
Abbreviations: 4t, fourth trochanter; fs, fem-
oral shaft; lt, lesser trochanter; pt, posterior
tubercle. Scale bars equal 5 cm (Aand B)
and 2 cm (Cand D).
McPhee et al.—A second species of Eucnemesaurus (e980504-13)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

FIGURE 10. Right distal epipodium of Eucnemesaurus entaxonis (BP/1/6234) in A, anterior view and B, posterior view. Abbreviations: aap, astraga-
lar ascending process; apt, ascending process of the distal tibia; ca, calcaneum; dpt, descending process of the distal tibia; fib, fibula; pmf, posteromedial
facet of the distal tibia; tib, tibia. Gray represents matrix/eroded bone. Scale bar equals 5 cm.
McPhee et al.—A second species of Eucnemesaurus (e980504-14)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

some large-bodied forms such as Plateosaurus and Coloradisau-
rus) tend to have posteriorly flattened, subquadrangular poste-
rior margins.
The distal end of the right tibia, including about 15 cm of its
distal shaft and a similar amount of the distal shaft of the adja-
cent fibula, is preserved in articulation with the astragalus dis-
tally and the distal end of the right fibula laterally. The anterior
half of an additional 13 cm of the right tibial shaft remains
embedded in a block of matrix and articulates precisely with the
other preserved tibial material. The shaft is strongly elliptical,
being wider mediolaterally than anteroposteriorly. Accordingly,
the transverse width of the distal articular surface far exceeds its
anteroposterior depth, with the anteroposterior depth of the
medial margin exceeding that of the lateral margin, as is primi-
tive for basal Sauropodomorpha. E. entaxonis also lacks the
anterolateral deflection of the anterior ascending process
(Dfacet for articulation with the astragalus), which, in some
derived basal sauropodomorphs (e.g., Antetonitrus;Melanoro-
saurus [SAM-PK-3449]), produces a subtriangular distal articular
surface of the tibia in which the depths of the medial and lateral
margins are rendered relatively subequal.
Nonetheless, the lateral margin of the distal end of the tibia in
E. entaxonis is notable insofar as the anterior ascending process
extends further laterally than the posterior descending process,
rendering the former clearly visible in posterior view. This differs
from the plesiomorphic condition for Sauropodomorpha, in
which the descending process is laterally expanded beyond the
level of the anterior ascending process and the anterior ascend-
ing process is not visible in posterior view. The condition
observed in E. entaxonis is, in fact, more similar to the derived
condition seen in Anchisaurus, Aardonyx, Melanorosaurus,
Antetonitrus, and Sauropoda (e.g., Yates and Kitching, 2003;
Yates, 2004). This morphology also distinguishes E. entaxonis
from TM 119 (E. fortis), which preserves a mediolaterally expan-
sive posterodistal end in which the ascending and descending dis-
tal processes are essentially level with each other (Van Hoepen,
1920; Yates, 2007a:fig. 1). Although the relationship of the distal
to the proximal end of the bone is unknown in BP/1/6234, and it
is possible that the reduced lateral expansion of the descending
process may have been taphonomically exaggerated by an
abstruse angle of preservation, the possibility remains that this
feature has a more varied distribution than previously recog-
nized within basal nodes of Sauropodomorpha—and possibly
even within individual genera. A further distinguishing feature
between E. fortis and E. entaxonis can be seen in the facets that
divide the distal edge of the posterior surfaces of the tibia: in E.
fortis this surface is divided into two broad, subequal facets that
are separated by a subtle, centrally located ridge, whereas in E.
entaxonis this same ridge is located medially, resulting in a con-
siderably reduced medial facet and a greatly expanded lateral
one (Fig. 10).
The lateral margin of the descending process is straight, lack-
ing the beveled morphology seen in the distal tibiae of the
‘Plateosauravus’ assemblage (SAM-PK-3341, 3349). As in most
non-sauropodan sauropodomorphs, the anteromedial corner of
the distal end is acute. The morphology of the distal articular sur-
face is obscured because the tibia is in articulation to the main
body of the astragalus.
Fibula—The distal fibula is a morphologically simple element
that in cross-section has a subcircular shaft and an anteroposter-
iorly expanded distal end (Fig. 10). There is a distinct swelling
evident on the posterodistal corner of the bone, and this connects
to a ridge that runs dorsomedially along the anterior face of the
shaft about 15 mm from the distal end. However, these latter
two features are interpreted as either taphonomic or diagenetic
artifacts.
Astragalus—Although the astragalus was clearly articulated
with the epipodium at death, most of its cortical bone has not
been preserved, with only the ascending process visible beneath
the ascending process of the tibia (Dfacet for reception of the
ascending process of the astragalus) (Fig. 10). Given that the
ascending process occupies an anterior position upon the astrag-
alus in sauropodomorph dinosaurs, the absence of any bone pos-
terior to this process in E. entaxonis is illustrative of the amount
of diagenetic attrition experienced by this element. The ascend-
ing process appears to have been transversely flat in posterior
view, lacking the posterolateral concavity seen in a number of
sauropodomorph taxa (e.g., Coloradisaurus;‘Aardonyx’ [BP/1/
386]).
Calcaneum—The calcaneum is unremarkable for basal Sauro-
podomorpha, being subtriangular in ventral view with a tapered
medial margin. The lateral surface is twice as long as it is tall.
Mild pitting in the form of two shallow fossae can be seen on the
lateral surface; however, it is difficult to substantiate the degree
of taphonomic exaggeration present in this feature.
Pes—An almost complete and articulated right foot is pre-
served (Fig. 11). However, although each digit can be readily dis-
tinguished, it is clear that some of the bones have experienced
the same distorting process evident in other parts of the skeleton
of E. entaxonis, resulting in the ‘withered’ appearance of pedal
digits III and IV.
Metatarsal I (mt I) is a relatively short element. Its widest
proximal breadth (measured obliquely from the ventromedial to
the dorsolateral corner of the proximal articular surface) is 0.61–
0.62 times its total proximodistal length. Although this is an
unexpectedly low ratio (e.g., Yates, 2008) for a basal sauropodo-
morph, it is nonetheless appreciably more gracile than the first
metatarsals of the comparably derived lower Elliot taxa Blikana-
saurus and Antetonitrus, which display an mt I proximal-width/
length ratio of 0.89 and 0.77, respectively. Other non-sauropodan
sauropodomorphs (in particular the Massospondylidae) tend to
display considerably more gracile dimensions (e.g., Massospon-
dylus: 0.38 [BP/1/5241]; Lufengosaurus: 0.39 [Young, 1941]; Mus-
saurus: 0.41 [Otero and Pol, 2013]; Plateosaurus: 0.46 [SMNS
13200, GPIT1]).
Interestingly, Riojasaurus (PVL 3526) possesses a relatively
elongate mt I with a proximal-breadth/length ratio of approxi-
mately 0.45, and Yates (2008) records the equivalent metric in
Melanorosaurus (NM QR1551) as 0.58, comparable in length to
E. entaxonis. However, the referred, fully articulated skeleton of
Melanorosaurus (MN QR 3314; Yates, 2007b) presents a first
metatarsal proximal-breadth/length ratio of approximately 0.80
(B.W.M., pers. observ.) (character conflict in the Melanorosau-
rus hypodigm is discussed in greater detail in Discussion).
As in most non-sauropodan sauropodomorphs, the subovoid
proximal articular surface of mt I in E. entaxonis is twisted dorso-
laterally with respect to the transverse axis of the distal condyles,
resulting in acute dorsal and ventral margins when approximat-
ing a ‘living’ orientation. The proximal surface is also orthogonal
to the main axis of the shaft, differing from more derived, gravi-
portal forms (e.g., Antetonitrus CSauropoda) that tend to exhibit
an anteriorly sloping proximal surface of the first metatarsal that
is obliquely oriented with respect to the proximodistal axis of the
bone. As in most basal sauropodomorphs, the proximal end of
the first metatarsal is volumetrically smaller than the proximal
surfaces of metatarsals II and III.
The shaft is markedly elliptical in cross-section and narrows
mediolaterally as it extends distally. The minimum transverse
width of the shaft is 0.40 times the total length of the bone. This
metric is intermediate for non-sauropodan taxa, being similar to
the dimensions displayed by the Early Jurassic sauropodiforms
Jingshanosaurus (0.46) and Aardonyx (0.45), but broader than
many other basal sauropodomorphs (e.g., Pantydraco [Yates,
2003b]; Anchisaurus [Galton, 1976]; Lufengosaurus [LV 003];
Massospondylus [BP/1/4377]; Plateosaurus [Huene, 1926]),
which tend to display a minimum midshaft width of between 0.21
McPhee et al.—A second species of Eucnemesaurus (e980504-15)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

FIGURE 11. Right pes of Eucnemesaurus entaxonis (BP/1/6234) in A, dorsal view and B, ventral view. C, hypothesized relationships of the proximal
metatarsus. Abbreviations: mtI, metatarsal I; mtII, metatarsal II; mtIII, metatarsal III; mtIV, metatarsal IV; mtV, metatarsal V; ph1.1, first phalanx of
pedal digit I; ph1.2, second phalanx of pedal digit I; ph2.1, first phalanx of pedal digit II; ph2.2, second phalanx of pedal digit II; ph2.3, third phalanx of
pedal digit II. Scale bar equals 5 cm.
McPhee et al.—A second species of Eucnemesaurus (e980504-16)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

and 0.31 total length. In contrast, the ‘near-sauropods’ Blikana-
saurus and Antetonitrus display an even stouter ratio of approxi-
mately 0.50–0.53.
Poor preservation and adherent matrix obscures an accurate
examination of the differential morphology of the distal con-
dyles. However, it is clear that the more proximally located
medial condyle was considerably reduced compared with the lat-
eral condyle, as is typical for basal Sauropodomorpha. A subtle
extensor depression can be seen in the middle of the anterior
face of the distal end.
The proximal articular surface of metatarsal II is dorsoven-
trally elongate. Based on its articulation with the first metatarsal,
it is obvious that the medial margin of the proximal articular sur-
face is deeply concave. Unfortunately, the lateral margin cannot
be as easily discerned, making it difficult to substantiate whether
or not the proximal surface of metatarsal II in E. entaxonis con-
formed to the biconcave (hourglass) morphology generally found
in basal Sauropodomorpha. It appears, however, that the lateral
margin was distinctly less concave than the medial, as is generally
observed in most basal sauropodomorphs (Fig. 11). The ventral
margin of the proximal end is less symmetrical than the dorsal
margin, possessing a large flange of bone that extends medially
and thus cradles the ventral margin of metatarsal I. A similar
medially directed flange can be seen in the proximal second
metatarsals of a number of non-massospondylid sauropodo-
morphs (e.g., Leonerasaurus;Melanorosaurus [NM QR1551];
Aardonyx [BP/1/6253]; Antetonitrus;Tazoudasaurus [contra
Allain and Aquesbi, 2008]; cf. Smith and Pol, 2007; Pol et al.,
2011). In contrast, the taxa grouped together within the Masso-
spondylidae (e.g., Massospondylus;Lufengosaurus;Coloradisau-
rus;Glacialisaurus) are united in having a ventrolateral flange
that is significantly more developed than the ventromedial pro-
cess (Smith and Pol, 2007). Other sauropodomorphs (e.g., Pla-
teosaurus;Riojasaurus; Blikanasaurus) display a ventral margin
of the proximal metatarsal II that is subequal in its development
at the medial and lateral corners.
The shaft of metatarsal II is subcircular in cross-section, simi-
lar to the condition in Melanorosaurus but different from a num-
ber of other sauropodomorph taxa that display a quadrangular
shaft in cross-section (e.g., Massospondylus;Coloradisaurus;Bli-
kanasaurus;Antetonitrus). The minimum transverse width of the
shaft is approximately 0.25 times the maximum length of the
bone, similar in basic proportions to the majority of other Elliot
taxa (e.g., Plateosauravus;Melanorosaurus;Aardonyx;Antetoni-
trus) with the exception of Massospondylus (0.17). Taxa such as
Riojasaurus and Plateosaurus also have comparably gracile sec-
ond metatarsals with a minimum shaft width/total length ratio of
approximately 0.18.
The distal condyles display a strong medial cant, as is seen—to
greater and lesser degrees—in the distal condyles of metatarsals
II–IV of all sauropodomorph dinosaurs. Neither condyle is
markedly differentiated from the other, although the medial con-
dyle appears to have been more dorsoventrally extensive than
the lateral. No collateral ligament pit can be observed on either
condyle, although the lateral extent of the lateral condyle (where
the deepest pit is generally located) is possibly abraded.
The proximal surface of metatarsal III—the longest bone in
the pes—is partially obscured in a manner similar to that of
metatarsal II. Nonetheless, it appears that the outline of the
proximal surface of metatarsal III in E. entaxonis was roughly an
isosceles triangle with a short dorsal margin and extensive medial
and ventrolateral margins that meet ventrally to form an acute
angle. This same morphology is seen in a number of non-sauro-
podan sauropodomorph taxa (e.g., Massospondylus [e.g., BP/1/
4377]; Antetonitrus;Plateosauravus), whereas a number of mas-
sospondylid genera display a subtrapezoidal proximal metatarsal
III with an obtuse ventral margin (see Smith and Pol, 2007). The
‘near-sauropod’ Blikanasaurus also displays an obtuse ventral
margin of proximal metatarsal III, although this may represent a
corollary of the convex lateral margin, which is divided into two
discrete facets that articulate firmly with the bifacial medial sur-
face of metatarsal IV—a possible autapomorphy of that genus.
Because Herrerasaurus displays a trapezoid-shaped metatarsal
III with a wide ventral margin (Novas, 1994), it is possible that
the acute, triangular morphology observed in some sauropodo-
morphs represents a derived feature of those taxa.
The length of metatarsal I relative to metatarsal III remains
conservative throughout basal Sauropodomorpha, and the ratio
of 0.56 observed in E. entaxonis is roughly consistent with a great
many closely related genera (e.g., Plateosaurus;Riojasaurus;
Massospondylus [BP/1/4377]; Aardonyx; Antetonitrus). Outside
of Eusauropoda, only Blikanasaurus displays a heightened mt I/
mt III ratio, with a value of 0.63.
The shaft of metatarsal III in E. entaxonis, although having
experienced a degree of taphonomic deflation, is subtriangular
in cross-section with a flat, broad dorsal surface and a rounded
ventral surface. As in Glacialisaurus, the medial side is slightly
broader than the lateral. The minimum midshaft width is
approximately 0.19 times the proximodistal width of the bone
(although this metric is tentative considering the aforemen-
tioned deflation). This is similar to the same measurement in
Plateosauravus (0.18), being greater than that observed for Mas-
sospondylus (0.15; BP/1/4377, BP/1/5241) and less than the com-
paratively robust Antetonitrus (0.21) and Blikanasaurus (0.25).
The distal end is badly deformed, although the medial condyle
appears to have been more transversely extensive than the lat-
eral condyle.
Metatarsal IV is poorly preserved, with the morphology of the
proximal end obscured due to its adherence to the underside of
the fibula and the distal end being badly eroded. As in most basal
sauropodomorphs, the fourth metatarsal is only negligibly
shorter than the third.
The funnel-shaped fifth metatarsal is preserved adhered to the
underside of the distal end of metatarsal IV. As with all non-sau-
ropodan sauropodomorphs (and most basal dinosaurs), metatar-
sal V is considerably shorter than the rest of the metatarsus
(although, it bears repeating, in no genera is it ever appreciably
shorter than the first metatarsal). The transverse width of the
proximal surface of the fifth metatarsal in E. entaxonis is
0.5 times its total proximodistal length. This is consistent with
most basal forms (e.g., Massospondylus: 0.53; BP/1/5241).
Derived forms such as Blikanasaurus and Vulcanodon (and most
likely Antetonitrus, although the distal end of metatarsal V is not
preserved in that taxon) have a comparatively wider fifth meta-
tarsal, with a width/length ratio of approximately 0.75–0.77. The
medial and lateral margins of metatarsal V in E. entaxonis taper
smoothly from the proximal to the distal ends. This differs from
a number of genera (e.g., Blikanasaurus;Antetonitrus) where the
proximal half expands at a larger angle (from the proximodistal
axis) than that observed in the distal half of the bone (Yates,
2003b). Unlike Massospondylus, metatarsal V of E. entaxonis
appears to have lacked the faint oblique ridge that runs distome-
dially from the ventrolateral corner of the proximal end towards
midshaft, whereupon a pronounced ventrally oriented swelling
of the attenuated distal half is generally observed. Instead, a
deeply concave pit can be observed within the proximal half of
the ventral surface of E. entaxonis, but this may be a preservatio-
nal artefact.
Digits I and II preserve their full allotment of non-terminal
phalanges (one in the former, two in the latter) as well as their
respective unguals, although the preservation of the latter in
digit II is poor. Also present is an isolated first pedal phalanx
from digit I of the left foot that is preserved in articulation with
its associated ungual, which in turn is missing the distal portion.
In keeping with the rest of the pedal morphology, the non-ter-
minal phalanges of E. entaxonis are relatively squat, approaching
McPhee et al.—A second species of Eucnemesaurus (e980504-17)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

proportions more typical of the derived sauropodiform condition
(e.g., Aardonyx;Melanorosaurus;Antetonitrus) than that of the
comparatively gracile Massospondylidae (e.g., Massospondylus;
Coloradisaurus). The maximum transverse width of the proximal
end of the first phalanx of digit II (the best preserved non-termi-
nal phalanx of the articulated pes) is subequal to its proximodis-
tal length. This contrasts with the same ratio in both
Massospondylus (0.65; BP/1/5241) and Plateosaurus (0.68;
GPIT1), as well as many other basal sauropodomorphs (e.g.,
Pantydraco;Riojasaurus;Adeopapposaurus;Seitaad;Leyesaurus;
Mussaurus) that tend to display pedal phalanges that are proxi-
modistally longer than transversely broad (Apaldetti et al.,
2013). The dorsal surface of the first phalanx of digit I is appreci-
ably narrower than the ventral surface and coplanar with respect
to the dorsal margins of the distal condyles. The difference in
mediolateral width between the dorsal and ventral surfaces may
be less in other phalanges; however, the level of preservation
makes this difficult to confirm. In all observable elements, the
ventral incursion of the distal condyles exceeds that of the dorsal.
The better-preserved, isolated first pedal phalanx from the left
foot exhibits deep collateral ligament fossa on both distal con-
dyles. In the articulated right pes, preservation of this element is
only sufficient to observe ligament fossa on the medial condyles
of the first phalanges of the first and second digits.
The first pedal ungual (I-2) of E. entaxonis is preserved in both
the articulated right pes and the isolated left digit, although in
both it is missing the distal apex. Although incomplete, this ele-
ment would have undoubtedly been shorter than the first meta-
tarsal, a symplesiomorphy of all sauropodomorphs basal to
Melanorosaurus (McPhee et al., 2014). The proximal surface of
ungual I-2 is laterally compressed and triangular in outline with
a narrow dorsal margin and comparatively broad ventral surface.
This differs from more derived forms (e.g., Melanorosaurus;
Antetonitrus) that display a more circular-shaped proximal end
of the first pedal ungual. The proximodorsal process is pointed
and well developed, in contrast to the practically absent flexor
tubercle—a morphology consistent with most basal Sauropodo-
morpha. The beginnings of a distinct vascular groove can be seen
on the lateral and medial surfaces of the better-preserved left
element, just proximal to the break.
The proximal portion of pedal ungual II is preserved in articu-
lation with the non-terminal phalanx; however, its poor preserva-
tional condition precludes any useful comment on its
morphology.
PHYLOGENETIC RESULTS
The phylogenetic analysis returned eight most parsimonious
trees (MPTs) of length 1244, consistency index (CI) D0.34, and
rescaled consistency index (RCI) D0.7 (Fig. 12). Overall, the
topologies of these trees do not differ greatly from previous cla-
distic analyses of basal sauropodomorphs (e.g., Yates, 2007a,
2007b; Yates et al., 2010; Otero and Pol, 2013; McPhee et al.,
2014). The genus Eucnemesaurus is monophyletic, with the sis-
ter-taxon relationship between E. entaxonis and E. fortis sup-
ported by a single unambiguous synapomorphy: length of first
caudal centrum less than its height (char. 184.1; but see below).
The family Riojasuridae are also recovered as a monophyletic
group, diagnosed by essentially the same femoral features listed
in Yates (2007a) (i.e., longitudinal axis of the femur offset less
than ten degrees [char. 280.1; novel to this analysis]; hemispheri-
cal femoral head [char. 283.1]; posterior ‘tubercle’ on femoral
head [char. 284.0; but see below]; lesser trochanter at least as
high as its basal width [char. 289.1]; rounded profile of the fourth
trochanter [char. 295.0]).
Interestingly, novel positions are hypothesized in this topology
for Plateosauravus, Ruehlia, Anchisaurus, and Leonerasaurus.
Plateosauravus and Ruehlia are located within the monophyletic
Plateosauridae in the current topology, not basal to it as in previ-
ous analyses (e.g., Yates, 2007a, 2007b; Otero and Pol, 2013).
Anchisaurus, used as the specifier taxon for the group Anchisau-
ria in most recent taxonomic treatments (e.g., Yates, 2007a,
2007b, 2010; Otero and Pol, 2013), is resolved in a novel position
basal to the Massospondylidae. Leonerasaurus is recovered in a
highly derived position near the base of Sauropoda, a somewhat
more derived position than in previous analyses (Pol et al., 2011;
Otero and Pol, 2013) (see below for further discussion). It should
also be noted that, congruent with the results of Mart
ınez et al.
(2012), the problematic basal saurischian Eoraptor was resolved
as one of the basal-most sauropodomorphs currently known
(Fig. 12A).
Our exploratory analyses yielded the following results: analy-
sis of the data matrix with the Riojasauridae constrained to be
non-monophyletic results in 65 MPTs with a best score only a
single step longer than in the initial analysis (1245). The majority
of the fundamental MPTs in that analysis place Eucnemesaurus
in a position at the base of Sauropodiformes, a position consis-
tent with the relatively derived condition of several characters
within E. entaxonis (Fig. 12B) (see below for further discussion).
It should also be noted that in the constrained analysis the mono-
phyly of Eucnemesaurus is not supported in a number of funda-
mental MPTs, with E. fortis considerably more basal with
respect to E. entaxonis in these instances. Accordingly, the strict
consensus tree places both taxa in a sizeable polytomy along
with most non-sauropodan massopods. Similar to a subset of the
constrained analysis, the monophyly of Eucnemesaurus is also
not supported in the implied weighting analyses, with E. entaxo-
nis considerably more derived than E. fortis in all tested scenar-
ios (K D1¡10). This result is not altogether surprising, however,
given the apomorphies shared between E. entaxonis and Mela-
norosaurus (see also below).
Analysis of the data matrix with the removal of Riojasaurus
(see Discussion) leads to a dramatic increase in the number of
MPTs (from 8 to 123) and a much more poorly resolved consen-
sus tree, again with Eucnemesaurus spp. placed in a considerably
more derived position amongst the core grouping of non-sauro-
podan sauropodiforms. Constraining against the monophyly of
the Riojasauridae results in the ‘traditional’ positioning of
Anchisaurus within Sauropodiformes, whereas its position
remains unchanged in the reduced majority consensus tree in
which Riojasaurus is absent.
DISCUSSION
E.entaxonis and the ‘Riojasauridae’
The features supporting Eucnemesaurus monophyly, much as
those diagnosing the genus itself, pertain predominantly to the
femur (see Results above). These same features, with the excep-
tion of the obliquely directed long axis of the fourth trochanter
(not included in the current data matrix), are also the primary
diagnostic features of the Riojasauridae. Given the continued
paucity of information pertaining to the non-hindquarter regions
of Eucnemesaurus spp., it is therefore likely that many of the
unknown features of Eucnemesaurus are being polarized in phy-
logenetic analysis by the substantially more complete Riojasaur-
us—probably influencing the phylogenetic position of
Eucnemesaurus considerably. For example, the sole unambigu-
ous synapomorphy uniting E. fortis and E. entaxonis within the
Riojasauridae pertains to the first caudal centrum being higher
than it is long. However, this is the typical condition for most
non-massospondylid sauropodomorphs and is rendered autapo-
morphic here simply because of the opposite state apparently
being present in Riojasaurus. Because there are no caudal verte-
brae figured in Bonaparte (1971), it is difficult to substantiate the
scoring of this character (originally scored from the mounted dis-
play skeleton within the museum of the Instituto Miguel Lillo,
McPhee et al.—A second species of Eucnemesaurus (e980504-18)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

Tucum
an [PVL])—a concern that applies to a number of ele-
ments amongst the vast array of material referred to Riojasaurus.
This uncertainty is emblematic of the myriad issues surround-
ing the Riojasaurus hypodigm, and the validity of phylogenetic
inferences dependent upon its inclusion. Riojasaurus was first
introduced in a preliminary fashion as a component of a larger
study of the tetrapod fauna of the Los Colorados Formation of
the Ischigualasto-Villa Uni
on Basin (Bonaparte, 1971). Since
then, only one additional study has explicitly dealt with Riojasau-
rus, when Bonaparte and Pumares (1995) referred a fairly com-
plete skull (PURL 56; found in association with an almost
complete subadult skeleton) to the genus. Although easily differ-
entiated from the contemporaneous Los Colorados taxa Colora-
disaurus and Lessemsaurus with respect to the relatively gracile
bauplan of the former, and the comparatively derived morphol-
ogy in the latter, Riojasaurus remains too poorly represented
within the current literature to confirm the monospecificity of
the large amount of material referred to it (see Bonaparte, 1971).
This uncertainty was recently underscored by author A. M. Yates,
who has expressed doubt regarding the systematic equivalence of
the femoral element(s) used to score Riojasaurus (using non-type
material within the collections of the Paleontolog
ıa de Vertebra-
dos, Instituto San de Miguel de Tucum
an, based on assumed
synonymy with PVL 3808) within the current matrix.
The above considerations have particular bearing on the pre-
sumed phylogenetic significance of the ‘posterior tubercle’—the
obtuse bulge on the posterior side of the femoral head that Yates
(2007a) suggested as a synapomorphic reversal to the non-dino-
saurian condition supporting a monophyletic Riojasauridae.
Although certain non-dinosaurian dinosauromorphs (e.g., Lager-
peton) have a distinct notch on the posterior surface of the femo-
ral head (Dthe medial tuberosity: Sereno and Arcucci, 1994),
most early dinosaurs—including basal sauropodomorphs—have
at most a subtle protuberance on the posterior surface of the
femoral head (although the homology of the latter to the former
is uncertain). Although it appears that in several specimens of
sauropodomorph this protuberance is moderately more pro-
nounced, its expression can be highly variable both within and
between taxa (e.g., the assemblage BP/1/386 [see Otero et al.,
2015] preserves three femoral heads, two without the feature and
one with it, while the femur of Rheulia [MB.R.4753] also pre-
serves a pronounced posterior tubercle). The possibility that
family- or genus-defining status has been afforded a feature that
may have only become exaggerated in individual specimens is
therefore a valid concern. Although this character remains as a
synapomorphy of the Riojasauridae for now, further investiga-
tion is into its homology and distribution within basal sauropodo-
morph taxa is warranted.
FIGURE 12. Results of cladistic analyses. A, abbreviated strict consensus tree of 8 MPTs with a best score of 1244 steps. Numbers below nodes rep-
resent Bremer support values higher than 1. Tree begins at basal-most nodes within Sauropodomorpha; B, 50% majority consensus when the mono-
phyly of the ‘Riojasauridae’ has been constrained against (65 MPTs with best score of 1245 steps).
McPhee et al.—A second species of Eucnemesaurus (e980504-19)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

The a priori removal of Riojasaurus from the current data
matrix, although dramatically affecting the overall resolution of
the analysis, ultimately favors a position for Eucnemesaurus
close to the base of Sauropodiformes—and closer to other Elliot
taxa such as Aardonyx and Melanorosaurus. The fact that this
same result occurs with the addition of a single step when con-
straining for the non-monophyly of the Riojasauridae empha-
sizes the degree to which Riojasaurus (as currently understood
and coded) has significant bearing on our understanding of char-
acter polarity and sister-taxon relationships within non-sauropo-
dan Sauropodomorpha. Clearly, a detailed, monographic
treatment (ideally one that codes PURL 56 and PVL 3808 as
separate operational taxonomic units [OTUs] in order to assess
their character-based affinities) of all the material associated
with Riojasaurus is required in order to corroborate the taxo-
nomic validity of Riojasurus, as well as strengthening phyloge-
netic hypotheses that center upon this taxon.
Novel Positions of Previously Existing Taxa—Several of our
phylogenetic results present novel positions for key taxa, and
these warrant further discussion. The strict consensus tree places
Plateosauravus and Ruehlia within the monophyletic Plateosaur-
idae, differing from all previous analyses (e.g., Yates, 2007a,
2007b, 2010; Otero and Pol, 2013; McPhee et al., 2014) that place
these taxa basal to this group. Although only weakly supported
by two synapomorphies (laterally expanded tables at the mid-
length of the dorsal surface of the neural spines in both cervical
and pectoral vertebrae [char. 149.2; known only for Ruehlia and
convergently acquired in Massospondylidae]; posteriorly projec-
ting heel on the distal end of the iliac ischial peduncle [253.1;
present also in Riojasaurus and some massospondylids]), this
result nonetheless suggests an expanded, global distribution of
plateosaurids that also encompasses southern Africa—a result
geographically congruent with the phylogenetic hypotheses of
Novas et al. (2011) (although see below for a discussion on prob-
lems pertaining to Plateosauravus).
Anchisaurus is resolved in a non-‘anchisaurid’ (sensu Yates,
2007a, 2007b, 2010) position basal to the Massospondylidae. The
derived position of the Massospondylidae relative to Anchisau-
rus is supported by several unambiguous synapomorphies (raised
postorbital rim of the orbit [char. 55.1]; absence of a deep septum
spanning the interbasipterygoid space [char. 85.0; process pres-
ent in Anchisaurus, Riojasaurus, Plateosaurus, and Efraasia];
length of the humerus 55–65% the length of the femur [char.
205.1]; proximal width of the first metacarpal 80–100% of its
length [227.2]; size of the ungual of pedal digit III less than 85%
of the ungual of pedal digit II in all linear dimensions [359.1]).
Two of these characters, however, are highly homoplastic
throughout Sauropodomorpha (205, 359), whereas a small grade
consisting of Massospondylidae, Yunnanosaurus, and Jingshano-
saurus all display the derived condition for characters 55 and 205
(including Mussaurus in the case of the latter), despite this condi-
tion being reversed in all more apical Sauropodomorpha. The
condition in Anchisaurus in these trees is therefore interpreted
as being plesiomorphic within Sauropodomorpha, rather than
being secondarily reversed. Unsurprisingly, only a single addi-
tional step is required to place Anchisaurus at the base of Sau-
ropodiformes, whereas the results of the implied weighting
analysis places Anchisaurus amongst Sauropodiformes when K
is larger than 2 (data not shown). Clearly, a more detailed
exploration of the effect of homoplasy on current sauropodo-
morph matrices is warranted, especially with respect to the con-
founding mosaic of characters evident in Anchisaurus. The
uncertainty surrounding the morphological relationships of
Anchisaurus, as well as its newly labile position, count among
the reasons why McPhee et al. (2014) opted instead for the use
of Sauropodiformes (the most inclusive clade containing Salta-
saurus but not Massospondylus) as a stem-based alternative to
‘Anchisauria.’
Leonerasaurus is retrieved in a relatively derived position near
the base of Sauropoda in our analysis, a surprising result given
the removal of the similarly gracile Anchisaurus from the pecti-
nate, large-bodied grouping of Sauropodiformes. This result is
potentially explicable via the large amounts of missing informa-
tion within this OTU, but a confluence of characters approaching
the ‘sauropodan’ condition (e.g., procumbent dentary teeth; four
sacral vertebrae; non-sinuous deltopectoral crest; preacetabular
process of the ilium exceeding the anterior margin of the pubic
peduncle; proximal surface of the first metatarsal subequal in
size to the second; straight lateral margin of the proximal surface
of the second metatarsal) nonetheless support a more derived
position. Although the possibility remains that the gracile mor-
phology of Leonerasurus is exaggerated by its subadult age (see
Pol et al., 2011), the basal sauropod Vulcanodon also bears an
elongate pes (Cooper, 1984), an observation consistent with our
hypothesized taxonomic position for Leonerasaurus (see also
Otero and Pol, 2013).
The relatively small number of coding alterations and new
characters added in this study has resulted in some modest
yet interesting changes to previous hypotheses of basal sauro-
podomorph phylogenetics. Further taxonomic work refining
taxon hypodigms and systemic work defining new cladistic
characters is necessary to test the stability of these phyloge-
netic results.
E.entaxonis and the Interrelationships of South African Basal
Sauropodomorpha
Sauropodomorph diversity within the Elliot Formation of
South Africa is strongly partitioned between the lower and upper
sections of the formation. Although the upper Elliot (Early
Jurassic) is showing signs of containing more taxonomic diversity
than just the abundant and well-known Massospondylus spp.
(Yates et al., 2010, 2011), the majority of sauropodomorph taxo-
nomic variation is exclusive to the lower Elliot (Late Triassic).
However, the understanding and taxonomic organization of this
variation has proven persistently elusive (e.g., Van Heerden,
1979). Besides Eucnemesaurus, we regard two other lower Elliot
sauropodomorph taxa as being of near-certain validity: Blikana-
saurus cromptoni, based on its distinctive ‘dwarfed’ morphology
(Galton and Van Heerden, 1998), and Antetonitrus ingenipes,
based on its clearly unique collection of relatively derived apo-
morphies (Yates and Kitching, 2003; McPhee et al., 2014). How-
ever, serious questions remain regarding the validity of the two
remaining lower Elliot taxa: Melanorosaurus readi and Plateo-
sauravus cullingworthi. The new information presented by E.
entaxonis thus has bearing on the stratigraphic, taxonomic, and
adaptive significance of both of the above genera—especially in
light of the outwardly progressive shift from primitive to rela-
tively derived forms observed with the LEF.
Plateosauravus cullingworthi—ostensibly the most basal of all
the Elliot taxa (see e.g., Yates, 2007a, 2007b, 2010)—has experi-
enced a taxonomic past almost as confused as that of Eucneme-
saurus (see Haughton, 1924; Huene, 1932; Van Heerden, 1979;
Yates, 2003a). Known primarily from a large assemblage of
bones recovered in 1918 near Hershel, Eastern Cape, the ‘type’
series (SAM-PK-3341–3356, 3602–3603, 3607–3609) is composed
of three or more moderately sized individuals as well as a consid-
erably more massive animal represented by a few dorsal/sacral
vertebrae and an almost complete pair of ischia. This latter,
larger material (SAM-PK-3607-3609) was originally selected as
the type material for Euskelosaurus africanus by Haughton
(1924), and doubt remains as to the conspecifity of this material
to the rest of the Plateosauravus type material. Furthermore,
although clearly plesiomorphic for Sauropodomorpha in a num-
ber of respects (see below), Plateosauravus nonetheless displays
the contradictory attributes of humeri that are elongate relative
McPhee et al.—A second species of Eucnemesaurus (e980504-20)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

to other bones in the assemblage and comparatively high poste-
rior dorsal neural spines—both derived features within Sauropo-
domorpha. For the time being, it is difficult to say whether this is
explicable via homoplastic convergence or evidence of a multi-
taxic assemblage.
E. entaxonis shares with Plateosauravus (to the exclusion of
other lower Elliot taxa) a number of characters consistent with
the non-sauropodan condition. These include (1) dorsal verte-
brae with subcircular neural canals; (2) an anteriorly positioned
lesser trochanter; (3) a fourth trochanter located mainly in the
proximal half of the shaft; and (4) a rectangular distal end of the
tibia with an acute anteromedial corner. Although most of these
features clearly distinguish Plateosauravus from the compara-
tively derived morphology of Melanorosaurus (with the excep-
tion of the elongate humeri and high posterior dorsal neural
arches), the exclusive autapomorphies distinguishing the former
from both Eucnemesaurus and Melanorosaurus are subtle: a
‘square’-shaped postacetabular process and the excavated
embayment of the lateral surface of the tibial posterior descend-
ing process (Yates, 2003a, 2007a). Unfortunately, the forelimb in
Eucnemesaurus remains unknown and the significance that such
information could bring to bear on the putatively elongate fore-
limb of Plateosauravus remains enigmatic.
Melanorosaurus is currently interpreted as occupying a rela-
tively close phylogenetic position to Sauropoda (e.g., Yates,
2007a, 2007b; Pol et al., 2011). Although known from a large col-
lection of syntypic and referred material (Haughton, 1924; Gal-
ton et al., 2005; Yates, 2007b; Bonnan and Yates, 2007), thus far
only the skull of a referred complete skeleton (NM QR3314;
Yates, 2007b) has received a formal description and diagnosis
with reference to both autapomorphies and unique character
constellations. Further uncertainties relate to the cohesiveness of
the syntype material itself, which includes bones collected from
at least two distinct—albeit neighboring—localities as well as
other disassociated material, possibly cataloged under SAM-PK-
3449 subsequent to initial collection (Haughton, 1924; Galton
et al., 2005; B.W.M., pers. observ.). Consequently, the distin-
guishing characters of the postcrania of Melanorosaurus are lim-
ited to a vaguely communicated unique suite of derived features
(i.e., dorsal neural spines that are approximately 1.5 times as tall
as they are long; four to five sacral vertebrae; sagittal furrow on
the ventral surface of the anterior caudal centra; distal migration
of the major femoral trochanters; femoral shaft elliptical in
cross-section with a reduced sigmoidal curvature; proximal
breadth of mtI roughly 0.6 times its proximodistal length), of
which at least two (anteroposterior constriction of the femoral
shaft; proportions of the first metatarsal) now appear present in
E. entaxonis (see Yates [2003a, 2007b, 2008] for a further com-
mentary on the distinguishing postcranial features of this taxon).
Yates (2007b) suggested the presence of a pitted excavation
(D‘non-articulating gap’) within the first primordial sacral rib as
a possible postcranial autapomorphy of Melanorosaurus (based
primarily on the morphology of the two major sets of referred
material: NM QR1551 and 3314). This observation is interesting
insofar as it means that the two sacral vertebrae anterior to the
pit-possessing element in NM QR1551 are in effect both dorso-
sacrals, contra Galton et al. (2005) who identified them as the
dorsosacral and first-primordial sacral vertebrae. This observa-
tion, along with the possibility that Melanorosaurus may have
also been in possession of a caudosacral (i.e., at least five sacral
vertebrae in total), suggests either the autapomorphic acquisition
of upwards of two additional sacral vertebrae in Melanorosaurus,
or character continuity shared exclusively with Eusauropoda
(Upchurch et al., 2004). However, a caudosacral is unambigu-
ously observable only in NM QR3314, whereas this same individ-
ual appears to lack the additional dorsosacral described for NM
QR1551 (pers. observ. B.W.M.), rendering the homology of the
non-articulating gap of Yates (2007b) contentious. Nonetheless,
although character conflict within the Melanorosaurus hypodigm
is not restricted to the sacral elements alone (the proportions of
the first metatarsals of NM QR1551 and 3314 are wildly diver-
gent; see above), and future care should be taken when regarding
referred material of Melanorosaurus as a single operational taxo-
nomic unit, the presence of a similar excavated pit in the first pri-
mordial sacral of E. entaxonis appears to corroborate Yates’
diagnosis of the same element in Melanorosaurus (NM
QR1551). The presence of this feature in both Eucnemesaurus
and referred specimens of Melanorosaurus suggests either a
wider distribution of this character than previously appreciated
or a closer taxic relationship between these two taxa than
implied by the current consensus tree.
These caveats aside, it is evident from the morphologies out-
lined above that Eucnemesaurus entaxonis appears to represent
something of an intermediate form between the relatively plesio-
morphic Plateosauravus and the relatively derived Melanorosau-
rus. However, although these morphologies could be taken as
evidence of either an anagenetic series, local adaptation, or pop-
ulation-level variation (or a combination thereof), testing these
various scenarios is hampered by the continued lack of resolu-
tion in regards to the temporal duration of the lower Elliot (and
of the Elliot Formation generally), as well as the poorly recorded
stratigraphic provenance of the majority of the sauropodomorph
specimens collected over the past century. Although Yates
(2008) has suggested that the lower Elliot forms a homogeneous
sedimentary unit in which the same large-bodied forms (i.e., Bli-
kanasaurus) are found throughout strata representing a short
depositional sequence—in which case the hypothesis of anagene-
sis would be convincingly falsified—the evidence cited for this is
slight (see Yates, 2008). Because E. entaxonis was discovered
just above the Molteno-Elliot contact, and as new, preliminary
data tentatively suggest that most of the more derived, sauropo-
diform taxa are found higher up in the sequence, it remains a
plausible possibility that E. entaxonis represents an ancestral
population of basal sauropodomorphs to more derived forms
such as Melanorosaurus. Nonetheless, the biostratigraphic and
temporal delineation of the Elliot Formation remains in its early
stages, with much more work (and fossil material) required yet if
we are to begin elucidating the true population dynamics of
lower Elliot Sauropodomorpha.
Functional Morphology of E. entaxonis
The most striking anatomical features observable within E.
entaxonis are the surprisingly robust foot architecture and the
presence of a deep brevis fossa on the ventral surface of iliac
postacetabular process. The stout metatarsus evinces the rela-
tively early occurrence (at least phylogenetically) of a robust,
subentaxonic pes amongst Late Triassic basal sauropodomorphs
and suggests an early experiment in a slower, subgraviportal
form of locomotion. Given the incipient adoption of a form of
pedal architecture later emblematic of the obligatorily quadrupe-
dal sauropods, it is interesting that the brevis fossa should be as
deeply developed as it is in E. entaxonis. Although Saturnalia,
Thecodontosaurus, and Efraasia retain relatively deep brevis
fossae (Benton et al., 2000; Yates, 2003b, 2003c), the tendency
early within Sauropodomorpha—especially in the large-bodied
forms—is to significantly reduce the extent of the brevis fossa,
with most taxa displaying at most a shallow embayment on the
ventral or ventrolateral surface of the postacetabular process
(e.g., Adeopapposaurus [PVSJ569]; Mussaurus [Otero and Pol,
2013]; Plateosauravus [SAM-PK-3609]). A brevis fossa as deeply
excavated as that exhibited by E. entaxonis is therefore unknown
in a sauropodomorph of its relatively derived position (with the
possible exception of Riojasaurus [PVL 3808]) and may be indic-
ative of a specialized locomotor strategy unique to E. entaxonis
amongst the lower Elliot sauropodomorph assemblage.
McPhee et al.—A second species of Eucnemesaurus (e980504-21)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

Recently, McPhee et al. (2014; see also Mallison, 2010a,
2010b) have suggested that a hypertrophied M. caudofemoralis
brevis complex may be related to the adducting forces required
to steady the feet beneath the body of a large, wide-gaited biped.
In this context, the brevis fossa of E. entaxonis may relate to a
specialized form of obligate/habitual bipedality within an early
radiation of broad-footed, massopodan sauropodomorphs. This
interpretation, if corroborated, would provide compelling evi-
dence of divergent locomotor strategies between E. entaxonis
and the putatively quadrupedal Melanorosaurus (Yates, 2007b;
Rauhut et al., 2011). However, it should be noted that the dis-
tinctly oblique passage of the fourth trochanter towards the
medial margin of the femur in Eucnemesaurus is possibly repre-
sentative of the ‘transitional’ condition whereby the fourth tro-
chanter becomes located entirely on the medial margin of the
femoral shaft in sauropodomorphs from Melanorosaurus
onwards. Unfortunately, because the forelimb of Eucnemesaurus
remains unknown, the complete skeletal anatomy and function
of the locomotor apparatus of the genus can only be guessed at.
Clearly, a great deal more well-preserved, associated fossil mate-
rial is required in order to more fully elucidate the functional,
evolutionary, and taxonomic constraints at work within the sau-
ropodomorphs both within the lower Elliot and beyond.
Biogeographic Implications of E. entaxonis
Although it is axiomatic that southern Pangean landmasses as
closely allied as Argentina and South Africa should have experi-
enced a large degree of faunal interexchange in the Late Triassic,
that this relationship should be embodied in at least two sister-
taxon relationships (Lessemsaurus CAntetonitrus and the
‘Riojasauridae’) appears to provide strong support for a pan-
Gondwanan fauna that, at least in terms of its observable range,
extended from the Free State-Lesotho border through roughly
2700 km to the La Rioja/San Juan provinces of Argentina. This
relationship is all the more remarkable given that region-specific
monophyly is rare even in areas that preserve a rich localized
sauropodomorph assemblage (e.g., China, South Africa). How-
ever, although support for the Lessemsaurus CAntetonitrus
clade appears fairly robust (e.g., McPhee et al., 2014; current
analysis), it is possible that the Riojasauridae may yet prove chi-
merical (see above). Therefore, any additional information
regarding the Riojasaurus hypodigm will be of considerable bio-
geographical value in allowing us to test further whether this
assemblage is represented by multiple taxa that expand this cos-
mopolitan, basal-massopodan fauna, or dissolve it entirely.
The faunal closeness of South America and southern Africa is
also hinted at within the Massospondylidae, which contains both
the Los Colorados genus Coloradisaurus and the upper Elliot
genus Massospondylus. The Late Triassic age generally ascribed
to Coloradisaurus (along with all Los Colorados fauna e.g.,
Bonaparte, 1971, 1973; Arcucci et al., 2004; Apaldetti et al.,
2013), in association with the unexpectedly robust pes of
E. entaxonis, highlights a previously overlooked and intriguing
inconsistency: although the complete metatarsus of Plateosaura-
vus remains unknown, there has as yet been no material recov-
ered from within the lower sections of the Elliot Formation that
is of comparable gracility to the wealth of foot remains known
for the Early Jurassic taxon Massospondylus. This suggests that
the elongate pes morphology of Massospondylus was either
inherited from a known lower Elliot taxon that subsequently
adopted a more gracile bauplan following the end-Triassic
extinction event, or—more likely—was retained from an ances-
tor that was not present during deposition of the lower Elliot
Formation. Although the possibility remains that the massospon-
dylid ‘proto-population’ exists undiscovered within the lower
Elliot, over a century’s worth of fossil-prospecting has yet to
recover a convincing candidate. If this discrepancy is therefore
not explicable due to sampling bias, an external origin becomes
the most likely explanation.
Although the Massospondylidae achieved a cosmopolitan dis-
tribution during the Early Jurassic—with forms known from
China (Lufengosaurus; Barrett et al., 2005), Antarctica (Glaciali-
saurus; Smith and Pol, 2007), Argentina (Adeopapposaurus and
Leyesaurus; Mart
ınez, 2009; Apaldetti et al., 2011), South Africa
(Massospondylus; Kitching and Raath, 1984), and possibly India
(Pradhania; Novas et al., 2011) and North America (Sarahsaurus
cf. Row et al., 2010)—only Coloradisaurus brevis hails from
rocks confidently datable to the Late Triassic. In further treat-
ments on the subject of massospondylid origins, it may therefore
prove useful to look to South America—in the capacity of a null
hypothesis—as the area of origin of the Massospondylidae.
Unfortunately, the paucity of similarly aged Late Triassic rocks
throughout the rest of Gondwana will undoubtedly render such
hypotheses difficult to substantiate.
CONCLUSION
The above work has further established Eucnemesaurus as a
valid genus that, although still mysterious with regard to substan-
tial areas of its anatomy, can now be shown to be represented by
two species—E. fortis and E. entaxonis. Additionally, although
the new anatomical data provided by this new specimen cur-
rently support the idea of a monophyletic radiation of
‘riojasaurids’ at the base of Massopoda, the possibility remains
that future phylogenetic analyses might resolve other relation-
ships, with the Riojasaurus hypodigm certainly warranting closer
scrutiny.
Serious questions also remain regarding the validity and
interrelationships of other LEF taxa, uncertainties further com-
plicated by the intriguing mosaic of primitive and derived fea-
tures present in E. entaxonis. Nonetheless, the question of
whether the derived characters within this suite are better
explained as synapomorphies supporting a closer relationship to
Sauropodiformes, or homoplasies exclusive to E. entaxonis,
remains clouded by our continued poor understanding of the
complete anatomy of Eucnemesaurus spp., as well as the contra-
dictory assemblage of characters within the prevailing Melanor-
osaurus hypodigm.
Clearly, a more robust understanding of the character com-
plexes of LEF Sauropodomorpha, set against a framework of
improved stratigraphic and temporal control, is required in
order to more fully elucidate the durations and variability of
lineages during this important period of sauropodomorph
evolution.
ACKNOWLEDGMENTS
We would like to thank S. Kaal, B. Zipfel, J. Botha-Brink, and
H. Fourie for access to specimens in their care. We are also
grateful for the hospitality of the proprietor of Cannon Rock
Farm, P. Prinsloo, for warmly welcoming us onto his property.
Gratitude is also expressed to M. Ezcurra and C. Apaldetti, for
comprehensive reviews that greatly improved the strength of the
manuscript. Funding for B.W.M. was supplied by an NRF Afri-
can Origins bursary to Bruce Rubidge and a DST/NRF Centre
of Excellence in Palaeosciences postgraduate bursary.
LITERATURE CITED
Allain, R., and N. Aquesbi. 2008. Anatomy and phylogenetic relation-
ships of Tazoudasaurus naimi (Dinosauria, Sauropoda) from the
late Early Jurassic of Morocco. Geodiversitas 30:345–424.
Anderson, J. M., H. M. Anderson, and A. R. Cruickshank. 1998. Late
Triassic ecosystems of the Molteno/lower Elliot biome of southern
Africa. Palaeontology 41:387–412.
McPhee et al.—A second species of Eucnemesaurus (e980504-22)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

Apaldetti, C., D. Pol, and A. Yates. 2013. The postcranial anatomy of
Coloradisaurus brevis (Dinosauria: Sauropodomorpha) from the
late Triassic of Argentina and its phylogenetic implications. Palae-
ontology 56:277–301.
Apaldetti, C., R. N. Mart
ınez, O. A. Alcober, and D. Pol. 2011. A new
basal sauropodomorph (Dinosauria: Saurischia) from Quebrada del
Barro Formation (Marayes-El Carrizal Basin), Northwestern
Argentina. PLoS ONE 6:e26964. doi: 10.1371/journal.pone.02696.
Arcucci, A. B., C. A. Marsicano, and A. T. Caselli. 2004. Tetrapod associ-
ation and palaeoenvironment of the Los Colorados Formation
(Argentina): a significant sample from Western Gondwana at the
end of the Triassic. Geobios 37:557–568.
Barrett, P. M., P. Upchurch, and W. Xiao-Lin. 2005. Cranial osteology of
Lufengosaurus huenei Young (Dinosauria: Prosauropoda) from the
lower Jurassic of Yunnan, People’s Republic of China. Journal of
Vertebrate Paleontology 25:806–822.
Benton, M. J., L. Juul, G. W. Storrs, and P. M. Galton. 2000. Anatomy
and systematics of the prosauropod dinosaur Thecodontosaurus
antiquus from the upper Triassic of southwest England. Journal of
Vertebrate Paleontology 20:77–108.
Bonaparte, J. F. 1971. Los tetrapodos del sector superior de la formacion
los colorados, La Rioja, Argentina (Triasico Superior) Opera Lil-
loana 22:1–183.
Bonaparte, J. F. 1973. Edades Reptil para el Tri
asico de Argentina y Bra-
zil. Actas V Congreso Geol
ogico Argentino 3:93–129.
Bonaparte, J. F., and J. A. Pumares. 1995. Notas sobre el primer craneo
de Riojasaurus incertus (Dinosauria, Prosauropoda, Melanorosauri-
dae) del Tri
asico superior de La Rioja, Argentina. Ameghiniana
32:341–349.
Bonnan, M. F., and A. M. Yates. 2007. A new description of the forelimb
of the basal sauropodomorph Melanorosaurus: implications for the
evolution of pronation, manus shape and quadrupedalism in sauro-
pod dinosaurs. Special Papers in Palaeontology 77:157–168.
Bordy, E. M., P. J. Hancox, and B. S. Rubidge. 2004. Basin development
during the deposition of the Elliot Formation (Late Triassic–Early
Jurassic), Karoo Supergroup, South Africa. South African Journal
of Geology 107:397–412.
Bordy, E. M., P. J. Hancox, and B. S. Rubidge. 2005. The contact of the
Molteno and Elliot formations through the main Karoo Basin,
South Africa: a second-order sequence boundary. South African
Journal of Geology 108:351–364.
Bordy, E. M., R. Prevec, and C. Makhwelo. 2006. Late Triassic (Norian)
palaeoecosystem of the lowermost Elliot Formation (Salpeterberg,
Eastern Cape, South Africa); pp. 169–170 in W. G. Parker, S. R.
Ash, and R. B. Irmis (eds.), A Century of Research at Petrified For-
est National Park: Geology and Paleontology. Museum of Northern
Arizona Bulletin 62.
Carrano, M. T. 2005. The evolution of sauropod locomotion; pp. 229–250
in K. A. Curry Rogers, and J. A. Wilson (eds.), The Sauropods:
Evolution and Paleobiology. University of California Press, Berke-
ley, California.
Carrano, M. T., and J. R. Hutchinson. 2002. Pelvic and hind limb muscu-
lature of Tyrannosaurus rex (Dinosauria: Theropoda). Journal of
Morphology 253:207–228.
Catuneanu, O., J. P. Hancox, and B. S. Rubidge. 1998. Reciprocal flex-
ural behaviour and contrasting stratigraphies: a new basin develop-
ment model for the Karoo retroarc foreland system, South Africa.
Basin Research 10:417–439.
Cooper, M. 1980. The first record of the prosauropod dinosaur Euskelo-
saurus from Zimbabwe. Arnoldia Zimbabwe 9:1–17.
Cooper, M. R. 1981. The prosauropod dinosaur Massospondylus carina-
tus Owen from Zimbabwe: its biology, mode of life and phyloge-
netic significance. Occasional Papers, National Museums and
Monuments of Rhodesia, Series B 6:689–840.
Cooper, M. R. 1984. A reassessment of Vulcanodon karibaensis Raath
(Dinosauria: Saurischia) and the origin of the Sauropoda. Palaeon-
tologia africana 25:203–231.
Ezcurra, M. D. 2010. A new early dinosaur (Saurischia: Sauropodomor-
pha) from the Late Triassic of Argentina: a reassessment of dino-
saur origin and phylogeny. Journal of Systematic Palaeontology
8:371–425.
Galton, P. M. 1976. Prosauropod dinosaurs (Reptilia: Saurischia) of
North America. Postilla 169:1–98.
Galton, P. M. 1985. The poposaurid thecodontian Teratosaurus suevi-
cus v. Meyer, plus referred specimens mostly based on
prosauropod dinosaurs, from the Middle Stubensandstein
(Upper Triassic) of Nordw€
urttemberg. Stuttgarter Beitr€
age zur
Naturkunde (B) 116:1–29.
Galton, P. M. 1999. Sex, sacra and Sellosaurus gracilis (Saurischia, Sauro-
podomorpha, Upper Triassic, Germany)—or why the character
“two sacral vertebrae” is plesiomorphic for Dinosauria. Neues Jahr-
buch f€
ur Geologie und Pal€
aontologie Abhandlungen 213:19–56.
Galton, P. M., and J. van Heerden. 1998. Anatomy of the prosauropod
dinosaur Blikanasaurus cromptoni (Upper Triassic, South Africa),
with notes on the other tetrapods from the lower Elliot Formation.
Pal€
aontologische Zeitschrift 72:163–177.
Galton, P. M., J. van Heerden, and A. M. Yates. 2005. Postcranial anat-
omy of referred specimens of the sauropodomorph dinosaur Mela-
norosaurus from the Upper Triassic of South Africa; pp. 1–37 in V.
Tidwell and K. Carpenter (eds.), Thunder-Lizards: The Sauropodo-
morph Dinosaurs. Indiana University Press, Bloomington, Indiana.
Goloboff, P. A., J. S. Farris, and K. C. Nixon. 2008. TNT, a free program
for phylogenetic analysis. Cladistics 24:774–786.
Hatcher, J. B. 1903. Osteology of Haplocanthosaurus, with description of
a new species and remarks on the probable habits of the Sauropoda
and the age and origin of the Atlantosaurus beds: additional remarks
on Diplodocus. Memoirs of the Carnegie Museum 2:1–75.
Haughton, S. H. 1924. The fauna and stratigraphy of the Stormberg
Series. Annals of the South African Museum 12:323–497.
Huene, F. v. 1906. Ueber die Dinosaurier der aussereurop€
aischen Trias.
Geologische und Pal€
aontologische Abhandlungen 8:99–156.
Huene, F. v. 1926. Vollst€
andige Osteologie eines Plateosauriden aus dem
schw€
abischen Keuper. Geologische und Pal€
aontologische Abhand-
lungen 15:129–179.
Huene, F. v. 1932. Die fossile Reptil-Ordnung Saurischia, ihre
Entwicklung und Geschichte. Monographien zur Geologie und
Pal€
aontologie 4:1–361.
Irmis, R. B. 2010. Evaluating hypotheses for the early diversification of
dinosaurs. Earth and Environmental Science Transactions of the
Royal Society of Edinburgh 101:397–426.
Kitching, J. W., and M. A. Raath. 1984. Fossils from the Elliot and Cla-
rens formations (Karoo sequence) of the northeastern Cape,
Orange Free State and Lesotho, and a suggested biozonation based
on tetrapods. Palaeontologia africana 25:111–125.
Langer, M. C. 2003. The pelvic and hind limb anatomy of the stem-sauro-
podomorph Saturnalia tupiniquim (Late Triassic, Brazil). PaleoBios
23:1–30.
Langer, M. C., and M. J. Benton. 2006. Early dinosaurs: a phylogenetic
study. Journal of Systematic Palaeontology 4:309–358.
Lucas, S. G., and J. P. Hancox. 2001. Tetrapod-based correlation of the
nonmarine Upper Triassic of southern Africa. Albertiana 25:5–9.
Mallison, H. 2010a. The digital Plateosaurus I: body mass, mass distribu-
tion, and posture assessed using CAD and CAE on a digitally
mounted complete skeleton. Palaeontologia Electronica 13:1–26.
Available at http://palaeoelectronica.org/2010_2/198/index.html
Mallison, H. 2010b. The digital Plateosaurus II: an assessment of the
range of motion of the limbs and vertebral column and of previous
reconstructions using a digital skeletal mount. Acta Palaeontologica
Polonica 55:433–458.
Mart
ınez, R. N. 2009. Adeopapposaurus mognai, gen. et sp. nov. (Dino-
sauria: Sauropodomorpha), with comments on adaptations of basal
Sauropodomorpha. Journal of Vertebrate Paleontology 29:142–164.
Mart
ınez, R. N., and O. A. Alcober. 2009. A basal sauropodomorph
(Dinosauria: Saurischia) from the Ischigualasto Formation (Trias-
sic, Carnian) and the early evolution of Sauropodomorpha. PLoS
ONE 4:e4397. doi: 10.1371/journal.pone.0004397.
Mart
ınez, R. N., C. Apaldetti, and D. Abelin. 2012. Basal sauropodo-
morphs from the Ischigualasto Formation. Journal of Vertebrate
Paleontology 32:51–69.
McPhee, B. W., A. M. Yates, J. N. Choiniere, and F. Abdala. 2014. The
complete anatomy and phylogenetic relationships of Antetonitrus
ingenipes (Sauropodiformes, Dinosauria): implications for the ori-
gins of Sauropoda. Zoological Journal of the Linnean Society
171:151–205.
Novas, F. E. 1994. New information on the systematics and postcranial
skeleton of Herrerasaurus ischigualastensis (Theropoda: Herrera-
sauridae) from the Ischigualasto Formation (Upper Triassic) of
Argentina. Journal of Vertebrate Paleontology 13:400–423.
Novas, F. E. 1996. Dinosaur monophyly. Journal of Vertebrate Paleon-
tology 16:723–741.
McPhee et al.—A second species of Eucnemesaurus (e980504-23)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015

Novas, F. E., M. D. Ezcurra, S. Chatterjee, and T. S. Kutty. 2011. New
dinosaur species from the Upper Triassic Upper Maleri and Lower
Dharmaram formations of Central India. Earth and Environmental
Science Transactions of the Royal Society of Edinburgh 101:333–
349.
Olsen, P. E., and P. M. Galton. 1984. A review of the reptile and amphib-
ian assemblages from the Stormberg of southern Africa, with special
emphasis on the footprints and the age of the Stormberg. Palaeonto-
logia africana 25:87–110.
Otero, A., and D. Pol. 2013. Postcranial anatomy and phylogenetic rela-
tionships of Mussaurus patagonicus (Dinosauria, Sauropodomor-
pha). Journal of Vertebrate Paleontology 33:1138–1168.
Otero, A., E. Krupandan, D. Pol, A. Chinsamy, and J. Choiniere. 2015.
A new basal sauropodiform from South Africa and the phylogenetic
relationships of basal sauropodomorphs. Zoological Journal of the
Linnean Society 174:589–634.
Owen, R. 1842. Report on British fossil reptiles. Part II; pp. 60–204 in
Report of the Eleventh Meeting of the British Association for the
Advancement of Science; Held at Plymouth in July 1841. Richard
and John E. Taylor, London.
Pol, D., and J. E. Powell. 2007. New information on Lessemsaurus sauro-
poides (Dinosauria: Sauropodomorpha) from the Upper Triassic of
Argentina. Special Papers in Palaeontology 77:223–243.
Pol, D., A. Garrido, and I. A. Cerda. 2011. A new sauropodomorph dino-
saur from the Early Jurassic of Patagonia and the origin and evolu-
tion of the sauropod-type sacrum. PLoS ONE 6:e14572. doi:
10.1371/journal.pone.0014572
Rauhut, O. W. M., R. Fechner, K. Remes, and K. Reis. 2011. How to get
big in the Mesozoic: the evolution of the sauropodomorph body
plan; pp. 119–149 in N. Klein, K. Remes, C. T. Gee, and P. M.
Sander (eds.), Biology of the Sauropod Dinosaurs: Understanding
the Life of Giants. Indiana University Press, Bloomfield, Indiana.
Rohm and Haas Company. 2007. Paraloid B-72 Solid Grade Thermoplas-
tic Resin. Material Safety Data Sheet. Rohm and Haas Company,
Philadelphia, Pennsylvania.
Row, T. B., H.-D. Sues, and R. R. Reisz. 2010. Dispersal and diversity in
the earliest North American sauropodomorph dinosaurs, with a
description of a new taxon. Proceedings of the Royal Society B: Bio-
logical Sciences 278:1044–1053.
Seeley, H. G. 1888. On the classification of the fossil animals commonly
named Dinosauria. Proceedings of the Royal Society of London
43:165–171.
Smith, N. D., and D. Pol. 2007. Anatomy of a basal sauropodomorph
from the Early Jurassic Hanson Formation of Antarctica. Acta
Palaeontologica Polonica 52:657–674.
Van Heerden, J. 1979. The morphology and taxonomy of Euskelosaurus
(Reptilia: Saurischia: Late Triassic) from South Africa. Navorsinge
van die Nasionale Museum 4:21–84.
Van Hoepen, E., C. N. 1920. Contributions to the knowledge of the rep-
tiles of the Karoo Formation. 6. Further dinosaurian material in the
Transvaal Museum. Annals of the Transvaal Museum 7:93–140.
Wilson, J. A. 1999. A nomenclature for vertebral laminae in sauropods
and other saurischian dinosaurs. Journal of Vertebrate Paleontol-
ogy 19:639–653.
Wilson, J. A. 2011. Anatomical terminology for the sacrum of sauropod
dinosaurs. Contributions from the Museum of Paleontology, Uni-
versity of Michigan 32:59–69.
Yates, A. M. 2003a. A definite prosauropod dinosaur from the lower
Elliot Formation (Norian: Upper Triassic) of South Africa. Palae-
ontologia africana 39:63–68.
Yates, A. M. 2003b. A new species of the primitive dinosaur Thecodonto-
saurus (Saurischia: Sauropodomorpha) and its implications for the
systematics of early dinosaurs. Journal of Systematic Palaeontology
1:1–42.
Yates, A. M. 2003c. The species taxonomy of the sauropodomorph dino-
saurs from the L€
owenstein Formation (Norian, Late Triassic) of
Germany. Palaeontology 46:317–337.
Yates, A. M. 2004. Anchisaurus polyzelus (Hitchcock): the smallest
known sauropod dinosaur and the evolution of gigantism among
sauropodomorph dinosaurs. Postilla 230:1–58.
Yates, A. M. 2007a. Solving a dinosaurian puzzle: the identity of Aliwalia
rex Galton. Historical Biology 19:93–123.
Yates, A. M. 2007b. The first complete skull of the Triassic dinosaur Mel-
anorosaurs Haughton (Sauropodomorpha: Anchisauria). Special
Papers in Paleontology 77:9–55.
Yates, A. M. 2008. A second specimen of Blikanasaurus (Dinosauria:
Sauropoda) and the biostratigraphy of the lower Elliot Formation.
Palaeontologia africana 43:39–43.
Yates, A. M. 2010. A revision of the problematic sauropodomorph dino-
saurs from Manchester, Connecticut and the status of Anchisaurus
Marsh. Palaeontology 53:739–752.
Yates, A. M., and J. W. Kitching. 2003. The earliest known sauropod
dinosaur and the first steps towards sauropod locomotion. Proceed-
ings of the Royal Society of London. Series B: Biological Sciences
270:1753–1758.
Yates, A. M., M. F. Bonnan, and J. Neveling. 2011. A new basal sauropo-
domorph dinosaur from the Early Jurassic of South Africa. Journal
of Vertebrate Paleontology 31:610–625.
Yates, A. M., J. P. Hancox, and B. S. Rubidge. 2004. First record of
a sauropod dinosaur from the Upper Elliot Formation (Early
Jurassic) of South Africa. South African Journal of Science
100:504–506.
Yates, A. M., M. J. Wedel, and M. F. Bonnan. 2012. The early evolution
of postcranial skeletal pneumaticity in sauropodomorph dinosaurs.
Acta Palaeontologica Polonica 57:85–100.
Yates, A. M., M. F. Bonnan, J. Neveling, A. Chinsamy, and M. G. Black-
beard. 2010. A new transitional sauropodomorph dinosaur from the
Early Jurassic of South Africa and the evolution of sauropod feed-
ing and quadrupedalism. Proceedings of the Royal Society B: Bio-
logical Sciences 277:787–794.
Young, C.-C. 1941. A complete osteology of Lufengosaurus huenei
Young (gen. et sp. nov.) Palaeontologica Sinica, Series C 7:1–53.
Young, C.-C. 1942. Yunnanosaurus huangi Young (gen. et sp. nov.), a
new Prosauropoda from the Red Beds at Lufeng, Yunnan. Bulletin
of the Geological Society of China 22:63–104.
Zhang, Y. H., and Z. Yang. 1994. A New Complete Osteology of Pro-
sauropoda in Lufeng Basin, Yunnan, China: Jingshanosaurus.
Yunnan Publishing House of Science and Technology, Kunming,
China.
Submitted April 21, 2014; revisions received August 9, 2014; accepted
October 5, 2014.
Handling editor: Richard Butler.
Citation for this article: McPhee, B. W., J. N. Choiniere, A. M. Yates, and
P. A. Viglietti. 2015. A second species of Eucnemesaurus Van Hoepen,
1920 (Dinosauria, Sauropodomorpha): new information on the diversity
and evolution of the sauropodomorph fauna of South Africa’s lower
Elliot Formation (latest Triassic). Journal of Vertebrate Paleontology.
DOI: 10.1080/02724634.2015.980504.
McPhee et al.—A second species of Eucnemesaurus (e980504-24)
Downloaded by [Jonah Choiniere] at 21:15 07 August 2015