Content uploaded by Lida Xing
Author content
All content in this area was uploaded by Lida Xing on Mar 17, 2015
Content may be subject to copyright.
ARTICLE
A NEW SAUROPOD DINOSAUR FROM THE LATE JURASSIC OF CHINA
AND THE DIVERSITY, DISTRIBUTION, AND RELATIONSHIPS OF MAMENCHISAURIDS
LIDA XING,
1
TETSUTO MIYASHITA,*
,2
JIANPING ZHANG,
1
DAQING LI,
3
YONG YE,
4
TORU SEKIYA,
5
FENGPING WANG,
6
and PHILIP J. CURRIE
2
1
School of the Earth Sciences and Resources, China University of Geosciences, Beijing 100083, China, xinglida@gmail.com;
zhjping@cugb.edu.cn;
2
Department of Biological Sciences, University of Alberta, Edmonton, Alberta, Canada T6G 2E9, tetsuto@ualberta.ca;
philip.currie@ualberta.ca;
3
Geological Museum of Gansu Province, Lanzhou 730040, China, daqingligs@gmail.com;
4
Zigong Dinosaur Museum, 238, Dashanpu, Zigong 643013, Sichuan, China, yeyozdm@126.com;
5
Fukui Prefectural Dinosaur Museum, Katsuyama, Fukui, Japan 911-8601, t.sekiya.jiu@gmail.com;
6
Qijiang County Bureau of Land and Resources, Chongqing 401420, China, 455039424@qq.com
ABSTRACT—Qijianglong guokr, gen. et sp. nov., represents a mamenchisaurid eusauropod from the Late Jurassic of
southern China. The holotype consists of an incomplete skull, partly articulated axial skeleton, and fragmentary appendicular
skeleton. A well-preserved braincase and skull roof provide rare insights into the poorly known neurocranial anatomy of
mamenchisaurids and reveal a unique combination of characters such as an accessory tuber at the base of planar
basipterygoid process and parietal excluding frontal from the anterior margin of the supratemporal fenestra. The cervical
vertebrae have a distinct �nger-like process extending from the postzygapophyseal process beside a zygapophyseal contact.
Qijianglong is the �rst mamenchisaurid from the Late Jurassic of China that is de�nitively distinct from Mamenchisaurus,
indicating greater morphological and taxonomic diversity of the poorly represented Late Jurassic mamenchisaurids. The
occurrence of Qijianglong is consistent with a scenario in which mamenchisaurids formed an endemic sauropod fauna in
the Late Jurassic of Asia. Phylogenetically, Qijianglong represents a relatively plesiomorphic mamenchisaurid lineage. The
mamenchisaurids form an ancient clade of basal eusauropod dinosaurs that likely appeared in the Early Jurassic. A cladistic
analysis highlights the interrelationships of mamenchisaurids and suggests guidelines for mamenchisaurid taxonomic revision.
It may be desirable to restrict generic names to the type species in order to avoid confusion.
http://zoobank.org/urn:lsid:zoobank.org:pub:F93276CF-71FE-472E-9114-68294547C2A9
SUPPLEMENTAL DATA—Supplemental materials are available for this article for free at www.tandfonline.com/UJVP
INTRODUCTION
The sauropod fauna from the Late Jurassic of Asia has previ-
ously been considered distinct from the contemporaneous sauro-
pod faunas from other landmasses in two respects: (1) the
dominance of mamenchisaurids, which likely form a clade of
basal eusauropods; and (2) the absence of de�nitive diplodocoids
and titanosauriforms (Upchurch et al., 2004; Mannion et al.,
2011; Whitlock et al., 2011). Mamenchisaurids replaced a diverse
assemblage of basal eusauropods and possible macronarian neo-
sauropods in the Late Jurassic of Asia (Xing et al., 2013). The
dominance of titanosauriform neosauropods followed in the
Early Cretaceous (Whitlock et al., 2011). The faunal turnovers
amongst sauropods of different phylogenetic grades seemingly
correlate with the geographic isolation and reconnection of Asia
during the Late Jurassic (Russell, 1993; Upchurch and Mannion,
2009; Mannion et al., 2011; Whitlock et al., 2011). This correla-
tion suggests that the Late Jurassic sauropods from Asia repre-
sent an endemic fauna. All mamenchisaurids from this critical
time interval have been assigned to the single genus Mamenchi-
saurus. Thus, the Late Jurassic times appears to have been a bot-
tleneck in Asian sauropod diversity, with low morphological and
systematic diversities. Alternatively, the diversity may be under-
estimated. It is possible that the genus Mamenchisaurus merely
serves as a wastebasket taxon for large basal eusauropods from
the Late Jurassic of Asia. However, it has been dif�cult to evalu-
ate either of the hypotheses because of the lack of comparative
studies, and because of the lack of cranial materials in many of
the species.
This biostratigraphic and biogeographic scheme for Asian sau-
ropods faces a modest challenge from revised chronological cor-
relations of the Jurassic localities in the Sichuan Basin. The
upper Shaximiao succession (Shangshaximiao) was recently
resolved as upper Middle Jurassic in age (Bathonian–Callovian;
G. Li et al., 2010; K. Li et al., 2010a; Wang et al., 2010). This
revised chronology reassigns the majority of sauropod taxa that
occur in the upper successions of the Jurassic Sichuan Basin
from the lower Upper Jurassic to the upper Middle Jurassic
(Table 1). This revision leaves only two valid sauropod taxa as
de�nitively Late Jurassic taxa from China: Mamenchisaurus any-
uensis from the Suining Formation (He et al., 1996) and M. sino-
canadorum from the upper Shishugou Formation (Russell and
Zheng, 1993). The postcranial skeletons from the Xiangtang For-
mation are referred to M. constructus (Young, 1958), but this
assignment is inadequate in the light of multiple species of
Mamenchisaurus and the lack of description of diagnostic
*Corresponding author.
1
Journal of Vertebrate Paleontology e889701 (17 pages)
Óby the Society of Vertebrate Paleontology
DOI: 10.1080/02724634.2014.889701
Downloaded by [Kangwon National University - Samcheok Campus] at 15:51 19 February 2015
TABLE 1. Chronological distribution of sauropodomorphs from the Early Jurassic to Early Cretaceous of Asia.
Early Jurassic Middle Jurassic
Sauropodomorpha Sauropodomorpha
Chuxiongosaurus lufengensis lLU Yunnanosaurus youngi ZG
Jingshanosaurus xinwaensis lLU Eusauropoda
Lufengosaurus huenei lLU, ZZ Chuanjiesaurus anaensis CH
Xixiposaurus suni lLU Datousaurus bashanensis lSX
Yimenosaurus youngi FJ Hudiesaurus sinojapanorum*QG
Yunnanosaurus huangi lLU Nebulasaurus taito ZG
Yunnanosaurus robustus FJ/ZG Shunosaurus lii lSX
Sauropoda Mamenchisauridae
Chinshakiangosaurus chunghoensis FJ Eomamenchisaurus yuanmouensis ZG
Gongxianosaurus shibeiensis ZL Mamenchisaurus constructus uSX
Eusauropoda Mamenchisaurus fuxiensis lSX
Mamenchisauridae Mamenchisaurus hochuanensis uSX
Tonganosaurus hei YM Mamenchisaurus jingyanensis uSX
Mamenchisaurus youngi uSX
Omeisaurus jiaoi lSX
Omeisaurus junghsiensis lSX
Omeisaurus maoianus uSX
Omeisaurus tianfuensis lSX
Xinjiangtitan shanshanensis QG
?Yuanmousaurus jingyiensis
x
ZG
Neosauropoda
Ferganasaurus verzilini BB
Macronaria
Abrosaurus dongpoi lSX
Bellusaurus sui WC
Daanosaurus zhangi uSX
Late Jurassic Early Cretaceous
Sauropodomorpha Sauropodomorpha
Eusauropoda Neosauropoda
Mamenchisauridae Macronaria
Mamenchisaurus anyuensis uSU Titanosauriformes
Mamenchisaurus sinocanadorum uSS Chiayusaurus lacustris KZ
?Mamenchisaurus constructus
z
XT Daxiatitan binglingi NP
?Mamenchisaurus sp. PK Dongbeititan dongi YX
Qijianglong guokr, gen. et sp. nov. SU Erketu ellisoni BS
Euhelopus zdanskyi MY
Fukuititan nipponensis KD
Fusuisaurus zhaoi NP
Gobititan shenzhouensis DG
Huanghetitan liujiaxiaensis HK
Jiangshanosaurus lixianensis JH
Jiutaisaurus xidiensis QT
Liubangosaurus hei NP
Mongolosaurus haplodon OG
Phuwiangosaurus sirindhornae SK
Pukyongosaurus millenniumi HD
Qiaowanlong kangxii XM
Tangvayosaurus hoffeti
Yunmenglong ruyangensis HL
This table is a revision of that in Xing et al. (2013). The upper Shaximiao Formation is now regarded as the upper Middle Jurassic (Bathonian–Callo-
vian). Qijianglong (this paper), Xijiangtitan (Wu et al., 2013), and Yunmenlong (L€
u et al., 2013) are added. Each taxon is denoted with two upper case
letters that stand for the stratigraphic unit from which it derives. Where a single formation contains demonstrably distinct faunas between upper and
lower levels, a lower case letter indicates whether the taxon occurs in the upper (u) or lower (l) levels. This list does not include taxa that are consid-
ered not diagnosable in the current literature, such as Tienshanosaurus (Upchurch et al., 2004) and Yuanmousaurus (Xing et al., 2013). Abbreviations:
BB, Balabansai Formation; BS, Baynshiree Svita; CH, Chuanjie Formation; DG, Digou Formation; FJ, Fengjiahe Formation; GS,Gr

es Sup
erior For-
mation; HD, Hasandong Formation; HK, Hekou Group; HL, Haoling Formation; JH, Jinhua Formation; KD, Kitadani Formation; KZ, Kalazha For-
mation; LU, Lufeng Formation; MY, Mengyin Formation; NP, Napai Formation; OG, On Gong Formation; PK, Phu Kradung Formation; QG, Qigu
Formation; QT, Quantou Formation; SK, Sao Khua Formation; SU, Suining Formation; SS, Shishugou Formation; SX, Shaximiao Formation; WC,
Wucaiwan Formation; XM, Xinminpu Group; XT, Xiangtang Formation; YM, Yimen Formation; YX, Yixian Formation; ZG, Zhanghe Formation;
ZL, Ziliujing Formation; ZZ, Zhenzhuchong Formation.
*Chronological age is uncertain. The type and only specimen of Hudiesaurus likely comes from the upper Middle Jurassic Qigu Formation (Wings
et al., 2011, 2012).
x
Taxonomic status is uncertain. This taxon may represent an indeterminate mamenchisaurid (Xing et al., 2013).
z
Postcranial skeletons are referred to M. constructus (Young, 1958). Pending description of proper diagnostic characters, it remains uncertain whether
or not these specimens pertain to the taxon.
Taxonomic status and chronological age are uncertain; from the Late Jurassic/Early Cretaceous of Thailand (Phu Kradung Formation; Suteethorn
et al., 2013).
e889701-2JOURNAL OF VERTEBRATE PALEONTOLOGY
Downloaded by [Kangwon National University - Samcheok Campus] at 15:51 19 February 2015
characters. Hudiesaurus sinojapanorum was initially reported
from the Upper Jurassic Kalazha Formation of Xinjiang, but the
original material was likely collected from the upper Middle
Jurassic Qigu Formation (Wings et al., 2011, 2012). Despite its
problematic taxonomy and despite the coarsely sampled Late
Jurassic record, Mamenchisaurus remains as the only sauropod
genus in the Late Jurassic of China.
The majority of the Middle–Late Jurassic sauropods from
China occur in the Sichuan Basin, which makes this a primary
locality for a revision of mamenchisaurids. In contrast to the rich
vertebrate fauna from the Middle Jurassic Shaximiao Formation
in the basin, however, dinosaur fossils are rare in the overlying
Upper Jurassic Suining and Penglaizhen formations (Zhang and
Li, 2003). The sole valid sauropod taxon from these formations is
M. anyuensis. This species is represented by teeth and postcra-
nial skeletons of more than 10 individuals from the Suining For-
mation (He et al., 1996) and by another 70% complete
postcranial skeleton from the uppermost Suining Formation just
below the contact with the overlying Penglaizhen Formation
(Ouyang and Ye, 2002). Additional sauropod materials were col-
lected from the Longjiaya fossil site in Anyue County (Kan
et al., 2005), but the materials have not been described. The
youngest record of Mamenchisaurus from China is fragmentary
postcranial material tentatively assigned to M. anyuensis from
the Penglaizhen Formation (He et al., 1996).
As a lead to the elusive vertebrate fossils from the Suining
Formation, a local farmer (Cai Changming) discovered sauropod
vertebrae in his backyard in Heba Village, Beidu Township, in
the early 1990s. In 2006, construction workers working 500 m
from the original locality spotted a 0.7 m long neopterygian �sh
(Lepidotes). These discoveries prompted Qijiang District to
commission a survey of the area by the Fossil Research and
Development Center of the Third Geology and Mineral Resour-
ces Exploration Academy of Gansu Province. A new mamenchi-
saurid sauropod was discovered during this �eld work. This
paper presents a description of the sauropod, Qijianglong guokr,
gen. et sp. nov., based on morphological features unique among
mamenchisaurids, such as extensive pneumatization of the cervi-
cal vertebrae. Qijianglong improves the resolution of mamenchi-
saurid interrelationships, hints at an imminent taxonomic
revision for multiple mamenchisaurid taxa, and strengthens the
endemism scenario for the Late Jurassic Asian sauropod fauna.
Institutional Abbreviations—PMU, Paleontological Museum
of Uppsala University, Uppsala, Sweden; QJGPM, Qijiang Pet-
ri�ed Wood and Dinosaur Footprint National Geological Park
Museum, Chongqing, China; ZDM, Zigong Dinosaur Museum,
Zigong, Sichuan, China.
SYSTEMATIC PALEONTOLOGY
SAURISCHIA Seeley, 1887
SAUROPODOMORPHA Huene, 1932
SAUROPODA Marsh, 1878
EUSAUROPODA Upchurch, 1995
MAMENCHISAURIDAE Young and Chao, 1972
QIJIANGLONG, gen. nov.
Type and Only Known Species—Qijianglong guokr, sp. nov.
Etymology—Qijiang, after Qijiang District where the type
specimen was collected and is accessioned; ‘long,’ dragon in
Chinese.
Diagnosis—As for the type and only known species.
QIJIANGLONG GUOKR, sp. nov.
(Figs. 2–14)
Holotype—QJGPM 1001. Skull consisting of the skull roof,
braincase, right pterygoid, fragments of right antorbital elements
(lacrimal, maxilla, palatine, ectopterygoid), right postorbital, and
right quadrate; a complete cervical series; thoracic dorsal series;
distal caudal series; numerous fragments of neural arches;
numerous rib fragments; numerous hemal arch fragments; left
pubis; and a pedal phalanx.
Horizon—Suining Formation (Upper Jurassic). Seven forma-
tions of terrestrial deposits represent the Jurassic of the Sichuan
Basin (in ascending order): Zhenzhuchong, Ziliujing, Xintian-
gou, Xiashaximiao (lower Shaximiao), Shangshaximiao (upper
Shaximiao), Suining, and Penglaizhen formations (Peng et al.,
2005). The last two formations are calibrated to the Upper Juras-
sic part of the succession. The Suining Formation overlies the
lower Shangshaximiao Formation, with a single lithology predo-
minated by red and reddish-brown calcareous mudstone, mixed
with some off-white and gray-green quartz sandstone. Based on
the lithology and the characteristics of associated ostracods, the
stratigraphic age of the Suining Formation is de�nitively Upper
Jurassic (Gu and Li, 1997; Peng et al., 2005; G. Li et al., 2010;
Wang et al., 2010; Xie et al., 2010).
FIGURE 1. Geographic and taphonomic information on the holotype
of Qijianglong guokr (QJGPM 1001). A, location of the locality for
QJGPM 1001 (indicated by a silhouette of a sauropod) and Qijiang
District in China (inset map); B, site map for QJGPM 1001. The skull
elements were found in the block indicated by a circle at the end of the
cervical series.
JOURNAL OF VERTEBRATE PALEONTOLOGY e889701-3
Downloaded by [Kangwon National University - Samcheok Campus] at 15:51 19 February 2015
Locality—Beidu site (29002500N, 1063405500 E), Qijiang Petri-
�ed Wood and Dinosaur Footprint National Geological Park,
Qijiang District, Chongqing Municipality, China (Fig. 1A). The
outcrops within the park consist of the Shangshaximiao, Suining,
and Penglaizhen formations and the mid-Cretaceous Jiaguan
Formation. The fossil assemblage from the Upper Jurassic strata
in the park includes coniferopsid petri�ed wood, theropod teeth
(Wang Feng-ping, pers. comm., 2011), and the sauropod remains
described in this paper.
Etymology—guokr (gu-OH-ke-r), named in honor of Guokr
(science social network; ‘nutshell’ in Chinese) for their support
of paleontology in Qijiang.
Diagnosis—A non-neosauropod basal eusauropod with the
following unique combination of characters and autapomor-
phies: (1) semi-equal anteroposterior lengths of frontal and
parietal (also in Omeisaurus and Shunosaurus); (2) parietal
forming the entire anterior margin of supratemporal fenestra
(also in Atlasaurus and Omeisaurus); (3) absence of fronto-
parietal fenestra and presence of postparietal foramen (also
in Spinophorosaurus); (4) plate-like basipterygoid process ori-
ented anteroventrally with an accessory tuber paralleling
basal tuber (autapomorphy); (5) a �nger-like process lateral
to postzygapophyses in cervical vertebrae (autapomorphy);
(6) pneumatopores in spinodiapophyseal fossa in posterior
cervical vertebrae (autapomorphy); (7) anterior outline of
spinous process of mid-caudal vertebra indented posteriorly
for more than half a length of centrum (also in Mamenchi-
saurus); and (8) pubis anteriorly concave such that the distal
end points more anteriorly than ventrally (autapomorphy).
Numbers refer to each diagnostic character indicated by an
arrow in relevant �gures (Figs. 2, 5, 11, 12, 14, 15).
DESCRIPTION
Skull
The incomplete skull consists of six fractured portions, each
collected as a single block from the quarry (postorbital, ptery-
goid, quadrate, skull roof–facial unit, occipital plane, and basi-
cranium). The largest portion (Fig. 2) includes the right maxilla,
right lacrimal, right palatine, right ectopterygoid, prefrontals,
frontals, and parietals. Among these elements, the only visible
suture is between the frontals and parietals. The maxilla, pala-
tine, and ectopterygoid are only fragmentarily preserved and
cannot be compared with other sauropod skulls. These facial and
palatal elements and the lacrimal are set in a vertical plane per-
pendicular to the skull roof, which does not accurately reflect
their respective positions in life. A crack at the anterior end of
the skull roof through the prefrontal indicates that the preorbital
bar was at an angle about 5shallower than perpendicular to the
skull roof.
Frontals and Parietals—The frontals and parietals form a flat
skull roof that is as long anteroposteriorly as it is wide trans-
versely (Fig. 2). The anteroposterior length of the frontal is only
10% longer than the maximum anteroposterior length of the
parietal. The semi-equal proportions of length of the frontals
and parietals also exist in Omeisaurus tianfuensis and Shunosau-
rus (He et al., 1988; Chatterjee and Zheng, 2002), whereas the
frontals are longer anteroposteriorly than the parietals in other
basal sauropods such as Jobaria,Mamenchisaurus youngi, and
Spinophorosaurus as well as in many derived taxa (Sereno et al.,
1999; Ouyang and Ye, 2002; Knoll et al., 2012). In dorsal view,
the orbital margin is gently concave laterally. The frontals meet
the parietals along the suture that forms a shallow, posteriorly
pointed ‘V’ of approximately 150. The parietal extends antero-
laterally to form the entire anterior margin of the supratemporal
fenestra as in Atlasaurus and O. tianfuensis (He et al., 1988;
Monbaron et al., 1999). The postorbital participates in this mar-
gin in Jobaria,M. youngi, and Spinophorosaurus, but even in
these sauropods the frontal is excluded from the supratemporal
fenestra by both the parietal and postorbital (Sereno et al., 1999;
Ouyang and Ye, 2002; Remes et al., 2009; Knoll et al., 2012). In
Shunosaurus and Turiasaurus, the frontal participates in the mar-
gin of the fenestra (Chatterjee and Zheng, 2002; Royo-Torres
and Upchurch, 2012). The supratemporal fenestra is transversely
wider than anteroposteriorly long as in most sauropods, but the
proportions are reversed in M. youngi and Turiasaurus (Ouyang
and Ye, 2002; Royo-Torres and Upchurch, 2012). As is the case
for most sauropods, the supratemporal fenestra is not enclosed
within a fossa in Qijianglong.
Whereas a frontoparietal fenestra is absent, the postparietal
foramen demarcates the posterior end of the parietal at the mid-
line. The posterior wing of the parietal is oriented posterolater-
ally. In ventral view, both the frontals and parietals preserve
impressions of various parts of the brain. The dural depression is
longer anteroposteriorly than wide transversely and deeper pos-
teriorly toward the postparietal foramen. The depression is not
divided bilaterally. In comparison with the previously described
cranial endocasts of sauropods, this depression likely housed the
dural peak and a longitudinal venous sinus rather than the cere-
bral hemispheres. Anteriorly, the ventral surface of the frontal
has the impressions of the cerebral hemispheres and olfactory
tract. At the anterior end of the tract is a pair of small depres-
sions for the olfactory bulbs. These depressions are narrower
transversely but deeper than the olfactory tract and clearly sepa-
rated at the midline. The pro�le of an endocast reconstructed
from these impressions generally agrees with the three-
dimensionally reconstructed cranial endocasts of Ampelosaurus,
Apatosaurus,Brachiosaurus,Camarasaurus,Dicraeosaurus,
Diplodocus,Nigersaurus,Shunosaurus,Spinophorosaurus,Tor-
nieria, and various titanosaurs (Janensch, 1935; Hopson, 1979;
Chatterjee and Zheng, 2002, 2004; Tidwell and Carpenter, 2003;
Sereno et al., 2007; Witmer et al., 2008; Balanoff et al., 2010;
Knoll et al., 2012, 2013; Paulina Carabajal, 2012) except for two
variable features: relative proportions of the space for the venous
sinus and dural peak and absence/presence of frontoparietal
fenestra and postparietal foramen. In these two characters, the
skull roof of Qijianglong is more similar to that of Spinophoro-
saurus than to those of other sauropods. Although the presence/
absence of the frontoparietal fenestra and postparietal foramen
individually varies within Camarasaurus and Diplodocus
(Witmer et al., 2008; Knoll et al., 2012), neither of these open-
ings has been identi�ed in any of the skulls of Mamenchisaurus
and Omeisaurus (Xing et al., 2013). Unless future discovery
shows individual variation in these openings within a mamenchi-
saurid taxon, the presence of the postparietal foramen in
Qijianglong is taxonomically signi�cant.
Laterosphenoid—The left laterosphenoid (Fig. 3) is longer
anteroposteriorly than tall dorsoventrally and has unfused
sutures with the frontal (dorsally), orbitosphenoid (anteriorly),
prootic (posteriorly), and basisphenoid (ventrally). The foramina
for the oculomotor and trochlear nerves (CNs III and IV) open
side by side, with the former at the laterosphenoid-basisphenoid
suture. The foramen for the trigeminal nerve (CN V) sits at the
laterosphenoid-prootic suture. The canal for CN V passes below
this suture from the endocranial cavity to the external surface of
the braincase. A foramen under a tuber anterior to the foramen
of CN V may have housed the anterior middle cerebral vein.
Exoccipital and Supraoccipital—The exoccipitals and the
lower half of the supraoccipital (Fig. 4) formed the lateral and
dorsal margins of the foramen magnum. Unlike Omeisaurus,
Shunosaurus,Spinophorosaurus, and cf. Cetiosaurus (He et al.,
1988; Tang et al., 2001; Chatterjee and Zheng, 2002; Galton and
Knoll, 2006; Remes et al., 2009; Knoll et al., 2012), the foramen
magnum is at least twice as tall dorsoventrally as wide trans-
versely, or may even be taller than that if fully reconstructed.
Amongst sauropods, the dorsoventrally tall foramen magnum is
e889701-4JOURNAL OF VERTEBRATE PALEONTOLOGY
Downloaded by [Kangwon National University - Samcheok Campus] at 15:51 19 February 2015
generally a condition observed in neosauropods (both within dip-
lodocoids and macronarians), M. youngi, and Turiasaurus
(Janensch, 1935; Salgado and Bonaparte, 1991; Calvo and Sal-
gado, 1995; Chatterjee and Zheng, 2002; Ouyang and Ye, 2002;
Tidwell and Carpenter, 2003; Curry Rogers and Forster, 2004;
Wilson, 2005; Wilson et al., 2005; Harris, 2006a; Paulina
Carabajal and Salgado, 2007; Garcia et al., 2008; Balanoff et al.,
2010; Royo-Torres and Upchurch, 2012).
The jugular foramen and the fenestra ovalis are separated by
an incomplete crista interfenestralis. The columellar canal and
the groove leading to the jugular foramen extend on the antero-
ventral surface of the paroccipital process in parallel. The
FIGURE 2. Skull roof of Qijianglong guokr (QJGPM 1001). A, photograph; B, interpretive drawing in dorsal view; C, photograph; D, interpretive
drawing in right lateral view; E, photograph; F, interpretive drawing in ventral view. Arrow with number indicates a character diagnostic to this taxon
(number refers to the list of characters in the Diagnosis).
JOURNAL OF VERTEBRATE PALEONTOLOGY e889701-5
Downloaded by [Kangwon National University - Samcheok Campus] at 15:51 19 February 2015
incompletely fused exoccipital-opisthotic suture crosses these
grooves posteroventrally. The foramen for the posterior middle
cerebral vein opens below the suture with the supraoccipital.
The canal for this vein passes posteriorly through a small sinus
before exiting the endocranial cavity. The canal for the posterior
branch of the hypoglossal nerve (CN XII) is enclosed within a
depression on the medial surface of the exoccipital, just anterior
to the margin of the foramen magnum. The paroccipital process
is slender and without marked distal expansion.
Basioccipital, Basisphenoid, and Parasphenoid—The basicra-
nium (Fig. 5) is nearly complete. The sutures with the latero-
sphenoid, prootic, opisthotic, and exoccipital are all unfused.
Overall, the basicranium is dorsoventrally low and anteroposter-
iorly long such that the diameter of the occipital condyle is
greater than the distance between the basal tuber and the base
of the neck for the occipital condyle, and such that the basal
tubera and the bases of the basipterygoid process form a square
in ventral view. The distance between the basal tubera is greater
than the width of the occipital condyle. The craniopharyngeal
foramen as described for some neosauropods (Balanoff et al.,
2010) is absent. Instead, the notch between the basal tubera leads
to the space between the basipterygoid processes on the ventral
surface of the basisphenoid.
The basipterygoid process extends anteroventrally and slightly
laterally. In lateral view, the extended axis of the basipterygoid
process meets the floor of the endocranial cavity at an angle of
120, and the angle between the parasphenoid rostrum and the
basipterygoid process is accordingly smaller than perpendicular.
In anterior view, the right and left basipterygoid processes meet
almost perpendicular to each other (87). The basipterygoid pro-
cess is plate-like, not round at the end as in Shunosaurus (Chatter-
jee and Zheng, 2002). The process has an accessory tuber near the
base in a direction roughly parallel with the basal tuber, which is
unique to Qijianglong among sauropods. The crista prootica is
anterodorsally oblique with respect to the floor of the endocranial
cavity. Behind this crista posteriorly at midheight is a large fora-
men for the internal carotid artery. On the anterior side of the
crista, a fossa sits around the external foramen for the abducens
nerve. The parasphenoid rostrum extends more anteriorly than
the basipterygoid process, and the base of the rostrum is at the
similar horizontal level with the basal tuber.
FIGURE 3. Left laterosphenoid of Qijianglong guokr (QJGPM 1001) in
A, lateral view; B, posterior view; C, medial view.
FIGURE 4. Occiput of Qijianglong guokr (QJGPM 1001). The supraoccipital and left exoccipital in A, posterior view; B, ventral view. The right
exoccipital in C, anteromedial view; D, anterolateral view.
e889701-6JOURNAL OF VERTEBRATE PALEONTOLOGY
Downloaded by [Kangwon National University - Samcheok Campus] at 15:51 19 February 2015
These features of the basipterygoid process and the parasphenoid
rostrum in Qijianglong are distinct in comparison with other basal
sauropods. The basipterygoid process extends vertically and is per-
pendicular to both the parasphenoid rostrum and the floor of the
endocranial cavity in Chebsaurus,M. youngi,Shunosaurus,Turia-
saurus, and basal non-sauropod sauropodomorphs such as Anchi-
saurus and Plateosaurus (Galton, 1984; Benton et al., 2000;
Chatterjee and Zheng, 2002; Ouyang and Ye, 2002; L€
ang and
Mahammed, 2010; Royo-Torres and Upchurch, 2012). Shunosaurus
also differs in that the base of the parasphenoid rostrum is lower in
position than the basal tuber and that the basipterygoid processes
meet at ‘U’-shape at an angle substantially broader than 90in ante-
rior view. In contrast, the basipterygoid process is oriented postero-
ventrally in parallel with the basal tuber in Atlasaurus and
Spinophorosaurus as if the accessory tuber of the basipterygoid pro-
cess in Qijianglong was greatly extended (Monbaron et al., 1999;
Remes et al., 2009; Knoll et al., 2012).
In the basicranium, the canal for the trigeminal nerve (CN V) is
posterior with respect to the crista prootica at the floor of the endo-
cranial cavity. The canal for the facial nerve (CN VII) is also near
the floor of the endocranial cavity and above the notch between the
basipterygoid process and the basal tuber in lateral view. The bro-
ken plane intersects these two canals and represents the contact
surface with the exoccipital-opisthotic, prootic, and laterosphenoid.
Prefrontal—The right prefrontal (Fig. 2) sits on the dorsal sur-
face of the frontal and does not contact with the postorbital pos-
teriorly. The main part of the element collapsed into the orbit.
With restoration in this region, the frontal and prefrontal likely
had subequal participation in the dorsal margin of the orbit. The
prefrontal contacts the nasal along the medial margin, the max-
illa at the anterior end, and the lacrimal at the lower end of the
preorbital ramus. A fragment of the lacrimal is still attached
near the contact, and the nasal sits medial to the prefrontal. This
region is too weathered to make comparisons with other taxa.
Postorbital—The right postorbital (Fig. 6) is ‘T’-shaped in
both lateral and dorsal views. The postorbital bar is expanded
laterally such that the outline of the bone is markedly convex
laterally in dorsal view. The frontal and squamosal processes are
subequal in anteroposterior length. The frontal process wraps
around the posterolateral corner of the frontal, whereas the
squamosal process extends posteriorly. The dorsal surface of the
postorbital is nearly flat and unlike the weakly concave dorsal
surface in Omeisaurus spp. and Turiasaurus (He et al., 1988;
Tang et al., 2001; Royo-Torres and Upchurch, 2012). In M.
FIGURE 5. Basicranium (parasphenoid, basisphenoid, and basioccipital) of Qijianglong guokr (QJGPM 1001) in A, left lateral view; B, dorsal view; C,
ventral view; D, posterior view. Arrow with number indicates a character diagnostic to this taxon (number refers to the list of characters in the Diagnosis).
FIGURE 6. Right postorbital of Qijianglong guokr (QJGPM 1001) in
A, lateral view; B, dorsal view; C, anterior view; D, posteromedial view.
JOURNAL OF VERTEBRATE PALEONTOLOGY e889701-7
Downloaded by [Kangwon National University - Samcheok Campus] at 15:51 19 February 2015
youngi, the dorsal margin is strongly concave such that the supra-
temporal fenestra is round in lateral view (Ouyang and Ye,
2002). On the medial surface of the postorbital bar, the ridge
extends along the anterior margin to delineate the orbit. The
jugal contacts the postorbital along the posterior margin.
Quadrate—The long axis of the right quadrate (Fig. 7) is
nearly vertical as in basal eusauropods but unlike those that are
inclined posterodorsally in diplodocoids and derived titanosauri-
forms (Upchurch et al., 2004). The shaft of the quadrate is
inflated by a large pneumatic fossa as in M. youngi and Turiasau-
rus (Ouyang and Ye, 2002; Royo-Torres and Upchurch, 2012),
whereas the fossa is shallow in Shunosaurus (Chatterjee and
Zheng, 2002). The apex of the anteriorly expanded pterygoid ala
is in the upper half of the element, whereas it is in the lower half
in O. tianfuensis (He et al., 1988).
Pterygoid—The partial right pterygoid (Fig. 8) consists of the
quadrate ala and basipterygoid process. The quadrate ala is tri-
angular in lateral view. The basipterygoid process forms a shelf
on the medial side of the ala as in M. youngi (Ouyang and Ye,
2002), rather than extending in a long process ventrally as in neo-
sauropods such as Dicraeosaurus and Giraffatitan (Janensch,
1935). The main shaft extends anteriorly under this shelf.
Articular—The right articular (Fig. 9) has a conspicuous retro-
articular process posterolateral to the articular surface with the
quadrate. With the exception of the retroarticular process, the
lateral surface of the articular was overlapped by the surangular
(not preserved). The medial surface of the articular has a fossa
within which the posterior end of the prearticular (not pre-
served) �t. The articular surface has a shallow pro�le, with slight
concavity in lateral view.
Postcranial Skeleton
Nomenclature for the vertebral laminae and pneumatic fossae
follows Wilson (1999) and Wilson et al. (2011), respectively.
Cervical Vertebrae—The completely preserved cervical series
of Qijianglong consists of 17 vertebrae (Figs. 10–12). The axis to
the 11th cervical vertebra were fully articulated in the quarry.
The atlas intercentrum and the 12th–17th cervical vertebrae
were closely associated with the series. Except for the amphicoe-
lous atlas intercentrum, the cervical vertebrae are all opisthocoe-
lous. Overall, the anterior cervical vertebrae (3rd–5th) have
relatively anteroposteriorly elongate centra in Qijianglong that
are typically more than three times as long as high (Supplemen-
tary Data, Table S1). The relative centrum length in this region
FIGURE 7. Right quadrate of Qijianglong guokr (QJGPM 1001) in A,
lateral view; B, medial view; C, posterior view, rotated 90
counterclockwise.
FIGURE 8. Right pterygoid of Qijianglong guokr (QJGPM 1001) in A,
dorsal view; B, medial view.
FIGURE 9. Right articular of Qijianglong guokr (QJGPM 1001) in A,
lateral view; B, dorsal view; C, ventral view.
e889701-8JOURNAL OF VERTEBRATE PALEONTOLOGY
Downloaded by [Kangwon National University - Samcheok Campus] at 15:51 19 February 2015
of the cervical series is greater in this taxon than in Chuanjiesau-
rus,M. anyuensis,M. hochuanensis,M. sinocanadorum, and Ton-
ganosaurus and is roughly comparable to M. youngi and O.
tianfuensis (Table S1; Russell and Zheng, 1993; He et al., 1996;
K. Li et al., 2010b; Sekiya, 2011).
The semilunar atlas intercentrum (Fig. 10) contacts the odon-
toid process of the axis posteroventrally (Fig. 11A). The anterior
articular surface of the axis is rugose and flat, indicating an
incompletely fused contact with the atlas intercentrum. The pla-
nar spinous process of the axis extends over the dorsomedially
oriented spinopostzygapophyseal lamina. As a result, the spino-
postzygapophyseal fossa forms a deep recess closed dorsally by
the spinous process. The posterior centrodiapophyseal lamina is
incomplete. The diapophysis is more anterior than the longitudi-
nal midpoint of the centrum and extends from the edge of the
anterior articular surface via the anterior centrodiapophyseal
lamina. The prezygodiapophyseal lamina is incomplete and not
directly connected to the diapophysis. The axis has three pleuro-
coels on the right side and two on the left. On the right side, each
pleurocoel is round and successively smaller posteriorly. On the
left side, the anterior pleurocoel is more than double the area of
the posterior one. The pleurocoels on the left side are anteropos-
teriorly elongate.
In the anterior to mid-cervical region (Fig. 11B–H), the spi-
nous process forms a longitudinal plate that is split posteriorly
into the bilaterally paired spinopostzygapophyseal laminae in all
of the 6th –8th vertebrae. The lamina is oriented primarily ante-
rodorsally such that the spinopostzygapophyseal fossa is open
dorsally. The spinodiapophyseal fossa is a shallow depression. In
lateral view, the prezygodiapophyseal lamina overhangs the cen-
trum and overlaps the postzygapophysis of the previous vertebra
from lateral side. The posterior centrodiapophyseal lamina
extends posteriorly in a straight line to the base of the neural
arch, thereby separating the postzygapophyseal centrodiapophy-
seal fossa from the deeply excavated centrodiapophyseal fossa.
This lamina is clearly more pronounced than the postzygodiapo-
physeal lamina in the anterior cervical series in Qijianglong, and
possibly in M. sinocanadorum, whereas this condition is typically
reversed in other mamenchisaurids, including Chuanjiesaurus,
M. youngi,Omeisaurus spp., and Tonganosaurus (He et al.,
1988; Russell and Zheng, 1993; Tang et al., 2001; Ouyang and
Ye, 2002; Sekiya, 2011). The postzygodiapophyseal lamina origi-
nates from the posterior centrodiapophyseal lamina posterodor-
sal to the diapophysis. The postzygodiapophyseal lamina is
prominent enough in the mid-cervical region (from the 5th/6th
onward) to overhang the centrum and extend posteriorly beyond
the posterior articular surface as in Chuanjiesaurus, but unlike
Mamenchisaurus spp., Omeisaurus spp., and Tonganosaurus (He
et al., 1988; Russell and Zheng, 1993; Tang et al., 2001; Ouyang
and Ye, 2002; K. Li et al., 2010b; Sekiya, 2011). The maximum
vertical height of the postzygapophysis is twice that of the spi-
nous process. The zygapophyses articulate with one another at
the level slightly above the centrum in the 7th and 8th, but at a
level noticeably higher above the centrum in the 6th cervical ver-
tebra. Although the centropostzygapophyseal fossa is not visible
in lateral view, it forms a deep pit set between the prominent
centropostzygapophyseal lamina and the postzygapophysis in
the 8th cervical vertebra. This fossa is absent in the 6th cervical
vertebra in which the postzygapophysis is raised high above the
level of the centrum.
In the 12th–14th cervical vertebrae of the posterior cervical
region (Fig. 12A–C), the postzygodiapophyseal lamina origi-
nates above the posterior centrodiapophyseal lamina as in the
cervical vertebrae of Chuanjiesaurus,M. youngi, and Omeisaurus
spp. (He et al., 1988, 1996; Tang et al., 2001; Ouyang and Ye,
2002; Sekiya, 2011), not from the latter lamina as in the more
anterior or posterior cervical vertebrae. The postzygapophysis is
well above the centrum. The 17th cervical vertebra is the last of
the cervical series (Fig. 12F) as in Omeisaurus. The neural arch
is relatively taller than in the more anterior cervical vertebrae.
The spinous process is as tall as the centrum, and the both zyg-
apophyses articulated well above the centrum. On the left side of
the process, the spinodiapophyseal fossa has three pneumato-
pores. There are at least seven, and likely more, pneumatopores
within the fossa on the right side of the spinous process. The pre-
zygapophysis is pneumatic and associated with a tubercle along
the anterior margin. Other posterior cervical vertebrae also have
pneumatopores in the spinous process, whereas the pneumato-
pores are absent in other mamenchisaurids such as Chuanjiesau-
rus,M. youngi, and Omeisaurus spp. (He et al., 1988, 1996; Tang
et al., 2001; Ouyang and Ye, 2002; Sekiya, 2011)
From the axis to at least the 14th cervical vertebra, a �nger-
like process extends posteriorly above the postzygapophysis and
overlaps onto the dorsolateral surface of the prezygapophysis of
the next vertebra (Fig. 11I, J). These processes are unique to
Qijianglong, unlike all previously known mamenchisaurids that
are preserved with cervical vertebrae (e.g., Chuanjiesaurus,
Mamenchisaurus spp., Omeisaurus spp., Tonganosaurus). There-
fore, the neck of Qijianglong presumably had a range of motion
restricted in sideways.
Dorsal Vertebrae—Six dorsal vertebrae are preserved from
the anterior thoracic region (Fig. 13). Although posterodorsal
crushing makes comparison dif�cult, the vertebrae likely repre-
sent the 1st–6th dorsals. These vertebrae are all opisthocoelous
and generally identical in morphology except for a dorsal shift in
position of the parapophysis posteriorly along the series. Based
on the better-preserved 3rd–6th vertebrae, the centra are more
strongly opisthocoelous rather than nearly amphiplatyan as in
Eomamenchisaurus (L€
u et al., 2008). The centra are anteroposter-
iorly longer than dorsoventrally tall as in Mamenchisaurus spp.
and Xinjiangtitan, but unlike Chuanjiesaurus,Eomamenchisaurus,
Omeisaurus spp., Tonganosaurus,andYuanmousaurus (He et al.,
1988, 1996; Tang et al., 2001; Ouyang and Ye, 2002; L€
uetal.,
2006, 2008; K. Li et al., 2010b; Sekiya, 2011; Wu et al., 2013). The
neural arches with the spinous processes are at least 1.5 times the
height of the centra. The bifurcated spinous processes of Qijiang-
long differ from those of Hudiesaurus in lacking the medial projec-
tions between the bifurcate processes (Dong, 1997).
FIGURE 10. Atlas intercentrum of Qijianglong guokr (QJGPM 1001)
in A, anterior view; B, posterior view.
JOURNAL OF VERTEBRATE PALEONTOLOGY e889701-9
Downloaded by [Kangwon National University - Samcheok Campus] at 15:51 19 February 2015
Based on the 3rd–6th dorsal vertebrae, the prezygapophyseal
centrodiapophyseal fossa extends onto the anterior surface of
the transverse process with four distinct pneumatopores,
whereas the fossa is smooth in other mamenchisaurids such as
M. youngi (Ouyang and Ye, 2002). The centrodiapophyseal fossa
is a low triangular depression with a pleurocoel tucked under-
neath the transverse process. The postzygapophyseal centrodia-
pophyseal fossa occupies about three times the area of the
FIGURE 11. Anterior cervical series of Qijianglong guokr (QJGPM 1001) in left lateral views unless otherwise noted. A, axis; B, cervical vertebra 3;
C, cervical vertebra 4; D, cervical vertebrae 5 and 6; E, cervical vertebra 7 and anterior half of cervical vertebra 8 (horizontally inverted; showing right
side); F, posterior half of cervical vertebra 8 and cervical vertebra 9; G, cervical vertebra 10; H, cervical vertebra 11; I, close-up of the prezygapophy-
sis-postzygapophysis contact between cervical vertebrae 3 and 4 in dorsolateral view, showing �nger-like process lateral to postzygapophysis; J, close-
up of the postzygapophysis of cervical vertebra 5 in dorsal view, showing �nger-like process lateral to postzygapophysis. Arrow with number indicates
a character diagnostic to this taxon (number refers to the list of characters in the Diagnosis). All scale bars equal 5 cm. Abbreviations:acdl, anterior
centrodiapophyseal lamina; cdf, centrodiapophyseal fossa; plc, pleurocoel; pocdl, postcentrodiapophyseal lamina; poz, postzygapophysis; pozcdf, post-
zygapophyseal centrodiapophyseal fossa; pozdl, postzygodiapophyseal lamina; ppoz, �nger-like process lateral to postzygapophysis; ppozc, groove for
contact with �nger-like process; przdl, prezygodiapophyseal lamina; sdf, spinodiapophyseal fossa.
e889701-10 JOURNAL OF VERTEBRATE PALEONTOLOGY
Downloaded by [Kangwon National University - Samcheok Campus] at 15:51 19 February 2015
prezygapophyseal centrodiapophyseal fossa in lateral view. This
condition is similar to that of O. tianfuensis and Tonganosaurus
(He et al., 1988; K. Li et al., 2010b). In other mamenchisaurids
such as Chuanjiesaurus,Mamenchisaurus spp., and Yuanmou-
saurus (He et al., 1996; Ouyang and Ye, 2002; L€
u et al., 2006;
Sekiya, 2011), the latter fossa is typically larger than the former.
The upper half of the postzygapophyseal centrodiapophyseal
fossa is separated into two chambers by a vertical lamina
descending from the spinopostzygapophyseal lamina. The ante-
rior of the two chambers represents an incipient postzygapophy-
seal spinodiapophyseal fossa. The postzygapophyseal
centrodiapophyseal fossa has a pneumatopore at the
FIGURE 12. Posterior cervical series of Qijianglong guokr (QJGPM 1001) in left lateral view unless otherwise noted. A, cervical vertebra 12;
B, cervical vertebra 13; C, cervical vertebra 14; D, cervical vertebra 15 (horizontally inverted; showing right side); E, cervical vertebra 16 (horizontally
inverted; showing right side); F, cervical vertebra 17 (horizontally inverted; showing right side). Arrow with number indicates a character diagnostic to
this taxon (number refers to the list of characters in the Diagnosis). All scale bars equal 5 cm. Abbreviations:spozf, spinopostzygapophyseal fossa;
spozl, spinopostzygapophyseal lamina.
FIGURE 13. Dorsal series of Qijianglong guokr (QJGPM 1001). The dorsal vertebrae are crushed dorsoventrally or transversely. A, dorsal vertebra
1 (dorsoventrally crushed) in dorsal view; B, incomplete dorsal vertebra 2 in left lateral view; C, dorsal vertebrae 3 and 4 in lateral view (horizontally
inverted; showing right side); D, dorsal vertebrae 5 and 6 in left lateral view. Abbreviations:cdf, centrodiapophyseal fossa; pozcdf, postzygapophyseal
centrodiapophyseal fossa; pozsdf, postzygapophyseal spinodiapophyseal fossa; przcdf, prezygapophyseal centrodiapophyseal fossa.
JOURNAL OF VERTEBRATE PALEONTOLOGY e889701-11
Downloaded by [Kangwon National University - Samcheok Campus] at 15:51 19 February 2015
dorsomedial corner in posterior view. The right and left counter-
parts of the fossa are set apart by a vertical ridge below the spi-
nopostzygapophyseal fossa.
Caudal Vertebrae—The caudal series is represented by 28 ver-
tebrae (Fig. 14). Although the precise identi�cation is dif�cult,
these vertebrae are from the middle to distal region of the tail.
One caudal centrum is procoelous and the other centra are
amphiplatyan.
Based on comparisons with Mamenchisaurus spp. and Omei-
saurus spp. (He et al., 1988; Tang et al., 2001; Ouyang and Ye,
2002), the procoelous centrum is from around the 10th caudal
positions. The posterior articular surface is convex for only 16%
of the diameter of the centrum. There is no pneumatic fossa on
the lateral surface of the centrum. The ventral surface of the cen-
trum has a longitudinal sulcus on the posterior half. The prezyga-
pophysis extends anterodorsally, and its distal tip is slightly
beyond the anterior articular surface of the centrum. The spino-
prezygapophyseal lamina is a simple low ridge on the dorsal half
of the spinous process. The transverse process has no centrodia-
pophyseal laminae. The spinous process is inclined posterodor-
sally at approximately 75.
The rest of the caudal vertebrae are from the middle to
distal caudal series (from the 15th onward). The spinous pro-
cess and the postzygapophysis are inclined posterodorsally
beyond the centrum and over half the length of the next cen-
trum as in Mamenchisaurus spp. (Ouyang and Ye, 2002). The
most distal two of the preserved caudal vertebral centra are
fused to each other (Fig. 13G). The fusion is possibly
pathologic.
Pubis—The left pubis (Fig. 15) is transversely flat and verti-
cally tall and has a vertically elongate inverted teardrop-shape in
cross-section. The anterior margin is deeply concave such that
the distal end of the pubis points more anteriorly than ventrally
in life position, whereas the general condition amongst sauro-
pods is a distal end of a pubis oriented more ventrally than ante-
riorly. The pubic foramen is enclosed within the peduncle for the
ischial contact, and the flange extends along the ventral margin
for half the length of the shaft. This suite of traits differs from
Eomamenchisaurus,M. youngi, and O. maoianus in being more
robust, from Chuanjiesaurus in having a pubic foramen, from O.
tianfuensis in having a clearly demarcated, concave acetabular
margin, and from Xinjiangtitan in lacking a marked constriction
proximal to the pubic apron (He et al., 1988; Tang et al., 2001;
Ouyang and Ye, 2002; L€
u et al., 2008; Sekiya, 2011; Wu et al.,
2013).
Other Postcranial Elements—The dorsal ribs are preserved in
fragments, with a pneumatopore at the base of the capitulum.
The hemal arches from the mid-caudal positions are closed dor-
sally by a bony bridge between the right and left articular facets,
whereas the arches from the distal caudal positions are open.
Both of the two pedal phalanges represent the proximal phalanx
of each digit, but the precise identi�cation of the digits is
uncertain.
PHYLOGENETIC ANALYSIS
A maximum parsimony analysis used the data set of Harris
(2006b), with modi�cations by Xing et al. (2013) and the addi-
tion of Qijianglong (Supplementary Data, Appendix S1). In this
matrix, ‘Prosauropoda’ were split into Plateosaurus and Theco-
dontosaurus.Mamenchisaurus and Omeisaurus were split into
multiple species for which information is available from the liter-
ature and the authors’ collections visits: M. anyuensis,M. con-
structus,M. hochuanensis,M. sinocanadorum,M. youngi,O.
maoianus, and O. tianfuensis (Young, 1958; Young and Zhao,
1972; He et al., 1988, 1996; Russell and Zheng, 1993; Tang et al.,
2001; Ouyang and Ye, 2002). The following taxa were added to
Harris’s (2006b) data set: Atlasaurus,Chuanjiesaurus,Liraino-
saurus,Nebulasaurus,Spinophorosaurus,Tornieria,Turiasaurus,
and Yuanmousaurus (Monbaron et al., 1999; Sanz et al., 1999;
L€
u et al., 2006; Remes, 2006; Royo-Torres et al., 2006; Remes
et al., 2009; D
ıaz et al., 2011; Sekiya, 2011; Knoll et al., 2012;
Royo-Torres and Upchurch, 2012; Xing et al., 2013). Harris’s
(2006b) characters 38 and 76 were modi�ed, and 13 braincase
characters were added (Xing et al., 2013). Character codes for
individual taxa were extensively modi�ed to follow recently
added information in the literature. For the current data set,
scorings were modi�ed for Barapasaurus,Brachiosaurus,
FIGURE 14. Caudal series of Qijianglong guokr (QJGPM 1001). A mid-caudal vertebra in A, anterior view; B, left lateral view; C, posterior view.
Distal caudal series in left lateral view: D, possible caudal vertebrae 15–20; E, possible caudal vertebrae 21–26; F, possible caudal vertebrae 27–39;
G, close-up photograph of possible caudal vertebrae 35–41; H, close-up photograph of possible caudal vertebrae 40 and 41. The centrum of the caudal
vertebra 41 is fused to that of the caudal vertebra 40. Arrow with number indicates a character diagnostic to this taxon (number refers to the list of
characters in the Diagnosis).
e889701-12 JOURNAL OF VERTEBRATE PALEONTOLOGY
Downloaded by [Kangwon National University - Samcheok Campus] at 15:51 19 February 2015
Nigersaurus, and Euhelopus following Sereno et al. (2007), Tay-
lor (2009), Wilson and Upchurch (2009), Nair and Salisbury
(2012), and Poropat and Kear (2013). The following changes
were made based on personal communications from J. D. Harris
(pers. comm., 2012): Camarasaurus (characters 101–149 as miss-
ing ‘?’); Limaysaurus (character 319 from ‘2’ to ‘?’); and Nemeg-
tosaurus (characters 301–305 as missing ‘?’). The following
characters were parsimony uninformative and therefore
removed from the analysis: 40, 47, 89, 136, 271, and 336. The cur-
rent data set includes 45 taxa (including three outgroups: Thero-
poda, Plateosaurus, and Thecodontosaurus) and 338 characters.
All characters were treated as unordered.
A heuristic search by PAUP b.4.01 (Swofford, 2003) with mul-
tiple TBR CTBR search strategy recovered more than 29,500
most parsimonious trees (MPTs; tree length [TL] D1056; consis-
tency index [CI] D0.400; retention index [RI] D0.643; rescaled
consistency index [RC] D0.257). A strict consensus of all MPTs
(Fig. 16A) supports a monophyletic Mamenchisauridae, with O.
tianfuensis as the sister taxon to the rest of the clade, and with
Chuanjiesaurus and Qijianglong nested outside a polytomy
including Mamenchisaurus spp., O. maoianus, and Yuanmousau-
rus.Qijianglong therefore represents a mamenchisaurid lineage
that extends back at least to the Middle Jurassic and is not the
closest relative of M. anyuensis, which occurs in the upper part
of the same formation. This result also provides the �rst phyloge-
netic support for the mamenchisaurid af�nity of Yuanmousau-
rus, which was originally compared with Patagosaurus (L€
u et al.,
2006). Coupled with the undiagnosable nature of the holotype
(Xing et al., 2013), Yuanmousaurus cannot be readily distin-
guished from Eomamenchisaurus from the same Zhanghe For-
mation on the basis of the available information (L€
u et al.,
2008). Although this paper accepts the priority of Eomamenchi-
saurus over Yuanmousaurus because of the undiagnostic nature
of the latter, these two genera await further detailed description
of the type materials and taxonomic revision.
In the second round of the heuristic search under the same set-
ting, M. sinocanadorum and Yuanmousaurus were removed from
the analysis. Scored characters for M. sinocanadorum have no
overlap with those of the type species of Mamenchisaurus,M.
constructus, in the data matrix. This second analysis recovered
542 MPTs (TL D1046; CI D0.403; RI D0.644; RC D0.260). A
strict consensus of the MPTs partly resolves the polytomy of
Mamenchisaurus spp. and O. maoianus and is identical to the
strict consensus of the whole data set in the rest of the tree
FIGURE 15. Left pubis of Qijianglong guokr (QJGPM 1001) in lateral
view. Arrow with number indicates a character diagnostic to this taxon
(number refers to the list of characters in the Diagnosis).
FIGURE 16. Results of maximum parsimony analyses for Qijianglong and its relationships to other sauropods. A, strict consensus of shortest trees
from a maximum parsimony analysis of 45 taxa including Qijianglong guokr based on the data set modi�ed from Harris (2006b) (see Supplementary
Data); B, a part of strict consensus of shortest trees using the same data set without Mamenchisaurus sinocanadorum and Yuanmousaurus. The
mamenchisaurid interrelationships are better resolved, and the rest of the tree is identical to A. See text for description of the trees, tree statistics, and
discussion.
JOURNAL OF VERTEBRATE PALEONTOLOGY e889701-13
Downloaded by [Kangwon National University - Samcheok Campus] at 15:51 19 February 2015
(Fig. 16B). Each of the pairs M. hochuanensis CO. maoianus
and M. youngi CM. anyuensis forms a clade, and these two
clades and M. constructus form a polytomy.
In both analyses, the Mamenchisauridae sits in a relatively
basal position, more derived than Shunosaurus but outside of the
rest of eusauropods. The clade is supported by 10 unambiguous
character changes in each of the analyses, but the supporting
character changes slightly differ between the two (*analysis 1;
**analysis 2; ‘0’ to ‘1’ for characters 64**, 100, 102, 110, 144, 146,
148, 307, 337*; ‘1’ to ‘2’ for character 126; ‘1’ to ‘2’ for character
126; ‘4’ to ‘6’ for character 105). These characters are sagittal and
transverse nuchal crests merging smoothly (character 64); den-
ticles absent on distal margin of tooth (100); procumbent teeth
(102); more than 15 cervical vertebrae (105); fossa above para-
pophysis of cervical vertebra confluent with lateral pneumatic
fossa (110); hypantrum-hyposphene contact in dorsal vertebrae
(126); spinopostzygapophyseal lamina unconnected to postspinal
lamina in posterior dorsal vertebra (144); spinodiapophyseal and
spinopostzygapophyseal laminae contacting each other in poste-
rior dorsal vertebra (146); moderate triangular process at distal
end of neural spine on dorsal vertebra (148); ossi�ed calcaneum
absent (307); and supraoccipital wider transversely than tall ver-
tically (337).
The clade of mamenchisaurids with the exclusion of O. tian-
fuensis is supported by six unambiguous character changes in
both analyses (‘0’ to ‘1’ for characters 28, 150, 166, 168, 207; ‘2’
to ‘3’ in character 143). These characters are frontal-parietal
suture anterior to supratemporal fenestra (28); prespinal and spi-
noprezygapophyseal lamina connected in posterior dorsal verte-
bra (143); opisthocoelous posterior dorsal vertebra (150);
procoelous �rst and proximal caudal vertebrae (166, 168); and
dorsal-most point of acromion process posteriorly displaced
(207).
Further internal nodes within the Mamenchisauridae are sup-
ported by two characters each. The polytomy of Mamenchisau-
rus spp. and O. maoianus in the second analysis is characterized
by characters 156 (‘1’ to ‘2’; �ve sacral vertebrae) and 169 (‘0’ to
‘2’; transversely compressed articular surface of proximal caudal
vertebra). The clade of M. hochuanensis and O. maoianus is sup-
ported by ‘0’ to ‘1’ unambiguous changes in characters 248 and
294 (longest metacarpal about 35% to 45% the length of radius;
tibia transversely twice wider at distal condyle as at midshaft).
The clade of M. anyuensis and M. youngi is supported by ‘0’ to
‘1’ unambiguous changes in characters 225 and 287 (proximolat-
eral process of humerus reduced; distal condyle for tibia more
than twice wider than that for �bula in femur).
DISCUSSION AND CONCLUSION
Qijianglong is a signi�cant addition to the Asian fossil record
of sauropods because it is the �rst mamenchisaurid de�nitively
distinct from Mamenchisaurus spp. from the Late Jurassic of
China. Not only does it increase the generic diversity of mamen-
chisaurids from that time interval, but the seemingly derived
morphology of Qijianglong and the basal phylogenetic position
of mamenchisaurids suggests that mamenchisaurids indepen-
dently evolved conditions convergent with other sauropods. The
postparietal foramen is absent in other mamenchisaurids but
widespread among eusauropods. The presence of an accessory
tuber of the basipterygoid process extending in parallel with the
basal tuber is unique among sauropods. The closest condition
occurs in Atlasaurus and Spinophorosaurus in which the basip-
terygoid process extends in parallel with the basal tuber. The
extensive pneumatization of the cervical vertebrae is unlike
other mamenchisaurids and reminiscent of diplodocoids.
The geographic isolation of Asia in Late Jurassic times
(Russell, 1993; Barrett and Upchurch, 2005; Mannion et al.,
2011) could explain both the low diversity of sauropodomorphs
and the convergent morphology of Qijianglong with distantly
related sauropods. During this time interval in Asia mamenchi-
saurids are currently the sole sauropods (Table 1). The low spe-
cies richness of mamenchisaurids from the Late Jurassic of Asia
may be attributed to a smaller number of fossiliferous terrestrial
localities than for preceding time intervals. However, the parsi-
mony analysis presented in this paper (Fig. 16) indicates that
both basal and derived lineages of mamenchisaurids existed in
Late Jurassic times. As such, mamenchisaurids do not appear to
have gone through a bottleneck across the Middle–Late Jurassic
boundary. Current evidence suggests that sauropod linages other
than mamenchisaurids did not survive into Late Jurassic times.
This endemic sauropod fauna in the Late Jurassic of Asia was
replaced by titanosauriforms across the Jurassic–Cretaceous
boundary.
A surprising result of the parsimony analysis was the relatively
basal position of mamenchisaurids (Fig. 16). In previous analy-
ses, Mamenchisaurus and Omeisaurus are typically recovered as
relatively derived non-neosauropod eusauropods, often just out-
side the Neosauropoda (Harris, 2006b; Royo-Torres et al., 2006,
2009; Remes et al., 2009; L€
ang and Mahammed, 2010; Sekiya,
2011; Nair and Salisbury, 2012; Royo-Torres and Upchurch,
2012). Instead, the current analysis suggests that mamenchisaur-
ids represent an ancient lineage of basal eusauropods only
slightly more derived than Shunosaurus. This basal position is
consistent with the Early Jurassic age of the putative mamenchi-
saurid Tonganosaurus (K. Li et al., 2010b).
The analysis also recovered a monophyletic Mamenchisauri-
dae and subclades within the lineage (Fig. 16). Mamenchisauri-
dae can be de�ned as a stem-based clade more closely related to
M. constructus,M. youngi,O. junghsiensis, and O. tianfuensis
than to Shunosaurus,Barapasaurus,Patagosaurus,orSpinophor-
osaurus. This result contradicts the recovery of Mamenchisauri-
dae excluding Omeisaurus as proposed by Sekiya (2011). In
further comparison with Sekiya’s (2011) analysis, Mamenchisau-
rus spp. form a paraphyletic assemblage in the present analysis
as opposed to a polyphyletic assemblage, and Chuanjiesaurus is
nested outside the ‘Mamenchisaurus’ polytomy as opposed to
being found in a derived position as the sister taxon to M.
hochuanensis. These differences of opinion regarding mamenchi-
saurid interrelationships highlight the need for a taxonomic revi-
sion of Mamenchisaurus and Omeisaurus. Taken at face value,
the present analysis suggests that (1) O. maoianus and O. tian-
fuensis are likely not congeneric; (2) the genus Mamenchisaurus
should be restricted to the type species, M. constructus, because
of the poorly resolved relationships among derived mamenchi-
saurids; and (3) each of the ‘Mamenchisaurus’ clades (M.
hochuanensis CO. maoianus;M. anyuensis CM. youngi), if
tested positive, represent distinct lineages. However, this analysis
does not include the type species O. junghsiensis, other described
species of Mamenchisaurus and Omeisaurus, and newly
described mamenchisaurids such as Eomamenchisaurus,Tonga-
nosaurus, and Xinjiangtitan (L€
u et al., 2008; K. Li et al., 2010b;
Wu et al., 2013). These taxa were omitted from the analysis
because published information was not suf�cient, because the
authors have not examined the materials, or because their inclu-
sion was not justi�ed like other highly incomplete taxa (e.g.,
Yuanmousaurus,M. constructus) that are crucial for tests of tax-
onomic validity. Therefore, it remains uncertain whether or not
some species might turn out to be synonyms, and whether either
or neither of the two species of Omeisaurus included in the anal-
ysis better represents that genus.
A mamenchisaurid taxonomic revision is far beyond the scope
of this paper. Reexamination of specimens referred to the type
species (M. constructus and O. junghsiensis) would be a reason-
able starting point, because the generic and speci�c diagnosis
remains challenging with referred materials from different strati-
graphic units (e.g., Young, 1958). Meanwhile, a temporary
e889701-14 JOURNAL OF VERTEBRATE PALEONTOLOGY
Downloaded by [Kangwon National University - Samcheok Campus] at 15:51 19 February 2015
solution may be to exercise caution in using the generic names
for other referred species until the generic diagnosis is resolved.
A thorough cladistic analysis of all mamenchisaurids should
serve as a guideline for recombination and formulation of taxo-
nomic names. This process is underway by the authors. In the
presence of many species referred to the two genera, and in the
absence of species-level phylogeny of mamenchisaurids, it may
be desirable to establish a generic distinction for distinct mamen-
chisaurid taxa.
The holotype of Qijianglong likely represents an immature
individual because the sutures between the braincase elements
remain unfused in the holotype. In comparison with the holotype
of M. youngi, the skull is approximately 25% larger, and the axis
is comparable in length (approximately 10% longer in Qijiang-
long; Tables S1, S2; Ouyang and Ye, 2002). Each anterior cervi-
cal vertebra (3rd–6th) of Qijianglong has a centrum both
absolutely and relatively longer than in M. youngi (Fig. 17).
However, each of the mid-cervical vertebrae (7th–10th) has a
both absolutely and relatively shorter centrum in Qijianglong
than in M. youngi. The posterior cervical vertebrae (11th
onward) of Qijianglong have comparable length/height ratio of
the centra with M. youngi, but absolutely shorter and lower than
those of M. youngi. Length/height ratios of the cervical vertebrae
follow similar trends in mamenchisaurids. M. hochuanensis dif-
fers from smaller mamenchisaurids in that the centra of the ante-
rior cervical vertebrae are not exceedingly more elongate than
those of the cervical vertebrae that follow in the series (Fig. 17).
Also in this taxon, the length/height ratios tend to be lower than
in other mamenchisaurids. However, the absolute (numerical)
length of the vertebral centra is greatest between the 9th and
13th cervical vertebrae in all mamenchisaurids examined,
regardless of the relative length.
Some of these proportional differences may be taxonomically
informative, but neither neck length nor individual vertebral
length necessarily increases isometrically with respect to body
size. It is possible that the cervical vertebrae have allometric
growth in length within each species of mamenchisaurids (onto-
genetic allometry) or among species (interspeci�c allometry).
Although sample size is insuf�cient to test these hypotheses, the
fact that the holotype of M. youngi has an absolutely longer neck
than that of Qijianglong but a similar skull size suggests that pro-
portional differences may be taxonomically meaningful at similar
body sizes. However, the proportional differences should not be
used in comparison with substantially larger or smaller speci-
mens of mamenchisaurids until the ontogenetic and interspeci�c
allometries of cervical vertebrae are well resolved among
mamenchisaurids.
Despite the immature status of the holotype, Qijianglong is
still distinguished from other mamenchisaurids and other sauro-
pods by a number of autapomorphies. With the exception of one
proportional character (semi-equal frontal and parietal lengths),
all diagnostic characters are discrete (see Diagnosis). Although
the discrete nature of the characters does not rule out the possi-
bility of ontogenetic or allometric transformation, these diagnos-
tic characters do not occur in M. youngi, which is of similar body
size. Stratigraphically, both Qijianglong and M. anyuensis occur
in the Suining Formation. Most notably among these differences,
the cervical vertebrae of M. anyuensis are clearly distinguished
from those of Qijianglong because they lack the �nger-like pro-
cess beside the postzygapophysis (He et al., 1996).
ACKNOWLEDGMENTS
The authors thank directors, collections managers, and cura-
tors at numerous institutions that they visited for this project. T.
M. especially thanks those at the Institute of Vertebrate Paleon-
tology and Paleoanthropology, Zigong Dinosaur Museum, and
Qijiang Petri�ed Wood and Dinosaur Footprint National Geo-
logical Park Museum (China) for access to specimens in their
care and for their hospitality. O. Mateus (Universidade Nova de
Lisboa, Portugal), J. D. Harris (Dixie State College, U.S.A.), E.
B. Koppelhus, K. Miyashita, and W. S. Persons (University of
Alberta, Canada), P. Upchurch (University College London, U.
K.), J. A. Wilson (University of Michigan, U.S.A.), and H.-L.
You (Institute of Geology, China) provided discussion, data sets,
and/or logistic support. A. Paulina Carabajal and O. Wings pro-
vided careful reviews, and P. Druckenmiller and H.-L. You’s
attention to detail improved the style and presentation of the
manuscript. Financial aid for this project came from Qijiang
County Bureau of Land and Resources, Chongqing, China (to L.
X.), Vanier CGS and Alberta Innovates PGS (to T.M.), and
NSERC (to P.J.C.). Author contributions: L.X., J.Z., D.L., and
F.W. conducted the �eld work and did the initial research. T.M.
and L.X. executed description and comparison. L.X., T.M., J.Z.,
D.L., Y.Y., T.S., F.W., and P.J.C. provided materials and analyti-
cal tools. T.M., L.X., and P.J.C. drafted the manuscript.
LITERATURE CITED
Balanoff, A. M., G. S. Bever, and T. Ikejiri. 2010. The braincase of
Apatosaurus (Dinosauria: Sauropoda) based on computed tomogra-
phy of a new specimen with comments on variation and evolution in
sauropod neuroanatomy. American Museum Novitates 3677:1–29.
Barrett, P. M., and P. Upchurch. 2005. Sauropodomorph diversity
through time: possible macroevolutionary and palaeoecological
implications; pp. 125–156 in K. A. Curry-Rogers and J. A. Wilson
(eds.), Sauropod Evolution and Paleobiology. University of Califor-
nia Press, Berkeley, California.
Benton, M. J., L. Juul, G. W. Storrs, and P. M. Galton. 2000. Anatomy
and systematics of the prosauropod Thecodontosaurus antiques
from the Upper Triassic of southwest England. Journal of Verte-
brate Paleontology 20:77–108.
FIGURE 17. Comparison of length/height ratios of the cervical verte-
bral centra among representative sauropods from China with a relatively
complete cervical series. Euhelopus is chronologically and systematically
distant from Qijianglong but is included because it shows similar propor-
tions of the cervical vertebral centra to those of Qijiangling. Measure-
ments for the atlas are excluded. The lengths were taken as the
maximum horizontal distance between the anterior and posterior articu-
lar surfaces, and the heights were taken as the vertical diameter of the
posterior articular surface. The original measurements are available in
Supplementary Data. Specimens from which measurements were taken:
Euhelopus zdanskyi (PMU 233); Mamenchisaurus hochuanensis (holo-
type; Young and Zhao, 1972); Mamenchisaurus youngi (ZDM 83;
Ouyang and Ye, 2002); Omeisaurus tianfuensis (ZDM T5703; He et al.,
1988); Qijianglong guokr (QJGPM 1001); and Shunosaurus (ZDM
T5401).
JOURNAL OF VERTEBRATE PALEONTOLOGY e889701-15
Downloaded by [Kangwon National University - Samcheok Campus] at 15:51 19 February 2015
Calvo, J. O., and L. Salgado. 1995. Rebbachisaurus tessonei sp. nov. A
new Sauropoda from Albian-Cenomanian of Argentina; new evi-
dence on the origin of the Diplodocidae. Gaia 11:13–33.
Chatterjee, S., and Z. Zheng. 2002. Cranial anatomy of Shunosaurus,a
basal sauropod dinosaur from the Middle Jurassic of China. Zoolog-
ical Journal of the Linnean Society 136:145–169.
Chatterjee, S., and Z. Zheng. 2004. Neuroanatomy and dentition of
Camarasaurus lentus; pp. 199–211 in V. Tidwell and K. Carpenter
(eds.), Thunder-Lizards: The Sauropodomorph Dinosaurs. Indiana
University Press, Bloomington, Indiana.
Curry Rogers, K., and C. A. Forster. 2004. The skull of Rapetosau-
rus krausei (Sauropoda: Titanosauria) from the Late Creta-
ceous of Madagascar. Journal of Vertebrate Paleontology
24:121–144.
D
ıaz, V. D., X. P. Suberbiola, and J. L. Sanz. 2011. Braincase anatomy of
titanosaurian sauropod Lirainosaurus astibiae from the Late Creta-
ceous of Iberian. Acta Palaeontologica Polonica 56:521–533.
Dong, Z. M. 1997. A gigantic sauropod (Hudiesaurus sinojapanorum,
gen. et sp. nov.) from the Turpan Basin, China; pp. 91–96 in Z.-M.
Dong (ed.) Sino-Japanese Silk Road Dinosaur Expedition. China
Ocean Press, Beijing, China. [Chinese, with English abstract]
Galton, P. M. 1984. Cranial anatomy of the prosauropod dinosaur Plateo-
saurus from the Knollenmergel (Middle Keuper, Upper Triassic) of
Germany. Geologica et Palaeontologica 18:139–171.
Galton, P. M., and F. Knoll. 2006. A saurischian dinosaur braincase from
the Middle Jurassic (Bathonian) near Oxford, England: from the
theropod Megalosaurus or the sauropod Cetiosaurus? Geological
Magazine 143:905–921.
Garcia, R. A., A. Paulina Carabajal, and L. Salgado. 2008. A new titano-
saurian braincase from the Allen Formation (Campanian–Maas-
trichtian), R
ıo Negro Province, Patagonia, Argentina. Geobios
41:625–633.
Gu, X., X. Liu, and Z.-F. Li. 1997. Stratigraphy of Sichuan Province.
China University of Geosciences Press, Wuhan, China, 417 pp.
Harris, J. D. 2006a. Cranial osteology of Suuwassea emilieae (Sauropoda:
Diplodocoidea: Flagellicaudata) from the Upper Jurassic Morrison
Formation of Montana, U.S.A. Journal of Vertebrate Paleontology
26:88–102.
Harris, J. D. 2006b. The signi�cance of Suuwassea emiliae (Dinosauria:
Sauropoda) for flagellicaudatan interrelationships and evolution.
Journal of Systematic Palaeontology 4:185–198.
He, X.-L., C. Li, and K.-J. Cai. 1988. The Middle Jurassic Dinosaur
Fauna from Dashampu, Zigong, Sichuan: Sauropod Dinosaurs, Vol-
ume 4, Omeisaurus tianfuensis. Sichuan Publishing House of Sci-
ence and Technology, Chengdu, 143 pp. [Chinese, with English
summary]
He, X., S. Yang, K. Cai, K. Li, and Z. Liu. 1996. [A new species of sauro-
pod, Mamenchisaurus anyuensis sp. nov.]; pp. 83–86 in Papers on
Geosciences Contributed to the 30th International Geological Con-
gress. Chengdu University of Technology, Chengdu, China. [Chinese]
Hopson, J. A. 1979. Paleoneurology; pp. 39–146 in C. Gans, R. G. North-
cutt, and P. Ulinski (eds.), Biology of the Reptilia, Volume 9. Aca-
demic Press, New York.
Huene, F. von. 1932. Die Fossile Reptil-ordnung Saurischia: Ihre
Entwicklung und Geschichte. Gebr€
uder Borntraeger. Monogra-
phien Geologie und Pal€
aontologie, Series 1 4:1–361.
Janensch, W. 1935. Die Sch€
adel der Sauropoden Brachiosaurus, Barosau-
rus und Dicraeosaurus aus den Tendaguru-Schichten Deutsch-Osta-
frikas. Palaeontographica Supplement to 7(1):147–298.
Kan, Z.-Z., B. Liang, Q.-W. Wang, and B. Zhu. 2005. Trace fossils in the
Suining and Penglaizhen Formations as buried places of dinosaur
fossils, in Anyue, Sichuan, and their environmental signi�cance.
Acta Geologica Sichuan 25:68–71.
Knoll, F., R. C. Ridgely, F. Ortega, J. L. Sanz, and L. M. Witmer. 2013.
Neurocranial osteology and neuroanatomy of a Late Cretaceous
titanosaurian sauropod from Spain (Amplelosaurus sp.). PLoS
ONE 8:e54991. doi: 10.1371/journal.pone.0054991.
Knoll, F., L. M. Witmer, F. Ortega, R. C. Ridgely, and D. Schwarz-
Wings. 2012. The braincase of the basal sauropod dinosaur Spino-
phorosaurus and 3D reconstructions of the cranial endocast and
inner ear. PLoS ONE 7:e30060. doi:10.1371/journal.pone.0030060
L€
ang, E., and F. Mahammed, 2010. New anatomical data and phyloge-
netic relationship of Chebsaurus algeriensis (Dinosauria, Sauro-
poda) from the Middle Jurassic of Algeria. Historical Biology
22:142–164.
Li, G., H. Hirano, T. Sakai, T. Kozai, and H. Ishiguro. 2010. Palae-
ontology and biostratigraphy of Jurassic clam shrimps of the
Sichuan Basin, China. Earth Science Frontiers 17, special issue
(Short Papers for the 8th International Congress on the Jurassic
System):34–36.
Li, K., C. Yang, J. Liu, and T. Jiang. 2010a. The vertebrate assemblages
from the Shaximiao Formation of the Sichuan Basin, China and dis-
cussion to its age. Earth Science Frontiers 17, supplement (Short
Papers for the 8th International Congress on the Jurassic Sys-
tem):123–124.
Li, K., C.-Y. Yang, J. Liu, and Z.-X. Wang. 2010b. A new sauropod from
the Lower Jurassic of Huili, Sichuan, China. Vertebrata PalAsiatica
48:185–202.
L€
u, J., T. Li, S. Zhong, Q. Ji, and S. Li. 2008. A new mamenchisaurid
dinosaur from the Middle Jurassic of Yuanmou, Yunnan Province,
China. Acta Geologica Sinica 82:17–26.
L€
u, J., T. Li, Q. Ji, G. Wang, J. Zhang, and Z. Dong. 2006. New eusauro-
pod dinosaur from Yuanmou of Yunnan Province. Acta Geologica
Sinica 80:1–10.
L€
u, J., L. Xu, H. Pu, X. Zhang, S. Jia, H. Chang, J. Zhang, and X. Wei.
2013. A new sauropod dinosaur (Dinosauria, Sauropoda) from the
late Early Cretaceous of the Ruyang Basin (central China). Creta-
ceous Research 44:202–213.
Mannion, P. D., P. Upchurch, M. T. Carrano, and P. M. Barrett. 2011.
Testing the effect of the rock record on diversity: a multidisciplinary
approach to elucidating the generic richness of sauropodomorph
dinosaurs through time. Biological Reviews 86:157–181.
Marsh, O. C. 1878. Principal characters of American Jurassic dinosaurs
(Part 1). American Journal of Science, Series 3 16:411–416.
Monbaron, M., D. A. Russell, and P. Taquet. 1999. Atlasaurus imerakei n.
g., n. sp., a brachiosaurid-like sauropod from the Middle Jurassic of
Morocco. Comptes Rendus de l’Acad
emie des Sciences, Paris: Sci-
ences de la Terre et des Plan
etes 329:519–526.
Nair, J. P., and S. W. Salisbury. 2012. New anatomical information on
Rhoetosaurus browni Longman, 1926, a gravisaurian sauropodo-
morph dinosaur from the Middle Jurassic of Queensland, Australia.
Journal of Vertebrate Paleontology 32:369–394.
Ouyang, H., and Y. Ye. 2002. The First Mamenchisaurian Skeleton with
Complete Skull: Mamenchisaurus youngi. Sichuan Science and
Technology Press, Chengdu, China, 111 pp. [Chinese, with English
summary]
Paulina Carabajal, A. 2012. Neuroanatomy of titanosaurid dinosaurs
from the Upper Cretaceous of Patagonia, with comments on endo-
cranial variability within Sauropoda. The Anatomical Record
295:2141–2156.
Paulina Carabajal, A., and L. Salgado. 2007. Un basicr
aneo de titano-
saurio (Dinosauria, Sauropoda) del Cret
acico Superior del norte de
Patagonia: descripci
on y aportes al conocimiento del o
ıdo interno
de los dinosaurios. Ameghiniana 44:109–120.
Peng, G.-Z., Y. Ye, and Y. Gao. 2005. Jurassic Dinosaur Faunas in
Zigong. People’s Publishing House of Sichuan Province, Chengdu,
China, 236 pp.
Poropat, S. F., and B. P. Kear. 2013. Photographic atlas and three-dimen-
sional reconstruction of the holotype skull of Euhelopus zdanskyi
with description of additional cranial elements. PLoS ONE 8:
e79932. DOI: 10.1371/journal.pone.0079932
Remes, K. 2006. Revision of the Tendaguru sauropod dinosaur Tornieria
africana (Fraas) and its relevance for sauropod paleobiogeography.
Journal of Vertebrate Paleontology 26:651–669.
Remes, K., F. Ortega, I. Fierro, U. Joger, R. Kosma, J. M. Marin Ferrer,
Project Paldes, Project SNHM, O. A. Ide, and A. Maga. 2009. A
new basal sauropod dinosaur from the Middle Jurassic of Niger and
the early evolution of Sauropoda. PLoS ONE 4:e6924. doi: 10.1371/
journal.pone.0006924.
Royo-Torres, R., and P. Upchurch. 2012. The cranial anatomy of the sau-
ropod Turiasaurus riodevensis and implications for its phylogenetic
relationships. Journal of Systematic Palaeontology 10:553–583.
Royo-Torres, R., A. Cobos, and L. Alcal
a. 2006. A giant European dino-
saur and a new sauropod clade. Science 314:1925–1927.
Royo-Torres, R., A. Cobos, L. Luque, A. Aberasturi, E. Esp
ılez, I.
Fierro, A. Gonz
alez, L. Mampel, and L. Alcal
a. 2009. High Euro-
pean sauropod dinosaur diversity during Jurassic-Cretaceous transi-
tion in Riodeva (Teruel, Spain). Palaeontology 52:1009–1027.
Russell, D. A. 1993. The role of Central Asia in dinosaur biogeography.
Canadian Journal of Earth Sciences 30:2002–2012.
e889701-16 JOURNAL OF VERTEBRATE PALEONTOLOGY
Downloaded by [Kangwon National University - Samcheok Campus] at 15:51 19 February 2015
Russell, D. A., and Z. Zheng. 1993. A large mamenchisaurid from the
Junggar Basin, Xinjiang, People Republic of China. Canadian Jour-
nal of Earth Sciences 30:2082–2095.
Salgado, L., and J. Bonaparte. 1991. Un nuevo saur
opodo Dicraeosauri-
dae, Amargasaurus cazaui gen. et sp. nov. de la Formacion La
Amarga, Neocomiano de la Provincia del Neuquen, Argentina.
Ameghiniana 28:333–346.
Sanz, J. L., J. E. Powell, J. Le Loeuff, R. Martinez, and X. Pereda-Suber-
biola. 1999. Sauropod remains from the Upper Cretaceous of La~
no
(north central Spain). Titanosaur phylogenetic relationships. Estu-
dios del Museo de Ciencias Naturales de Alava 14:235–255.
Seeley, H. G. 1887. On the classi�cation of the fossil animals commonly
named Dinosauria. Proceedings of the Royal Society of London
43:165–171.
Sekiya, T. 2011. Re-examination of Chuanjiesaurus anaensis (Dinosauria:
Sauropoda) from the Middle Jurassic Chuanjie Formation, Lufeng
County, Yunnan Province, southwest China. Memoir of the Fukui
Prefectural Dinosaur Museum 10:1–54.
Sereno, P. C., J. A. Wilson, L. M. Witmer, J. A. Whitlock, A. Maga, O.
Ide, and T. A. Rowe. 2007. Structural extremes in a Cretaceous
dinosaur. PLoS ONE 2:e1230. doi: 10.1371/journal.pone.0003303.
Sereno, P. C., A. L. Beck, D. B. Dutheil, H. C. E. Larsson, G. H. Lyon, B.
Moussa, R. W. Sadleir, C. A. Sidor, D. J. Varricchio, G. P. Wilson,
and J. A. Wilson. 1999. Cretaceous sauropods from the Sahara and
the uneven rate of skeletal evolution among dinosaurs. Science
286:1342–1347.
Suteethorn, S., J. Le Loeuff, E. Buffetaut, V. Suteethorn, and K.
Wongko. 2013. First evidence of a mamenchisaurid dinosaur from
the Late Jurassic/Early Cretaceous Phu Kradung Formation of
Thailand. Acta Palaeontologica Polonica 58:459–469.
Swofford, D.L. 2003. PAUP*: Phylogenetic Analysis Using Parsimony,
version 4.0b10. Sinaeur Associates, Sunderland, Massachusetts.
Tang, F., X. Jing, X. Kang, and G. Zhang. 2001. Omeisaurus maoianus:A
Complete Sauropod from Jingyuan, Sichuan. China Ocean Press,
Beijing, China, 128 pp. [Chinese, with English summary]
Taylor, M. P. 2009. A re-evaluation of Brachiosaurus altithorax Riggs
1903 (Dinosauria, Sauropoda) and its generic separation from
Giraffatitan brancai (Janensch 1914). Journal of Vertebrate Paleon-
tology 29:787–806.
Tidwell, V., and K. Carpenter. 2003 Braincase of an Early Cretaceous
titanosauriform sauropod from Texas. Journal of Vertebrate Pale-
ontology 23:176–180.
Upchurch, P. 1995. The evolutionary history of sauropod dinosaurs. Phil-
osophical Transactions of the Royal Society of London, Series B
349:365–390.
Upchurch, P., and P. D. Mannion. 2009. The �rst diplodocid from Asia
and its implications for the evolutionary history of sauropod dino-
saurs. Palaeontology 52:1195–1207.
Upchurch, P., P. M. Barrett, and P. Dodson. 2004. Sauropoda; pp. 259–
322 in D. B. Weishampel, H. Osm
olska, and P. Dodson (eds.), The
Dinosauria, second edition. University of California Press, Berke-
ley, California.
Wang, Y., B. Fu, X. Xie, G. Li, Q. Huang, X. Yang, Y. Pan, N. Tian, and
Z. Jiang. 2010. The non-marine Triassic and Jurassic system in the
Sichuan Basin, stratigraphic sequences, biodiversity and major
strato-boundaries. Earth Science Frontiers 17, special issue (Short
Papers for the 8th International Congress on the Jurassic Sys-
tem):19–21.
Whitlock, J. A., M. D. D’Emic, and J. A. Wilson. 2011. Cretaceous dip-
lodocids in Asia? Re-evaluating the phylogenetic af�nities of a frag-
mentary specimen. Palaeontology 54:351–364.
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. 2005. Redescription of the Mongolian sauropod Nemegto-
saurus mongoliensis Nowinski (Dinosauria: Saurischia) and com-
ments on Late Cretaceous sauropod diversity. Journal of Systematic
Palaeontology 3:283–318.
Wilson J. A., and P. Upchurch. 2009. Redescription and reassessment of
the phylogenetic af�nities of Euhelopus zdanskyi (Dinosauria: Sau-
ropoda) from the Early Cretaceous of China. Journal of Systematic
Palaeontology 7:199–239.
Wilson, J. A., M. S. Malkani, and P. D. Gingerich. 2005. A sauropod
braincase from the Pab Formation (Upper Cretaceous, Maastrich-
tian) of Balochistan, Pakistan. Gondwana Geological Magazine
8:101–109.
Wilson, J. A., M. D. D’Emic, T. Ikejiri, E. M. Moacdieh, and J. A. Whit-
lock. 2011. A nomenclature for vertebral fossae in sauropods and
other saurischian dinosaurs. PLoS ONE 6:e17114. doi: 10.1371/jour-
nal.pone.0017114.
Wings, O., D. Schwarz-Wings, and D.W. Fowler. 2011. New sauropod
material from the Late Jurassic part of the Shishugou Formation
(Junggar Basin, Xinjiang, NW China). Neues Jahrbuch f€
ur Geologie
und Pal€
aontologie, Abhandlungen 262:129–150.
Wings, O., M. Rabi, J. Schneider, L. Schwermann, G. Sun, C.-F. Zhou,
and W. Joyce. 2012. An enormous Jurassic turtle bone bed from
the Turpan Basin of Xinjiang, China. Naturwissenschaften 99:925–
935.
Witmer, L. M., R. C. Ridgely, D. L. Dufeau, and M. C. Semones. 2008.
Using CT to peer into the past: 3D visualization of the brain and ear
regions of birds, crocodiles, and nonavian dinosaurs; pp. 67–87 in H.
Endo and R. Frey (eds.), Anatomical Imaging: Towards a New
Morphology. Springer, Tokyo.
Wu, W.-H., C.-F. Zhou, O. Wings, T. Sekiya, and Z.-M. Dong. 2013. A
new gigantic sauropod dinosaur from the Middle Jurassic of Shan-
shan, Xinjiang. Global Geology 32:437–446.
Xie, X., Y. Wang, and B. Fu. 2010. Sedimentary facies of the Upper
Jurassic Suining and Penglaizhen formations in the central Sichuan
Basin, SW China. Earth Science Frontiers 17, special issue (Short
Papers for the 8th International Congress on the Jurassic Sys-
tem):313–314.
Xing, L.-D., T. Miyashita, P. J. Currie, H.-L. You, and Z.-M. Dong. 2013.
A new basal eusauropod from the Middle Jurassic of Yunnan Prov-
ince, People’s Republic of China. Acta Palaeontologica Polonica.
doi: 10.4202/app.2012.0151.
Young, C.-C. 1958. New sauropods from China. Vertebrata PalAsiatica
2:1–29.
Young, C.-C., and X.-J. Zhao. 1972. Mamenchisaurus hochuanensis.
Institute of Vertebrate Paleontology and Paleoanthropology Mono-
graph Series I 8:1–30. [Chinese]
Zhang, Y.-G., and J.-J. Li. 2003. Stratigraphy of the Mamenchisaurus
fauna in Jingyan, Sichuan. Journal of Stratigraphy 27:50–53.
Submitted May 19, 2013; revisions received December 20, 2013; accepted
January 25, 2014.
Handling editor: You Hailu.
JOURNAL OF VERTEBRATE PALEONTOLOGY e889701-17
Downloaded by [Kangwon National University - Samcheok Campus] at 15:51 19 February 2015
