A new long-tailed basal bird from the Lower Cretaceous of north-eastern China

Abstract

A new basal Avialae, Jeholornis curvipes sp. nov., from the Yixian Formation (Lower Cretaceous) of Liaoning Province (north-eastern China) is described. A revision of long-tailed birds from China and a phylogenetic analysis of basal Avialae suggest that Jeholornithiformes were paraphyletic, with Jixiangornis orientalis being the sister-taxon of pygostylia. The phylogenetic analysis also recovered that the tail reduction is a unique event in the evolution of birds. Jeholornis species were cursorial, nonperching, and seed-eating birds. © 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 790–804.

Full text

A new long-tailed basal bird from the Lower Cretaceous
of north-eastern China
ULYSSE LEFÈVRE1,2*, DONGYU HU3,4, FRANÇOIS ESCUILLIÉ5, GARETH DYKE6and
PASCAL GODEFROIT7
1Department of Geology, University of Liège, allée du 6 Août B18-B20, 4000 Liège, Belgium
2Department of Paleontology, Royal Belgian Institute of Natural Sciences, rue Vautier 29, 1000
Bruxelles, Belgium
3Paleontological Institute, Shenyang Normal University, 253 North Huanghe Street, Shenyang
110034, China
4Key Laboratory of Vegetation Ecology, Ministry of Education, Northeast Normal University, 5268
Renmin Street, Changchun 130024, China
5Eldonia, 9 avenue des Portes Occitanes, 3800 Gannat, France
6Ocean and Earth Science, National Oceanography Centre, University of Southampton, Southampton
SO14 3ZH, UK
7Department of Paleontology, Royal Belgian Institute of Natural Sciences, rue Vautier 29, 1000
Bruxelles, Belgium
Received 18 February 2014; revised 30 April 2014; accepted for publication 30 April 2014; published 17 October
2014
A new basal Avialae, Jeholornis curvipes sp. nov., from the Yixian Formation (Lower Cretaceous) of Liaoning
Province (north-eastern China) is described. A revision of long-tailed birds from China and a phylogenetic analysis
of basal Avialae suggest that Jeholornithiformes were paraphyletic, with Jixiangornis orientalis being the
sister-taxon of pygostylia. The phylogenetic analysis also recovered that the tail reduction is a unique event in the
evolution of birds. Jeholornis species were cursorial, nonperching, and seed-eating birds. © 2014 The Linnean
Society of London, Biological Journal of the Linnean Society, 2014, 113, 790–804.
ADDITIONAL KEYWORDS: Avialae – Jehol group – Jeholornis – Jeholornithiformes – systematic revision.
INTRODUCTION
For more than 150 years, Archaeopteryx lithographica
remained the only known Mesozoic bird with a long
bony tail. Since 1992, however, numerous fairly
complete skeletons of long-tailed Mesozoic birds
have been discovered, mainly from the Jehol Biota of
north-eastern China (Chang et al., 2008). Indeed,
several clades of basal Avialae are known only from
this part of the globe (O’Connor et al., 2012).
The first bird specimen from the Jehol biota was
described in 1992 as Sinornis santensis (Sereno &
Rao, 1992). Ten years later, Zhou & Zhang (2002)
described Jeholornis prima, a basal avialan from the
Jiufotang Formation, which is characterized in par-
ticular by a long bony tail of 22 caudal vertebrae
as in A. lithographica. At around the same time,
Shenzhouraptor sinensis and Jixiangornis orientalis
were also described: the first from the Jiufotang For-
mation and the latter from the Yixian Formation.
Jeholornithiformes as a group was coined in 2006
(Zhou & Zhang, 2006) to encompass all of the long-
tailed birds that are also characterized by a convex
rostroventral margin and a concave ventral margin
of the dentary (O’Connor et al., 2012). This group
now includes J. prima,S. sinensis,J. orientalis,
Dalianraptor cuhe, and Jeholornis palmapenis (Zhou
& Zhang, 2002; Ji et al., 2002a, b; Gao & Liu, 2005;
O’Connor et al., 2012). Indeed, some studies have
suggested that S. sinensis and J. orientalis might be
*Corresponding author. E-mail: lefevre.u@gmail.com
ZooBank Registration: urn:lsid:zoobank.org:act:882FACA6-
CB48-4D5D-ADEE-6E0DAD031101
bs_bs_banner
Biological Journal of the Linnean Society, 2014, 113, 790–804. With 10 figures
© 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 790–804790

synonyms of J. prima because they are based on
juvenile specimens (Zhou & Zhang, 2006; Ji et al.,
2003; Li et al., 2010; Zhou & Wang, 2010).
In the present study, we describe a new specimen of
a long-tailed bird from the Yixian Formation placed
within Jeholornithiformes. The new specimen pro-
vides new information on the anatomy, systematics
and phylogeny of this important group of basal
birds.
MATERIAL AND METHODS
GEOLOGY
The Jehol biota is formed from two formations: the
Yixian Formation, dated 124.6 ± 0.3 Mya using U-Pb
SHRIMP dating (Yang, Li & Jiang, 2007), and the
younger Jiufotang Formation, dated to 120 ± 0.7 Mya
using 40Ar/39Ar decay (Chang et al., 2008). These for-
mations are comprised of lacustrine deposits embed-
ded with volcanic rocks (Jiang et al., 2011) and
contain the most diversified avifauna known to date
closely related to A. lithographica (O’Connor et al.,
2012).
ANATOMY
Our use of anatomical nomenclature is applied in
accordance with Baumel et al. (1993). Although the
Latin terminology used by Baumel et al. (1993) is
retained for muscles, osteological structures are
described using the English equivalents of Latin
terms (sensu Howard, 1929) as is standard in descrip-
tions of non-neornithine fossil birds.
INSITUTIONAL ABBREVIATIONS
IVPP, Institute of Vertebrate Paleontology and
Paleoanthropology, Beijing, China; SDM, Shandong
Museum, Jinan, China; YFGP, Yizhou Fossil and
Geology Park, Yizhou, China.
SYSTEMATIC PALAEONTOLOGY
DINOSAURIA OWEN, 1842
THEROPODA MARSH, 1881 SENSU GAUTHIER, 1986
MANIRAPTORA GAUTHIER, 1986
AVIALAE GAUTHIER, 1986
JEHOLORNIS ZHOU AND ZHANG, 2002
Type species: Jeholornis prima Zhou and Zhang, 2002
Referred species: Jeholornis palmapenis O’Connor,
Sun, Xu, Wang and Zhou, 2012; Jeholornis curvipes
sp. nov.
Revised diagnosis: Pheasant-sized early diverging
avialans that share: (i) premaxilla edentulous; (ii) a
concave ventral margin and a convex rostro-ventral
margin of the dentary; (iii) a straight tibia with
a ratio tibia:femur between 1.19 and 1.24; (iv) a
nonretroverted caudomedially oriented hallux; (v) a
‘J’-shaped first metatarsal; (vi) third metatarsal that
is the longest; (vii) penultimate pedal phalanxes not
longer than terminal ones; and (viii) each caudal
vertebra three or four times as long as wide.
JEHOLORNIS CURVIPES SP.NOV.(FIG.1)
Holotype: YFGP-yb2, an almost completely articu-
lated skeleton including a partial caudal series
(Fig. 1).
Type locality and horizon: Yizhou Fossil and Geology
Park, Liaoning Province, (north-eastern China)
Dakangpu Member (equivalent to the Dawangzhangzi
Beds) of the Yixian Formation, lower Aptian (Lower
Cretaceous) (Swisher et al., 1999; Yang et al., 2007;
Chang et al., 2009)
Etymology: From the Latin curvus (curved) and pes
(foot), in reference to the lateral deviation of the distal
part of the metatarsals in YFGP-yb2.
Diagnosis: This new species of basal bird is charac-
terized by the following features: (i) a dentary
without an ossified symphysis; (ii) a strut-like cora-
coid with a lateral process not detached from the
lateral margin; (iii) medial margin of coracoid
concave along its entire length; (iv) deltopectoral
crest of humerus slightly deflected from the shaft;
(v) ratio between metacarpal II and humerus length
0.58 (see Supporting information, Fig. S1); (vi) ratio
between metacarpal I and metacarpal II 0.17; (vii)
ratio manual phalanx I-1:phalanx II-1 1.29; (viii) the
presence of two prominent proximal condyles on the
caudal end of the tibia; and (ix) distal half of meta-
tarsals laterally deviated.
Description and comparisons: YFGP-yb2 is an almost
complete and articulated skeleton that just lacks the
caudal region of the tail (Fig. 1). Also of note, however,
is that the rostral part of the skull, the autopodal
segment of the right forelimb, and the distal
autopodal segment of the right hindlimb were recon-
structed during preparation of the specimen. We have
confirmed this with X-ray radiographs and so these
parts of the skeleton are not described (Fig. 2). X-ray
radiography and computed tomography (CT) scans
also confirm that YFGP-yb2 is not a chimera in the
region of the pelvis as suspected at first examination
(see Supporting information, Fig. S2).
A NEW LONG-TAILED BASAL BIRD 791
© 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 790–804

Ontogenic stage: The following anatomical features
suggest that YFGP-yb2, the holotype of J. curvipes,
was an adult at death (Forster, 1998; Zhou & Zhang,
2003a; Zhou and Zhang, 2003b; Xu & Norell, 2004;
Turner et al., 2007; Gao et al., 2012; Godefroit et al.,
2013a): (i) the texture of the bones on the proximal
and distal ends of the coracoid, the humerus, the
ulna, the pubis, and the femur is regular and con-
tinuous; (ii) the frontals are fused together along their
sagittal line; (iii) the neural arches of the caudal
cervical vertebrae are completely fused to correspond-
ing centra (this suture is not visible in X-rays); (iv) CT
scans show that the sacral vertebrae are all fused
together; (v) metacarpals II and III are proximally
fused; and (v) metatarsals II, III, and IV are proxi-
mally fused.
Figure 1. Photograph, drawing and X-ray radiographs of the basal bird Jeholornis curvipes sp. nov., from the Lower
Cretaceous of north-eastern China. A, YFGP-yb2 photograph. B, Line drawing; grey parts represent restored skeletal
regions.
792 U. LEFÈVRE ET AL.
© 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 790–804

Skull: Because the antorbital region is incompletely
preserved, it is impossible to check whether this
region is rostrocaudally shorter than dorsoventrally
high, as in Xiaotingia zhengi and other Maniraptora
including A. lithographica (Xu et al., 2011) (Fig. 3).
No premaxillary or maxillary fenestrae are preserved
as a result of crushing. The frontals are fused
together along their whole length and, as in
J. palmapenis (O’Connor et al., 2012), the left frontal
is petal-shaped in dorsal view. The frontals are also
convex and wider caudally. There is no thickening of
the orbital rim, unlike in Confuciusornis sanctus,
where this thickening is interpreted as the fusion
between the palpebral and the frontal (Chiappe et al.,
1999). The frontals are perpendicularly fused to the
parietals, as described for C. sanctus (Chiappe et al.,
1999). Unlike Mei long, the interfrontal suture is
not separated from the parietals by a triangular
notch (Gao et al., 2012). The symphysial region of the
dentaries is not ossified in contrast to the condition
in J. prima. The dentaries are robust, as in other
members of Jeholornithiformes, with a convex
rostroventral margin and a concave ventral margin.
The ventral process of the dentary is located under
the rostral part of the surangular, whereas the dorsal
process is absent. A wide groove extends along the
lateral side of the dentary and does not contain any
foramina, unlike in Troodontidae and A. lithographica
(Weishampel, Dodson & Osmólska, 2007; Wellnhofer,
2009). The angular is rostrally thin and becomes
wider caudally; the dorsal margin of its rostral end
takes part in the ventral edge of the external man-
dibular fenestra. Dorsal to the angular, there is a
wide surangular. The dorsal edge of the surangular is
convex and has a groove in its caudal region, which is
not prolonged into the rostral region of the same bone.
Binocular observations and X-ray radiographs reveal
that the jaws are edentulous along their whole pre-
served length.
Vertebrae: The cervical region of YFGP-yb2 is poorly
preserved, making distinction between vertebrae dif-
ficult. Although X-ray radiographs allow us to distin-
guish four unequivocal cervical vertebrae, others are
trapped under the scapular region and distinction
between the dorsal and the cervical series remains
impossible. The most anterior cervical vertebra is
87% craniocaudally shorter than the most caudal
cervical vertebrae. The prezygapophyses form a
craniomedially oriented articular surface. A cervical
rib can be observed in articulation with the second
cervical vertebra. This rib is 11.7 mm long, compris-
ing 109% of the length of the corresponding centrum.
X-ray radiographs also reveal eight unequivocal
dorsal vertebrae. The anterior part of the dorsal
region, however, is crushed under the scapular region
and so the most caudal part cannot be distinguished
from the sacral region even by X-ray radiographs and
CT scans. There is also no size difference between the
Figure 2. X-ray radiography of YFGP-yb2. Arrows point
to reconstructed regions of the specimen (rostral part of
the skull, distal part of the left forelimb, and distal part of
the right hindlimb). Scale bar = 40 mm.
Figure 3. Skull of Jeholornis curvipes sp. nov., from
the Lower Cretaceous of north-eastern China. a, angular;
d, dentaries; f, frontals; p, parietals; s, surangular; sy,
symphyseal region; vp, ventral process of the dentary.
Scale bar = 40 mm.
A NEW LONG-TAILED BASAL BIRD 793
© 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 790–804

eight unequivocal dorsal vertebrae. X-ray radiographs
reveal trapezoidal neural spines with dorsal margins
that are only slightly shorter than corresponding
centra. Neural arches are all completely fused to
corresponding centra, which suggests that this speci-
men is adult (as discussed above).
The sacral region cannot be observed directly
because of its poor state of preservation. However, one
transverse process that is located caudally to the left
femur is as wide as in J. prima. This process has a
straight lateral margin that forms an angle of less
than 45° with the corresponding caudal margin.
Only the first nine caudal vertebrae are preserved.
Within this series, proximal centra are 57% shorter
than distal ones and the transition point is located
after the second caudal vertebrae, as in J. prima
(Zhou & Zhang, 2002). The neural spines and trans-
verse processes are not visible. Both the proximal
and distal articular surfaces of the caudal vertebra
are slightly convex. The zygapophyses become longer
proximodistally, although their length remains within
30% of the length of the corresponding vertebra. CT
scans show that the chevrons of the proximal caudal
vertebrae are rectangular and dorsoventrally higher
than proximodistally long as in J. palmapenis. They
become anterocaudally as long as the corresponding
centrum in the distal series. As in A. lithographica,
Anchiornis huxleyi, and Aurornis xui, a groove
extends along the junction between the neural arches
and the centrum (Hu et al., 2009; Godefroit et al.,
2013a, 2013b).
Scapular girdle: The scapula is long, thin and ribbon-
shaped (Fig. 4). Its distal end is thin and round, as
opposed to being sharp as in modern birds (Baumel
et al., 1993; Chiappe et al., 1999; Wellnhofer, 2009). It
is ventrally curved and its thickness decreases cau-
dally as in J. prima and S. sinensis (Ji et al., 2002a;
Zhou & Zhang, 2002). Its ventral margin is distinctly
thicker than its dorsal margin. There is no distal
groove near the glenoid fossa at the ventral side of the
scapula, as is the case in X. zhengi (Xu et al., 2011).
A prominent acromion process is present on the
proximal end of the scapula as in A. lithographica,
S. sinensis, and Rahonavis ostromi (Forster, 1998;
Wellnhofer, 2009; Ji et al., 2003). The glenoid fossa is
located caudally to the acromion process and is
laterodorsally oriented as in C. sanctus (Chiappe
et al., 1999; Zhou & Zhang, 2003a). The scapula is
significantly shorter (66%) than the humerus, as in
derived Maniraptora and basal Avialae (Zhang et al.,
2008) and lies parallel with the dorsal vertebrae as in
A. lithographica (Wellnhofer, 2009) and M. long (Gao
et al., 2012).
X-ray radiographs reveal a suture between the
scapula and the coracoid, indicating that those bones
were not completely fused (Fig. 2). The two coracoids
contact their respective scapulae at an angle of less
than 90°, as in Zhongjianornis yangi and Sapeornis
chaoyangensis (Zhou & Zhang, 2003b; Zhou & Li,
2010). The left coracoid is strut-like as in J. prima,
Z. yangi, and S. chaoyangensis (Wellnhofer, 2009)
and is approximately half the length of the scapula
(Fig. 5). Its sternal margin is sub-horizontal as
opposed to being convex as in J. prima (Zhou &
Zhang, 2002). Its medial margin is clearly concave,
whereas it is slightly convex in J. prima. The lateral
process is not as well developed as in J. prima,in
which this process is clearly detached from the lateral
margin of the coracoid. The procoracoid process is also
Figure 4. Scapular girdle of Jeholornis curvipes sp.
nov., from the Lower Cretaceous of north-eastern China.
ap, acromial process of scapula; gfc, glenoid fossa of cora-
coid; gfs, glenoid fossa of scapula; ls, left scapula; rs, right
scapula. Scale bar = 40 mm.
Figure 5. Left coracoid of Jeholornis curvipes sp. nov.,
from the Lower Cretaceous of north-eastern China. ap,
acrocoracoid process; lp, lateral process; mm, medial
margin; pro, procoracoid process; sm, sternal margin.
Scale bar = 40 mm.
794 U. LEFÈVRE ET AL.
© 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 790–804

poorly developed at the junction between the coracoid
and the scapula. A hemispherical acrocoracoid process
is prominently developed. There is no subglenoid
fossa on the proximal region of the coracoid, unlike
the condition in X. zhengi and Eosinopteryx brevi-
penna (Godefroit et al., 2013b).
The furcula is visible completely on X-ray radio-
graphs. It is robust and boomerang-shaped as in
J. prima,A. lithographica,C. sanctus,X. zhengi, and
Dromaeosauridae (Zhou & Zhang, 2002; Xu et al.,
2011). The furcula branches are less than one-third of
the femur length. There is no hypocleideum at the
junction between the two branches, as also seen
in J. prima,S. sinensis,C. sanctus, and A. huxleyi
(Chiappe et al., 1999; Norell & Clarke, 2001; Zhou &
Zhang, 2002; Ji et al., 2003). The intraclavicular angle
is 62° and the furcula branches are craniocaudally
compressed.
The ribs of YFGP-yb2are long and thin as in
many basal birds (e.g. A. lithographica,J. prima,
S. sinensis, and S. chaoyangensis) and there are no
traces of uncinate processes. Elements of gastralia
are present near the pelvic region, proximally
expanded and tapered distally.
A large brownish element is also present in the
sternal region of the skeleton. To determine whether
this bone is the remains of the sternum, we performed
a scanning electron microscopy (SEM) analysis
coupled with energy dispersive spectrometry (EDS).
EDS results show that the Ca/P ratio in this region
varies between 1.5 and 2 (see Supporting information,
Fig. S3), confirming that this element is formed of
apatite (Newesely, 1989; Hubert et al., 1996). SEM
analysis further reveals the presence of concentric
lamellar structures, similar to those seen in fresh
bones. A Haversian canal and a Volkmann’s canal are
preserved together with lenticular pores, which lie
parallel to the concentric lamellae; these could be the
remains of fossilized osteoblasts (Fig. 6A) (Gartner &
Hiatt, 2004). Close-up images reveal the presence of
mineralized collagen fibres inside a Haversian canal,
as also observed in Iguanodon bones (Leduc, 2012).
It is therefore likely that this brownish element rep-
resents the ossified sternum, rarely preserved in
basal birds (e.g. A. lithographica and S. chaoyan-
gensis) (Zhou & Zhang, 2003b). Unfortunately, the
poor preservation of this bone prevents further com-
parisons with other specimens.
Forelimbs: The ratio between forelimb length
(humerus + ulna + metacarpal II) and hindlimb
length (femur + tibia + metatarsal III) is approxi-
mately 1.22, which is similar to the condition
observed in J. prima.
The humerus is approximately 120% the length of
the femur and is more robust. The diaphysis is
straight in cranial view (Fig. 7). The deltopectoral
crest is wide (117% of the width of the humerus at
mid-shaft) and extends along 40% of the humerus
length as in J. prima and S. sinensis (Ji et al., 2003).
This crest is sub-rectangular as in J. prima,
S. sinensis, and S. chaoyangensis. Its dorsal margin is
straight as in S. chaoyangensis. Its dorsodistal
portion does not form an acute angle as in J. prima.
The humeral head, the bicipital groove, the ventral
tubercle, and the pneumotricipital fossa are not
visible, even on X-ray radiographs. However, X-ray
radiographs clearly reveal the absence of an extended
bicipital crest. The distal condyles of the humerus are
mainly located on the cranial face of the bone as in
S. chaoyangensis and advanced birds (Zhou & Zhang,
2003b). The ventral condyle is prominent and ball-
shaped, whereas the dorsal condyle is smaller and
sub-oval. It is impossible to determine whether these
Figure 6. Histological elements from the sternum of
Jeholornis curvipes sp. nov. A, general structure of the
bone. B, close-up of mineralized collagen fibres inside a
Haversian canal.
A NEW LONG-TAILED BASAL BIRD 795
© 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 790–804

two condyles are caudally separated by the olecranon
as in M. long (Gao et al., 2012). Proximally to these
two condyles lies a slight depression, which can be
interpreted as the coronoid fossa for insertion
of the brachialis muscle. The entepicondyle and
ectepicondyle are incompletely preserved. Unlike
C. sanctus, the proximal part of the humerus is not
fenestrated (Chiappe et al., 1999; Zhou & Zhang,
2003a).
The ulna is slightly longer than the humerus
(106% of the humeral length) as in S. sinensis and
S. chaoyangensis, whereas it is exactly the same
length in J. prima (Zhou & Zhang, 2002; Ji
et al., 2002a; Pu et al., 2013). The diaphysis is
slightly curved dorsally as in J. prima,S. sinensis,
Yandangornis longicaudus,M. long,Zhongjianornis
elegans,R. ostromi, and Epidexypteryx hui (Forster,
1998; Cai & Zhao, 1999; Zhou & Zhang, 2002; Ji et al.,
2002a; Zhang et al., 2008; Zhou & Li, 2010; Gao et al.,
2012). The proximal portion of the right ulna bears a
flat and oval surface probably for the insertion of the
brachialis muscle.
The radius is slightly shorter (97%) than the
humerus. The diaphysis of the left radius is expanded
as in C. sanctus, although it is impossible to know if
this expansion forms a rectangular transverse surface
for articulation with the radial. Although a groove
extends along the total length of the radius, this is
interpreted as the result of crushing and so cannot be
compared with the similar and characteristic groove
present in enanthiornithines (Chiappe et al., 1999).
The manus (here, the combined length of metacar-
pal II + phalanx II-1 + phalanx II-2 + ungual II) is
longer than the humerus (123%). Metacarpal I is
rectangular and is not fused to metacarpal II as in
C. sanctus (Chiappe et al., 1999), whereas metacarpal
I is extremely short (1/16 of the metacarpal II length)
as in J. prima. Metacarpal II is straight and as wide
as metacarpal III, whereas the proximal end of meta-
carpal II terminates at the same level as metacarpal
III, unlike in C. sanctus and S. chaoyangensis
(Chiappe et al., 1999; Pu et al., 2013). Metacarpals II
and III are fused proximally and form a prominent
carpal trochlea as in J. prima,C. sanctus,S. chao-
yangensis, and Z. elegans (Cai & Zhao, 1999; Chiappe
et al., 1999; Zhou & Zhang, 2003b; Zhou & Li, 2010).
Metacarpals II and III are also fused distally and
form a ginglymoid articular surface as in J. prima
and M. long (Zhou & Zhang, 2002; Gao et al., 2012).
X-ray radiographs confirm the total fusion of both the
proximal and distal ends of metacarpals II and III,
with a complete absence of a suture line between
these bones. Metacarpal III is bowed as in J. prima
and S. sinensis (Zhou & Zhang, 2002; Ji et al., 2002a)
and the interosseous space is as wide as the ulna at
midshaft.
The phalangeal formula of YFGP-yb2 is 2-3-3/4-
X-X. The penultimate phalanx and the ungual of
digit III are only discernable on X-ray radiographs.
Thus, it is impossible to determine the exact number
of phalanxes for this finger. Phalanx I-1 does
not reach the distal end of metacarpal II as in
S. sinensis,J. prima,S. chaoyangensis,C. sanctus,
enantiornithines, and ornithurines (Zhou & Zhang,
2003a, b; Ji et al., 2003). This phalanx is straight
and thin (one-third of the width of phalanx II-1).
Ungual I is slightly shorter than ungual II, unlike
in S. sinensis. Phalanx II-1 is the most robust
phalanx of the manus as in J. prima,S. sinensis, and
S. chaoyangensis (Zhou & Zhang, 2002; Ji et al.,
2002a; Pu et al., 2013). The length of phalanx II-1 is
approximately half the length of phalanx II-2, unlike
in S. sinensis where phalanx II-1 is approximately the
same length as the phalanx II-2 (Ji et al., 2002a).
Phalanx II-2 is as long as phalanx II-1 but three
times thinner than the latter, whereas the width of
phalanx II-2 is constant along its length. Ungual II is
the largest in the manus. It is not more curved than
unguals I and III. It is impossible to determine
whether phalanges III-1 and III-2 formed an angle as
in S. sinensis (Ji et al., 2003). Ungual III is the small-
est and is only visible under X-ray. No flexor tubercle
can be seen and a deep pit for the insertion of the
collateral ligament is present at the level of the distal
end of phalanx II-2.
Pelvic girdle: This region has been highly crushed
during fossilization (Fig. 8). The right ilium is 44% the
length of the femur and its dorsal margin is broken
off. CT scans show that the dorsal margin of the left
ilium was reconstructed during fossil preparation:
slices clearly show that compact bone surrounding the
Figure 7. Right forelimb of Jeholornis curvipes sp.
nov., from the Lower Cretaceous of north-eastern China.
dc, deltopectoral crest; h, humerus; mc, metacarpal; r,
radius; u, ulna. Scale bar = 40 mm.
796 U. LEFÈVRE ET AL.
© 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 790–804

spongy bone is missing (see Supporting information,
Fig. S4). 3D reconstruction of the pelvic girdle of
YFGP-yb2 shows that the distal part of the dorsal
margin is regularly convex. CT scans also show that
the preacetabular process is partly reconstructed.
There is no trace of a brevis fossa as seen in
C. sanctus. The left pubis is twisted and broken off.
CT scans show that the right pubis is slightly
sigmoidal and terminates into a spoon-shaped pubic
foot with a posterior process as long as the distal
width of the femur. The anterior process of the pubic
foot is absent. The 3D reconstruction also shows that
the pubic symphysis was not ossified. It is impossible
to determine how far the pubic symphysis extended
as a result of displacement of the two bones. Only the
left ischium is preserved. Its distal proximal part is
badly defined, making the measurement of the total
length impossible. Its distal part has no ventrodistal
and dorsodistal processes. Only the intermediate
process is located at mid-shaft and does not project as
far caudally as in A. lithographica (Wellnhofer, 2009).
Hindlimbs: The left hindlimb is well-preserved and
covers the right one (Fig. 9). The femur is 20% shorter
than the tibia and its proximal half is slightly bowed
laterally as in S. chaoyangensis,C. sanctus,A. xui,
and S. sinensis (Chiappe et al., 1999; Ji et al., 2002a;
Pu et al., 2013; Godefroit et al., 2013a). The proximal
half of the femur is narrower than its distal half,
as in S. chaoyangensis,Buitreraptor gonzalezorum
(Makovicky, Apesteguía & Agnolín, 2005), R. ostromi,
and A. huxleyi. The femoral head is sub-oval and is
not separated from the trochanteric region by a dis-
tinct neck, contrary to what can be seen in C. sanctus.
CT scans reveal that the fossa for the capital ligament
is absent at the top of both femoral heads, resembl-
ing the condition in R. ostromi,Z. elegans, and
A. lithographica (Forster, 1998; Wellnhofer, 2009;
Zhou & Li, 2010). The greater trochanter is not
prominent and is hardly discernible from the femoral
shaft. An oval area, the position of which corresponds
to that of fourth trochanter, is present along the
caudal side of the femur (Weishampel et al., 2007;
Lu et al., 2009). Distally, the medial and the lateral
condyles are caudally developed and, unlike in
Y. longicaudus, the medial condyle is wider than the
lateral one. The fibular trochlea is well developed on
the lateral condyle, whereas the tibiofibular crest is
poorly preserved. The intercondylar sulcus is nar-
rower than the lateral condyle. The popliteal fossa is
not as deep as in A. lithographica,R. ostromis, and
J. prima (Forster, 1998; Zhou & Zhang, 2002;
Wellnhofer, 2009). The tibia is straight as in
S. sinensis,C. sanctus,R. ostromi, and M. long
(Forster, 1998; Chiappe et al., 1999; Ji et al., 2002a;
Gao et al., 2012). This bone is as wide as the femur as
in J. prima,J. palmapenis, and S. sinensis (Zhou &
Zhang, 2002; Ji et al., 2002a; O’Connor et al., 2012).
There is no trace of a cnemial crest at the proximal
part of the bone as in many basal birds. The caudal
side of the tibia shows two prominent condyles
surroundering a deep fossa for insertion of the flexor
muscle. The fibular crest is laterally located and
extends over one-ninth of the length of the tibia.
Distally, the medial condyle is as wide as the lateral
condyle, as in S. chaoyangensis, whereas, in non-
avian theropods, the medial condyle is wider than the
lateral one (Chiappe et al., 1999).
The left fibula lies on its medial face and is iden-
tical to that of Jeholornis specimens in having a wide
Figure 8. Pelvic girdle of Jeholornis curvipes sp. nov.,
from the Lower Cretaceous of north-eastern China. ip,
intermediate process of ischium; lil, left ilium; lis, left
ischium; lpu, left pubis; lpap, left preacetabular process of
ilium; lpoap, left postacetabular process of ilium; ril, right
ilium; rpoap, right postacetabular process of the ilium;
rpu, right pubis. Scale bar = 40 mm.
Figure 9. Hindlimb of Jeholornis curvipes sp. nov.,
from the Lower Cretaceous of north-eastern China. mt,
metatarsal. Scale bar = 40 mm.
A NEW LONG-TAILED BASAL BIRD 797
© 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 790–804

proximal part that rapidly tapers distally. This bone
is half the length of the tibia and shows a
proximocranial process that projects anteriorly.
The left pes lies on its dorsal side. There are five
metatarsals as in A. lithographica,C. sanctus,
S. chaoyangensis,J. prima,J. palmapenis, and
S. sinensis (Chiappe et al., 1999; Zhou & Zhang, 2002;
Ji et al., 2002a; Wellnhofer, 2009; O’Connor et al.,
2012; Pu et al., 2013). Metatarsal V is reduced to less
than one-third the length of metatarsal IV. The proxi-
mal part of the five metatarsals is straight, whereas
the distal part is bent laterally. This configuration of
the metatarsals is not artefactual because X-ray
radiographs do not show any fracture or counterfeits.
The metatarsals are fused proximally, together with
the distal tarsals. They are not fused distally and
there are no vascular foramina in this area, unlike in
C. sanctus and S. choayangensis, and modern birds
(Baumel et al., 1993; Chiappe et al., 1999; Zhou &
Zhang, 2003b).
Metatarsal I is ‘J’-shaped in ventral view with a
sharp proximal end and a wider distal end as in
X. zhengi,C. sanctus, and Z. elegans (Chiappe et al.,
1999; Zhou & Li, 2010; Xu et al., 2011). Its distal
trochlea does not align distally with the distal
trochlea of the other metatarsals. It lays in the
same direction as metatarsals II–IV, suggesting that
metatarsal I dorsolaterally articulated with metatar-
sal II as in S. sinensis,J. palmapenis,C. sanctus,
A. huxleyi,E. brevipenna,A. lithographica, and A. xui
(Chiappe et al., 1999; Ji et al., 2002a; Hu et al., 2009;
Wellnhofer, 2009; O’Connor et al., 2012; Godefroit
et al., 2013a, 2013b) Metatarsal III is the largest and
reaches half the length of the tibia. Its proximal part
is not pinched as in the arctometatarsalian condition.
Metatarsals II and III terminate in a ginglymoid
articulation, whereas the metatarsal IV terminates,
as in C. sanctus, with a rounded, laterally-compressed
condyle.
The phalangeal formula is 2-3-4-5-0(X). Digit I is
short and does not reach the proximal end of the
distal trochlea of phalanx II-2. Phalanx II-2 is longer
than phalanx II-1, as in Microraptor gui and all
the basal avialans (Zhou & Zhang, 2002). Unlike
Dromaeosauridae and Troodontidae, phalanx II-2 is
not reduced in length and does not show a prominent
proximoventral heel (Xu et al., 2011). In all the digits,
the penultimate phalanx is longer than the others. The
flexor tubercles are not well developed, as in J. prima
(Zhou & Zhang, 2003a). The width of digits II and III
is constant, whereas the width of digit IV decreases
proximodistally. The unguals are broad and curved;
the ungual of digit II is slightly longer than that of digit
III. The ungual groove is centered on the lateral side of
each unguals. The proximoventral heel of each ungual
is more developed than the proximodorsal one.
Body mass estimation: Graphic double integration
(Seebacher, 2001) cannot be applied for estimating
the body mass of J. curvipes because of mediolateral
crushing of the fossil. A more empirical method, based
on the work of Christiansen & Farina (2004), is used
in the present study. This method gives a body mass
of 0.59 kg, which is consistent with the body mass
estimate of J. prima (see Supporting information,
Fig. S5).
DISCUSSION
COMPARISONS WITH OTHER JEHOLORNITHIFORMES
YFGP-yb2 shares with all species referred to
Jeholornithiformes the presence of a robust dentary
with a convex rostroventral margin and a concave
ventral margin. However, this character cannot be
regarded as diagnostic for the group because it is also
present in S. chaoyangensis, Oviraptorosauria, and
Therizinosauroidea (Zhou & Zhang, 2002; O’Connor
et al., 2012).
Dalianraptor cuhe is not considered for compari-
sons because it requires further preparation to clarify
its anatomy and systematic position (O’Connor et al.,
2012).
Shenzhouraptor sinensis and J. prima share the
following characters (Table 1) (Ji et al., 2003): (i)
robust mandibles with ossified symphyses; (ii) typical
U-shaped furculae; (iii) a lachrymal with two vertical
and elongated pneumatic fossa; (iv) a ratio of forelimb
to hindlimb length of approximately 1.2–1.27; (v) the
deltopectoral crest extends along 41% of the humeral
length; (vi) the phalanx II-1 is very wide; (vii) the
phalanx III-1 is twice as long as the phalanx III-2;
(viii) the flight feathers are distinctly longer than the
total length of both ulna and manus; (ix) a long tailed
composed of 23 or more caudal vertebrae; and (x) the
length of each vertebra is three to four times its
width. However, the number of caudal vertebrae
is at first sight different in both taxa: 23–25 in
Shenzhouraptor,27inJ. prima IVPP 13274, and 24
in J. prima IVPP V13353 (Ji et al., 2003; Zhou &
Zhang, 2003a; Zhou and Zhang, 2002). However, this
difference cannot be regarded as diagnostic because
considerable intraspecific variation in the number of
caudal vertebrae is reported in A. lithographica and
S. chaoyangensis, although the number of cervical
vertebrae and dorsal vertebrae is relatively constant
(Berger, 1956; Webster & Goff, 1977; Zhou & Zhang,
2003b; Mayr et al., 2007; Pu et al., 2013). The absence
of teeth in S. sinensis cannot be considered as a
character that differs from J. prima because the skull
and mandibles are poorly preserved (Ji et al., 2003).
We therefore regard S. sinensis as a synonym of
J. prima sensu Ji et al. (2003), Zhou & Zhang (2006),
Li et al. (2010) and Zhou & Wang (2010).
798 U. LEFÈVRE ET AL.
© 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 790–804

Table 1. Character comparisons within Jeholornithiformes
Character
Shenzhouraptor
sinensis
Jeholornis
prima
Jeholornis
palmapenis
Jeholornis
curvipes
Jixiangornis
orientalis
Robust mandibles with ossified symphysis ? Yes Yes Yes Yes
Typical U-shaped furcula Yes Yes ? No ?
Lachrymal with two elongated pneumatic fossa Yes Yes ? Yes Yes
Ratio of forelimb to hindlimb length 1.2–1.27 1.2 ? 1.23 1.35
The deltopectoral crest extends along of the humeral length 41% 41% ? 40% 38%
The phalanx II-1 is very wide Yes Yes ? Yes Yes
The phalanx III-1 twice as long as the phalanx III-2 Yes Yes ? ? ?
The flight feathers disctinctly longer than the total length of both ulna and manus Yes ? ? ? ?
The length of each caudal vertebra is three to four times its width Yes Yes Yes Yes Yes
A toothed maxillary No No Yes No No
Thoracic vertebra laterally excavated proximally by two fenestra that converge
centrally and forming a single lateral opening in the caudalmost thoracics
No ? Yes ? ?
The dorsal margin of ilium strongly convex No No Yes No No
Post-acetabular wing of the ilium strongly concave ventrally No No Yes No ?
Ischium curved dorsally No No Yes ? No
The presence of a large one-piece sternum with a faint keel No No Yes ? Yes
Metatarsals 55554
Strut-like coracoid with a lateral process not detached from the lateral margin No No ? Yes No
Medial margin of coracoid concave along its entire length ? No ? Yes No
Sternal margin of the coracoid sub-horizontal ? No ? Yes No
Deltopectoral crest less deflected from the shaft of the humerus No No ? Yes No
Ratio metacarpal II to humerus length 0.46 0.43 ? 0.58 0.40
Ratio metacarpal I to metacarpal II length ? 0.28 ? 0.17 0.25
Ratio manual phalanx I-1 to phalanx II-1 length ? 1.12 ? 1.29 1.23
Presence of two prominent condyles at the caudal side of the tibia projecting
far posteriorly
No No No Yes No
Distal half of metatarsals with a lateral deviation No No No Yes No
A NEW LONG-TAILED BASAL BIRD 799
© 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 790–804

Many characters considered diagnostic to
S. sinensis are also found in J. curvipes (Table 1) but
many others characters also distinguish J. curvipes
from S. sinensis (see J. curvipes diagnosis, above).
Comparing J. palmapenis and S. sinensis is more dif-
ficult because J. palmapenis does not preserve its
scapular girdle and its forelimbs.
Jeholornis palmapenis is well characterized by the
following autapomorphies that are not found in
J. prima,J. curvipes, and J. orientalis. This taxon is
therefore regarded as a valid species rather than a
synonym of J. prima (Table 1): (i) a toothed maxillary;
(ii) thoracic vertebra laterally excavated proximally
by two fenestra that converge centrally and forming a
single lateral opening in the caudalmost thoracics;
(iii) dorsal margin of ilium strongly convex; (iv) post-
acetabular wing of the ilium strongly concave ven-
trally; and (v) ischium curved dorsally (O’Connor
et al., 2012).
Jixiangornis orientalis also has characters that are
unique among Jeholornithiformes, including: (i) the
presence of a large single-element sternum with a
faint keel; (ii) ratio of forelimb to hindlimb length of
1.31, and (iii) a deltopectoral crest that extends less
than 40% of the humeral length. The presence of
four metatarsals in J. orientalis is an other character
that can be used to distinguish it from J. prima,
J. palmapenis, and J. curvipes.
Several characters also clearly differentiate
J. curvipes from the other members of Jeholor-
nithiformes (Table 1): (i) dentary without ossified
symphysis contrary to what is seen in J. prima and
S. sinensis (in J. palmapenis, the dentaries appear to
be unfused rostrally, as in J. curvipes, although the
region is very poorly preserved; O’connor et al., 2012);
(ii) strut-like coracoid with a lateral process not
detached from the lateral margin, contrary to
J. prima; (iii) medial margin of coracoid concave along
its entire length (this is straight to slightly convex in
J. prima and J. orientalis); (iv) sternal margin of the
coracoid sub-horizontal and not convex as in J. prima
and J. orientalis; (v) deltopectoral crest less deflected
from the shaft of the humerus than in J. prima,
S. sinensis, and J. orientalis; (vi) ratio between meta-
carpal II to humerus length of 0.58, greater than
J. prima,S. sinensis, and J. orientalis (0.43, 0.46, and
0.40, respectively); (vii) ratio metacarpal I to meta-
carpal II length 0.17, smaller than J. prima and
J. orientalis (0.28 and 0.25, respectively); (viii) ratio
manual phalanx I-1 to phalanx II-1 length of 1.29,
slightly greater than J. orientalis (1.12) but closer to
J. prima (1.23); (ix) the presence of two prominent
proximal condyles on the caudal side of the tibia that
project further posteriorly than in the other species,
and (x) distal half of metatarsals with a lateral devia-
tion unique within Jeholornithiformes. These ten
characters distinguish J. curvipes from all previously
described specimens.
PHYLOGENETIC ANALYSIS
A phylogenetic analysis was conducted to assess the
relationships of J. curvipes within Avialae. Although
several phylogenies of Avialae have been proposed
(Xu & Norell, 2004; Xu & Zhang, 2005; Hu et al.,
2009; Xu et al., 2009; Xu et al., 2011; Lee & Worthy,
2012), our analysis is based on the data matrix
published by Godefroit et al. (2013a). The
Shenzhouraptor operational taxonomic unit (OTU)
was removed from the original matrix because it
included all species referred to the genus Jeholornis,
and three new OTUs were included, J. prima,
J. palmapenis, and J. curvipes. Our new matrix now
includes 103 OTUs for 991 informative characters.
The results of our phylogenetic analysis are pre-
sented in Figure 10, with focus on Avialae. The char-
acter distribution for the three new taxa added to
the original matrix is presented in the Supporting
information (Fig. S6).
Nine hundred and ninety-one characters were
equally weighted and analyzed with TNT, version 1.1
(Goloboff, Farris & Nixon, 2008). A heuristic search of
1000 replicates using random addition sequences,
followed by branch swapping by tree-bisection–
reconnection (holding ten trees per replicate) was
conducted. The trees were subsequently analyzed
using WINCLADA, version 1.00.08 (Nixon, 2002).
Bremer support was assessed by computing decay
indices with TNT, version 1.1.
The maximum parsimony analysis resulted in a
single tree of 4570 steps. The consistency index is
0.26 and the retention index is 0.54. The consensus
tree together with its description is presented in the
Supporting information (Fig. S7). Species referred to
Jeholornithiformes occupy a position within Avialae
confirming their status as basal birds. Jeholornis
prima,J. palmapenis, and J. curvipes form a clade
supported by nine homoplasious synapomorphies: (i)
the dorsoventral process of the dentary is ventrally
bowed (character 1359 [1]); (ii) the anteroposterior
length of the pre-acetabular process is less than the
anteroposterior length of the post-acetabular process
(character 398 [0]); (iii) the length of the pre-
acetabular process does not reach 6/5th of its proxi-
mal dorsoventral height (character 803 [0]); (iv) the
proximodistal length of the post-acetabular process
comprises between 2/5th and all of the space between
the pre-acetabular and post-acetabular embayement
of the bone (character 1055 [1]); (v) the distal end of
the ischial peduncle in lateral view is a broad and flat
articular surface (character 628 [1]); (vi) the ratio of
the dorsoventral depth to the basal anteroposterior
800 U. LEFÈVRE ET AL.
© 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 790–804

length of the pubic peduncle comprises between 1 and
1/2th (character 1143 [1]); (vii) the tibiofibular crest at
the distal end of the femur is sharply demarcated
from the fibular condyle by a sulcus or a concavity
(character 735 [1]); (viii) the distal half of the femur
is as thick as its proximal half (character 1093 [0]),
and (ix) the distal extensor pits of the pedal phalan-
ges are deep with defined margins (character 1450
[1]). These nine homoplasious synapomorphies
can be used to redefine the genus Jeholornis. The
J. prima +J. curvipes clade is supported by three
homoplasious synapomorphies, namely: (i) the ante-
rior surface of the cervical and anterior dorsal verte-
brae centra is moderately convex (character 194 [1]);
(ii) the shape of chevrons for the middle caudal ver-
tebrae represents a hyperelongated processes, longer
than two times adjacent centra (character 355 [4]),
and (iii) the posteroventral surface of the dentary for
the symphysis is present and prominent (character
1240 [1]).
In our phylogeny, J. orientalis is the sister-taxon to
the pygostylian lineage, as already recovered by
Turner, Makovicky & Norell (2012). The two lineages
share these synapomorphies: (i) the anteroposterior
diameter of the external mandibular fenestra is
subequal or longer than one-quarter of the length of
the mandible (character 171 [1]); (ii) seven sacral
vertebrae have their transverse processes in articu-
lation with the medial wall of the iliac blades (char-
acter 343 [5]); (iii) the proximodistal length of the
tibia comprises between 1 and 6/5 (character 460 [1]);
(iv) metatarsals are proximally fused together
(character 476 [1]); (v) the distalmost extent of the
laterodistal end of the tibia is distally placed to the
distal extent of the mediodistal end (character 561
[1]); (vi) the sternal plates are ossified (character 1037
[1]), and (vii) the presence of a hypocleidum at the
junction of the branches of the furcula (character
1385 [1]). As previously discussed, Jeholornithi-
formes (J. prima,J. palmapenis,S. sinensis, and
J. orientalis) is paraphyletic. Jixiangornis orientalis
should no longer be considered as a member of
Jeholornithiformes because its phylogenetic position
is closer to the pygostylians than to J. prima.By
contrast to the study of Zhou & Li (2010) and Turner
et al. (2012), S. chaoyangensis is recovered here as the
most basal pygostylian. The position of this taxon
suggests that the reduction of the tail into a pygostyle
was a unique event in the evolution of birds as has
been suggested by previous studies (Chiappe &
Witmer, 2002; Zhou & Zhang, 2002; Gao et al., 2008;
Godefroit et al., 2013a).
PALEOECOLOGY
All of the species referred to Jeholornis are charac-
terized by a robust triangular dentary, likely related
to their granivorous diet (Zhou & Zhang, 2002).
Jeholornis curvipes has shorter penultimate pedal
phalanges, as has J. prima,J. palmapenis, and
nonperching birds (Zhang et al., 2002); species
referred to the genus Jeholornis therefore probably
had terrestrial habits (Zhang et al., 2002). The
absence of a reversed hallux in Jeholornis supports
this hypothesis (O’Connor, Chiappe & Bell, 2011). In
J. prima and J. palmapenis, the rectrices taper dis-
tally and do not overlap significantly proximally and
distally so no airfoil is formed. Such plumage, which
resembles the configuration in Caudipteryx zoui
(Dyke & Norell, 2005), is possibly related to a visual
display function (Zhou & Zhang, 2003b; O’Connor
et al., 2012). Recent accurate observations of J. prima
Figure 10. Strict consensus tree resulting from our phylogenetic analysis (length = 4570, consistency index = 0.26,
retention index = 0.54). Only clade Avialae is presented (for the complete consensus tree, see the Supporting information,
Fig. S7). A, Jeholornithiformes. B, pygostylia. Bremer support values superior to 1 are indicated next to the internodes.
A NEW LONG-TAILED BASAL BIRD 801
© 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 790–804

and J. palmapenis have highlighted a fan-shaped
tract of feathers above the proximal caudal vertebrae.
This tract appears to be more implicated in ornamen-
tation than in flight (O’Connor et al., 2013).
In view of our results, it is likely that the genus
Jeholornis only includes terrestrial and seed-eating
birds, although quantitative analyses are necessary
to understand the aerodynamic capacity of these
basal birds.
CONCLUSIONS
The genus Jeholornis encompasses three species:
J. prima (Zhou & Zhang, 2002), J. palmapenis
(O’Connor et al., 2012), and J. curvipes sp. nov.
Shenzhouraptor sinensis is here considered as a
junior synonym of J. prima. Our phylogenetic analy-
sis shows that J. orientalis is not a member of the
Jeholornithiformes but the sister-taxon to the
pygostylians. This position is supported by seven
homoplasious synapomorphies, although further
investigation of the specimen will be required to fully
determine its phylogenetic position. Our phylogenetic
analysis also corroborates the suggestion that the
reduction of the tail into a pygostyle was a unique
event in the evolution of birds. The presence of a
nonreversed caudomedially positioned hallux and a
unique caudal plumage combined with a phalangeal
proportion that are not characteristic for perching
birds indicate that the Jeholornis species were most
likely terrestrial, seed-eating birds.
ACKNOWLEDGEMENTS
We thank A. Cau for discussions and information
about the phylogenetic matrix and C. Moors who
helped to define the species name. We thank T. Hubin
for photographs. We also thank M. Mortimer and X.
Wang for their help with earlier iterations of this
manuscript and three anonymous reviewers for their
helpful comments.
REFERENCES
Baumel JJ, King AS, Breazile JE, Evans HE,
Vanden Berge JC. 1993. Handbook of avian anatomy:
nomina anatomica avium, 2nd edn. Cambridge, MA: The
Nuttall Ornithological Club.
Berger A. 1956. Anatomical variation and avian anatomy.
Condor 58: 433–441.
Cai Z, Zhao L. 1999. A long tailed bird from the Late
Cretaceous of Zhejiang. Science in China 42: 434–441.
Chang MM, Chen PJ, Wang YQ, Wang Y, Miao D. 2008.
The Jehol fossils: the emergence of feathered dinosaurs,
beaked birds and flowering plants. Amsterdam: Academic
Press.
Chang S, Zhang H, Renne PR, Fang Y. 2009.
High-precision 40Ar/39Ar age for the Jehol biota.
Palaeogeography, Palaeoclimatology, Palaeoecology 280:
94–104.
Chiappe LM, Norell MA, Ji Q, Ji S. 1999. Anatomy and
systematics of the confuciusornithidae (Theropod: Aves)
from the late Mesozoic of northeastern China. American
Museum of Natural History 242: 1–89.
Chiappe LM, Witmer LM. 2002. Mesozoic birds: above the
heads of dinosaurs. London: University of California Press.
Christiansen P, Farina RA. 2004. Mass prediction in
theropod dinosaurs. Historical biology 16: 85–92.
Dyke GJ, Norell MA. 2005. Caudipteryx as a non-avialan
theropod rather than a flightless bird. Acta Palaeontologica
Polonica 50: 101–116.
Forster CA. 1998. The theropod ancestry of birds: new evi-
dence from the Late Cretaceous of Madagascar. Science 279:
1915–1919.
Gao CL, Chiappe LM, Meng QJ, O’Connor JK, Wang XR,
Cheng XD, Liu JY. 2008. A new basal lineage of Early
Cretaceous birds from China and its implications on the
evolution of the avian tail. Palaeontology 51: 775–791.
Gao CL, Liu JY. 2005. A new avian taxon from Lower
Cretaceous Jiufotang Formation of western Liaoning.
Global Geology 24: 313–316.
Gao CL, Morschhauser EM, Varrichio DJ, Liu J, Zhao B.
2012. A second soundly sleeping dragon: new anatomical
details of the Chinese troodontid Mei long with implications
for phylogeny and taphonomy. PloS ONE 7: e45203.
Gartner PL, Hiatt JL. 2004. Atlas en couleur d’histologie,
2nd edn. Paris: Edition Pradel.
Gauthier J. 1986. Saurischian monophyly and the origin of
birds. Memoirs of the California Academy of Sciences 8:
1–55.
Godefroit P, Cau A, Hu DY, Escuillié F, Wu W, Dyke G.
2013a. A Jurassic avialan dinosaur from China resolves
the early phylogenetic history of birds. Nature 498: 359–
362.
Godefroit P, Demuynck H, Dyke G, Hu D, Escuillié F,
Claeys P. 2013b. Reduced plumage and flight ability of a
new Jurassic paravian theropod from China. Nature Com-
munications 4: 1394.
Goloboff P, Farris J, Nixon K. 2008. TNT, a free program
for phylogenetic analysis. Cladistics 24: 774–786.
Howard H. 1929. The avifauna of Emeryville Shellmound.
University of California Publications in Zoology 32: 301–
394.
Hu D, Hou L, Zhang L, Xu X. 2009. A pre-Archaeopteryx
troodontid theropod from China with long feathers on the
metatarsus. Nature 461: 640–643.
Hubert JF, Panish PT, Chure DJ, Prostak KS. 1996.
Chemistry, microstructure, petrology and diagenetic model
of Jurassic dinosaur bones, Dinosaur National Monument,
Utah. Journal of Sedimentary Research 66: 531–547.
Ji Q, Ji S, You H, Zhang J, Yuan C, Ji X, Li J, Li Y, By
T, Downs W. 2002a. Discovery of an Avialae bird from
China, Shenzhouraptor sinensis.Geological Bulletin of
China 21: 363–369.
802 U. LEFÈVRE ET AL.
© 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 790–804

Ji Q, Ji S, You H, Zhang J, Zhang H, Zhang N,
Yuan C, Ji X. 2003. An early Cretaceous avialan bird,
Shenzhouraptor sinensis from Western Liaoning, China.
Acta Geologica Sinica 77: 21–27.
Ji Q, Ji S, Zhang H, You H, Zhang J, Wang L, Yuan C, Ji
X. 2002b. A new Avialan bird – Jixiangornis orientalis gen. et
sp. nov. – from the Lower Cretaceous of Western Liaoning,
NE China. Journal of Nanjing University 38: 723–736.
Jiang B, Fürisch FT, Sha J, Wang B, Niu Y. 2011.
Early Cretaceous volcanism and its impact on fossil pres-
ervation in western Liaoning, NE China. Palaeogeography,
Palaeoclimatalogy, Palaeoecology 302: 255–269.
Leduc T. 2012. Les iguanodons de Bernissart : leur évolution
diagénétique et les processus de dégradation. DPhil Thesis,
University of Liège (Belgium).
Lee M, Worthy T. 2012. Likelihood re-instates Archaeopteryx
as a primitive bird. Biology Letters 8: 299–303.
Li D, Sullivan C, Zhou Z, Zhang F. 2010. Basal birds from
China: a brief review. Chinese Birds 1: 83–96.
Lu J, Xu L, Jiang X, Kia S, Li M, Yuan C, Zhang X, Ji Q.
2009. A preliminary report on the new dinosaurian fauna
from the Cretaceous of the Ruyang Basin, Henan province
of central China. Journal of the Paleontological Society of
Korea 25: 43–56.
Makovicky PJ, Apesteguía S, Agnolín FL. 2005. The
earliest dromaeosaurid theropod from South America.
Nature 437: 1007–1011.
Marsh OC. 1881. Principal characters of American Jurassic
dinosaurs – part V. American Journal of Science and Arts 3:
417–423.
Mayr G, Pohl B, Hartman S, Peters DS. 2007. The tenth
skeletal specimen of Archaeopteryx.Zoological Journal of
the Linnean Society 149: 97–116.
Newesely H. 1989. Fossil bone apatite. Applied Geochemistry
4: 233–245.
Nixon KC. 2002. Winclada, Version 1.00.08. Available at:
http://www.cladistics.com/
Norell MA, Clarke JA. 2001. Fossil that fills a critical gap in
avian evolution. Nature 409: 181–184.
O’Connor J, Chiappe LM, Bell A. 2011. Pre-modern birds:
avian divergences in the Mesozoic. In: Dyke G, Kaiser G,
eds. Living dinosaurs: the evolutionary history of modern
birds. Oxford: Wiley-Blackwell, 39–114.
O’Connor J, Sun C, Xu X, Wang X, Zhou Z. 2012. A new
species of Jeholornis with complete caudal integument. His-
tological Biology 24: 29–41.
O’Connor J, Wang X, Sullivan C, Zheng X, Tubaro P,
Zhang X, Zhou Z. 2013. Unique caudal plumage of
Jeholornis and complex tail evolution in early birds.
Proceedings of the National Academy of Sciences of the
United States of America October 7, 2013, doi: 10.1073/
pnas.1316979110.
Owen R. 1842. Report on British fossil reptiles. Report of the
British Association for the Advancement of Sciences 9:
60–204.
Pu H, Chang H, Lü J, Wu Y, Xu L, Zhang J, Jia S.
2013. A new juvenile specimen of Sapeornis (Pygostylia:
Aves) from the Lower Cretaceous of Northeast China
and allometric scaling of this basal bird. Paleontological
Research 17: 27–38.
Seebacher F. 2001. A new method to calculate allometric
length-mass relationships of dinosaurs. Journal of Verte-
brate Paleontology 21: 51–60.
Sereno PC, Rao C. 1992. Early evolution of avian flight and
perching: new evidence from the Lower Cretaceous of
China. Science 55: 845–848.
Swisher CC, Wang YQ, Wang XL, Xu L, Wang Y. 1999.
Cretaceous age for the feathered dinosaurs of Liaoning
China. Nature 400: 58–61.
Turner AH, Pol D, Clarke JA, Erickson GM, Norell MA.
2007. A basal dromaeosaurid and size evolution preceding
avian flight. Science 317: 1378–1381.
Turner AH, Makovicky PJ, Norell MA. 2012. A review of
dromaeosaurid systematics and paravian phylogeny. Bulle-
tin of the American Museum of Natural History 371: 1–206.
Webster D, Goff G. 1977. Variation in the vertebral column
and ribs of songbirds. Proceedings of the Indiana Academy
of Science 87: 450–459.
Weishampel DB, Dodson P, Osmólska H. 2007. The
Dinosauria, 2nd edn. Berkeley, CA: University of California
Press.
Wellnhofer P. 2009. Archaeopteryx lithographica: the icon of
evolution. München: Verlag Dr. Friedrich Pfeil.
Xu X, Norell MA. 2004. A new troodontid dinosaur from
China with avian-like sleeping posture. Nature 475: 465–
470.
Xu X, You H, Du K, Han F. 2011. An Archaeopteryx-like
theropod from China and the origin of Avialae. Nature 475:
465–470.
Xu X, Zhang F. 2005. A new maniraptoran dinosaur
from China with long feathers on the metatarsus.
Naturwissenschafter 92: 173–177.
Xu X, Zhao Q, Norell M, Sullivan C, Hone D, Erickson G,
Wang X, Han F, Guo Y. 2009. A new feathered
maniraptoran dinosaur fossil that fills a morphological gap
in avian origins. Chinese Science Bulletin 54: 430–435.
Yang W, Li S, Jiang B. 2007. New evidence for Cretaceous
age of the feathered dinosaurs of Liaoning: zircon U-Pb
SHRIMP dating of the Yixian Formation in Shietun, north-
east China. Cretaceous Research 23: 297–305.
Zhang F, Zhou Z, Xu X, Wang X. 2002. A juvenile
coelurosaurian theropod from China indicates arboreal
habits. Die Naturwissenschaften 89: 394–398.
Zhang F, Zhou Z, Xu X, Wang X, Sullivan C. 2008. A
bizarre Jurassic maniraptoran from China with elongate
ribbon-like feathers. Nature 455: 1105–1108.
Zhou Z, Li FZZ. 2010. A new Lower Cretaceous bird from
China and tooth reduction in early avian evolution. Proceed-
ings of the Royal Society of London Series B, Biological
Sciences 277: 219–227.
Zhou Z, Wang Y. 2010. Vertebrate diversity of the Jehol biota
as compared with other lagerstätten. Science China Earth
Sciences 53: 1894–1907.
Zhou Z, Zhang F. 2002. A long-tailed, seed-eating bird from
the Early Cretaceous of China. Nature 418: 1754–1756.
Zhou Z, Zhang F. 2003a. Jeholornis compared to
A NEW LONG-TAILED BASAL BIRD 803
© 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 790–804

Archaeopteryx with a new understanding of the earliest
avian evolution. Die Naturwissenschaften 90: 220–225.
Zhou Z, Zhang F. 2003b. Anatomy of the primitive bird
Sapeornis chaoyangensis from the Early Cretaceous of
Liaoning, China. Canadian Journal of Earth Sciences 40:
731–747.
Zhou Z, Zhang F. 2006. Mesozoic birds of China – a synoptic
review. Vertebrata Palasiatica 44: 74–98.
SUPPORTING INFORMATION
Additional Supporting Information may be found in the online version of this article at the publisher’s web-site:
Figure S1. Selected measurements of YFGP-yb2.
Figure S2. Three-dimensional reconstruction of the pelvic region from computed tomography scans.
Figure S3. Energy dispersive spectrometry analysis.
Figure S4. Slices and three-dimensional reconstruction of the ilium in Jeholornis curvipes.
Figure S5. Body mass estimation.
Figure S6. Operational taxonomic units added.
Figure S7. Phylogeny.
804 U. LEFÈVRE ET AL.
© 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 790–804