ArticlePDF Available

Abstract and Figures

Birds are one of the most successful groups of vertebrates. The origin of birds from their reptilian ancestors is traditionally rooted near the Jurassic "Urvogel" Archaeopteryx, an approach that has contributed in defining the dichotomy between the "reptilian" (pre-Archaeopteryx) and "avian" (post-Archaeopteryx) phases of what is instead a single evolutionary continuum. A great and still ever increasing amount of evidence from the fossil record has filled the gaps between extinct dinosaurs, Mesozoic birds and modern avians, and led to the revision of the misleading dichotomy between pre-and post-Archaeopteryx stages in the evolution of bird biology. Herein, the progressive assembly of the modern avian body plan from the archosaurian ancestral condition is reviewed using a combination of phylogenetic methods. The stem lineage leading to modern birds is described using 38 internodes, which identity a series of progressively less inclusive ancestors of modern birds and their Mesozoic sister taxa. The 160-million-year long assembly of the avian bauplan is subdivided into three main stages on the basis of analyses of skeletal modularity, cladogenetic event timing, divergence rate inference and morphospace occupation. During the first phase ("Huxleyian stage": Early Triassic to Middle Jurassic), the earliest ancestors of birds acquired postcranial pneumatisation, an obligate bipedal and digitigrade posture, the tridactyl hand and feather-like integument. The second phase ("Ostromian stage": second half of Jurassic) is characterised by a higher evolutionary rate, the loss of hypercarnivory, the enlargement of the braincase, the dramatic reduction of the caudofemoral module, and the development of true pennaceous feathers. The transition to powered flight was achieved only in the third phase ("Marshian stage": Cretaceous), with the re-organisation of both forelimb and tail as flight-adapted organs and the full acquisition of the modern bauplan. Restricting the investigation of the avian evolution to some Jurassic paravians or to the lineages crown-ward from Archaeopteryx ignores the evolutionary causes of over 60% of the features that define the avian body. The majority of the key elements forming the third phase are exaptations of novelties that took place under the different ecological and functional regimes of the Huxleyian and Ostromian stages, and cannot be properly interpreted without making reference to their original historical context. RIASSUNTO - [La costruzione del piano corporeo aviano: un processo lungo 160 milioni di anni] - Gli uccelli sono uno dei gruppi di vertebrati di maggiore successo. L'origine degli uccelli dai loro antenati rettiliani è tradizionalmente ancorata intorno allo "Urvogel" giurassico Archaeopteryx; questo approccio ha consolidato la distinzione tra una fase "rettiliana" (precedente Archaeopteryx) ed una "aviana" (successiva ad Archaeopteryx) in quello che è invece un singolo continuum evolutivo. Una crescente quantità di evidenze dal registro fossilifero ha colmato le lacune esistenti tra i dinosauri non-avialiani, gli uccelli mesozoici e quelli moderni, e ha portato alla revisione della fuorviante dicotomia tra fasi pre-e post-Archaeopteryx nell'evoluzione della biologia aviana. Il progressivo assemblaggio del moderno piano corporeo aviano è qui discusso usando una combinazione di metodi fi logenetici. La linea filetica che conduce agli uccelli moderni è descritta da 38 internodi, che identifi cano una serie progressiva di antenati condivisi tra gli uccelli attuali e i loro sister group mesozoici. I 160 milioni di anni di durata della costruzione del bauplan aviano sono suddivisi in tre fasi principali sulla base di analisi della modularità scheletrica, della cronologia degli eventi cladogenetici, dei tassi di divergenza, e delle regioni del morfospazio occupate. Durante la prima fase (detta "huxleyiana": dal Triassico Inferiore al Giurassico Medio), gli antenati degli uccelli svilupparono la pneumatizzazione postcraniale, una postura bipede obbligata e digitigrada, la mano tridattila e un tegumento simile al piumaggio. La seconda fase ("ostromiana": seconda metà del Giurassico) è caratterizzata da un più elevato tasso di evoluzione divergente, la perdita dell'ecologia ipercarnivora, l'espansione dell'endocranio, la drammatica riduzione del modulo caudofemorale, e lo sviluppo di piumaggio pennaceo. La transizione al volo battuto fu sviluppata solo nella terza fase ("marshiana": Cretacico), con la riorganizzazione dell'arto anteriore e della coda in organi adatti al volo, e la completa acquisizione del bauplan moderno. Restringere l'indagine sull'evoluzione aviana ad alcuni paraviani giurassici o alle linee successive ad Archaeopteryx significa ignorare la causa di oltre il 60% delle caratteristiche che definiscono il modello corporeo degli uccelli. La maggioranza degli elementi chiave che defi niscono la moderna fase dell'evoluzione aviana sono exaptation di novità occorse sotto diff erenti regimi ecologico-funzionali nelle fasi huxleyiana e ostromiana, e non possono essere propriamente interpretati senza fare riferimento al contesto storico della loro origine.
Content may be subject to copyright.
Bollettino della Società Paleontologica Italiana, 57 (1), 2018, 1-25. Modena
ISSN 0375-7633 doi:10.4435/BSPI.2018.01
The assembly of the avian body plan: a 160-million-year long process
Andrea CAU
A. Cau, Museo Geologico e Paleontologico “Giovanni Capellini”, Via Zamboni 63, I-40126 Bologna, Italia; cauand@gmail.com
KEY WORDS - Aves, Dinosauria, Macroevolution, Mesozoic, Phylogenetics, Theropoda.
ABSTRACT - Birds are one of the most successful groups of vertebrates. The origin of birds from their reptilian ancestors is traditionally
rooted near the Jurassic “Urvogel” Archaeopteryx, an approach that has contributed in de ning the dichotomy between the “reptilian” (pre-
Archaeopteryx) and “avian” (post-Archaeopteryx) phases of what is instead a single evolutionary continuum. A great and still ever increasing
amount of evidence from the fossil record has  lled the gaps between extinct dinosaurs, Mesozoic birds and modern avians, and led to the
revision of the misleading dichotomy between pre- and post-Archaeopteryx stages in the evolution of bird biology. Herein, the progressive
assembly of the modern avian body plan from the archosaurian ancestral condition is reviewed using a combination of phylogenetic methods.
The stem lineage leading to modern birds is described using 38 internodes, which identity a series of progressively less inclusive ancestors of
modern birds and their Mesozoic sister taxa. The 160-million-year long assembly of the avian bauplan is subdivided into three main stages
on the basis of analyses of skeletal modularity, cladogenetic event timing, divergence rate inference and morphospace occupation. During
the  rst phase (“Huxleyian stage”: Early Triassic to Middle Jurassic), the earliest ancestors of birds acquired postcranial pneumatisation,
an obligate bipedal and digitigrade posture, the tridactyl hand and feather-like integument. The second phase (“Ostromian stage”: second
half of Jurassic) is characterised by a higher evolutionary rate, the loss of hypercarnivory, the enlargement of the braincase, the dramatic
reduction of the caudofemoral module, and the development of true pennaceous feathers. The transition to powered  ight was achieved only
in the third phase (“Marshian stage”: Cretaceous), with the re-organisation of both forelimb and tail as  ight-adapted organs and the full
acquisition of the modern bauplan. Restricting the investigation of the avian evolution to some Jurassic paravians or to the lineages crown-ward
from Archaeopteryx ignores the evolutionary causes of over 60% of the features that de ne the avian body. The majority of the key elements
forming the third phase are exaptations of novelties that took place under the diff erent ecological and functional regimes of the Huxleyian
and Ostromian stages, and cannot be properly interpreted without making reference to their original historical context.
RIASSUNTO - [La costruzione del piano corporeo aviano: un processo lungo 160 milioni di anni] - Gli uccelli sono uno dei gruppi
di vertebrati di maggiore successo. L’origine degli uccelli dai loro antenati rettiliani è tradizionalmente ancorata intorno allo “Urvogel”
giurassico Archaeopteryx; questo approccio ha consolidato la distinzione tra una fase “rettiliana” (precedente Archaeopteryx) ed una
“aviana” (successiva ad Archaeopteryx) in quello che è invece un singolo continuum evolutivo. Una crescente quantità di evidenze dal registro
fossilifero ha colmato le lacune esistenti tra i dinosauri non-avialiani, gli uccelli mesozoici e quelli moderni, e ha portato alla revisione della
fuorviante dicotomia tra fasi pre- e post-Archaeopteryx nell’evoluzione della biologia aviana. Il progressivo assemblaggio del moderno piano
corporeo aviano è qui discusso usando una combinazione di metodi  logenetici. La linea  letica che conduce agli uccelli moderni è descritta
da 38 internodi, che identi cano una serie progressiva di antenati condivisi tra gli uccelli attuali e i loro sister group mesozoici. I 160 milioni
di anni di durata della costruzione del bauplan aviano sono suddivisi in tre fasi principali sulla base di analisi della modularità scheletrica,
della cronologia degli eventi cladogenetici, dei tassi di divergenza, e delle regioni del morfospazio occupate. Durante la prima fase (detta
“huxleyiana”: dal Triassico Inferiore al Giurassico Medio), gli antenati degli uccelli svilupparono la pneumatizzazione postcraniale, una
postura bipede obbligata e digitigrada, la mano tridattila e un tegumento simile al piumaggio. La seconda fase (“ostromiana”: seconda
metà del Giurassico) è caratterizzata da un più elevato tasso di evoluzione divergente, la perdita dell’ecologia ipercarnivora, l’espansione
dell’endocranio, la drammatica riduzione del modulo caudofemorale, e lo sviluppo di piumaggio pennaceo. La transizione al volo battuto
fu sviluppata solo nella terza fase (“marshiana”: Cretacico), con la riorganizzazione dell’arto anteriore e della coda in organi adatti al
volo, e la completa acquisizione del bauplan moderno. Restringere l’indagine sull’evoluzione aviana ad alcuni paraviani giurassici o alle
linee successive ad Archaeopteryx signi ca ignorare la causa di oltre il 60% delle caratteristiche che de niscono il modello corporeo degli
uccelli. La maggioranza degli elementi chiave che de niscono la moderna fase dell’evoluzione aviana sono exaptation di novità occorse
sotto diff erenti regimi ecologico-funzionali nelle fasi huxleyiana e ostromiana, e non possono essere propriamente interpretati senza fare
riferimento al contesto storico della loro origine.
INTRODUCTION
Plato had defined Man as an animal, biped and
featherless, and was applauded. Diogenes plucked a
fowl and brought it into the lecture-room with the words,
“Behold Plato’s man!”
(Diogenes Laërtius, in Hicks, 1925, p. 40)
While I appreciate their acceptance of my conclusions
about the ancestral affi nities of Archaeopteryx and later
birds, I reject the assertion by Bakker & Galton that
the avian radiation is merely an aerial exploitation of
basic dinosaurian physiology and structure, as well as
their reasoning that birds should therefore be classi ed
as dinosaurs. […] I confess that I am unable to accept
such theropods as Tyrannosaurus and Allosaurus as
“birds”, and therefore have little sympathy with this re-
classi cation scheme either.
(John Ostrom, 1976, p. 172)
The birds ( Linnaeus, 1758) represent the most
speciose lineage among the six forming the extant
tetrapod vertebrates (the other lineages being amphibians,
mammals, lepidosaurs, turtles, and crocodiles). Under
S. P. I.
S
O
C
I
E
T
A
'
P
A
L
E
O
N
T
O
L
O
G
I
C
A
I
T
A
L
I
A
N
A
Invited Paper
P
A
L
E
O
N
NN
Bollettino della Società Paleontologica Italiana, 57 (1), 2018
2
an evolutionary and phylogenetic framework, birds are
diapsid amniotes and form the archosaurian clade with
crocodiles and their relatives. Traditionally, the evolution
of birds is anchored to Archaeopteryx Meyer, 1861, with
the iconic “Urvogel” more or less explicitly assumed as
“transitional step” between the reptilian and the avian
grades (Ostrom, 1976). The use of Archaeopteryx as
key point along the reptile-avian transition has aimed to
conciliate the evolutionary paradigm with the traditional
separation between reptiles and birds (which precedes
        
is followed in non-phylogenetic taxonomies (Hennig,
1975; Dupuis, 1984), and is established in the academic
- and terminological - independence of herpetology and
ornithology (Prum, 2002; Harris, 2004).
This typological framework is among the most
consolidated in vertebrate zoology, and is still retained,

This is evidenced by the common use of vernacular
expressions including the negative adjective “non-avian”
associated to clade names now known to include birds
(e.g., “non-avian dinosaur”). Even after the recognition of
the evolutionary continuity between the (other) dinosaurs
and the extant birds (Ostrom, 1976; Gauthier & Padian,

of the “non-avian” grade, which is thus considered as an
useful biological category (even if explicitly 
non-monophyletic), qualitatively distinct from the birds
(Fig. 1). In literature, the adjective “non-avian” is more
frequently associated with the terms “theropods” and
“maniraptorans” than to any other name of clades that
include birds (pers. obs., Fig. 2). This shows that the
typological (non-phylogenetic) concepts of “theropod”
and “maniraptoran” (implicitly assumed in the “non
avian” categorisation) are more frequently-mentioned
than those of other groups. Ironically, while “Theropoda”
was erected over a century ago (Marsh, 1881) and was
used for decades as a “strictly reptilian” group (i.e., not
including birds), “Maniraptora” was explicitly erected
in a phylogenetic systematic context as a bird-bearing
clade (Gauthier, 1986)! This is further paradoxical,
given the considerable amount of studies establishing the
theropodan and maniraptoran nature of birds, published
during the last four decades (e.g., Gauthier & Padian,
1985; Gauthier, 1986; Padian & Chiappe, 1998; Norell
et al., 2001; Paul, 2002; Prum, 2002; Agnolín & Novas,
2013; Xu et al., 2014).
If “birds are maniraptoran theropods” is so vehemently
remarked, why the paraphyletic “non-avian” subgroups
of both Theropoda and Maniraptora are so frequently
used? It must be remarked that paraphyletic groups
represent arbitrary categories that do not correspond to
biological phenomena (Gauthier & Padian, 1985), but
may be retained as taxonomic tools due to their established
explanatory value (Rieppel, 2005). At least for the
members of Maniraptora, the analysis of the morphological
disparity rejects an explanatory value for the use of the
paraphyletic “non-avian maniraptoran” group, as it does
not represent a coherent ecomorphological cluster distinct
from that including Archaeopteryx and birds (Brusatte
et al., 2014). This result for Maniraptora automatically
invalidates any possible explanatory value also for “non-
avian theropods”, because the latter category includes all
taxa included in the “non-avian maniraptorans” category
(Gauthier, 1986; Prum, 2002). Thus, the persistent use
in literature of some paraphyletic grades of the avian
lineage (even if used as just vernacular terms) is not

This analysis of the vernacular taxonomy illuminates an
implicit pre-Darwinian background, still persistent in the
current age of phylogenetic systematics and feathered

taxonomic nomenclature, vernacular expressions are a

consolidate) the theories on the structure of the world
(Gould & Vrba, 1982). In this case, the persistent use
of paraphyletic tools in avian evolutionary literature not
only underestimates the evolutionary continuity between
    
   
Archaeopteryx, placed close to the arbitrary boundary
between the two categories of “non-avian” and “avian”.
Although the gradual evolutionary continuity between
birds and other dinosaurs is probably well-consolidated
among archosaur palaeontologists, the perception of the
actual distance between the avian and “reptilian” body
plans is more problematic among non-palaeontologists
(see Prum, 2002).
Fig. 1- The arbitrary boundary between birds and non-birds. The
traditional representation of the avian evolution is a single linear
transition from reptiles to birds, with Archaeopteryx as “origin” of
the avian lineage. This misleading scenario is implicitly retained
even under the phylogenetic paradigm: Archaeopteryx
boundary between birds and an arbitrary grade, the “non-avian
reptiles/dinosaurs”. Skeletal reconstructions by Marco Auditore
and Lukas Panzarin.
3
A. Cau - Avian Body Plan
The traditional division between herpetological
(“pre-Archaeopteryx”) and ornithological (“post-
Archaeopteryx”) parts of the avian evolution should
be abandoned, as it is fundamentally misleading (e.g.,
Harris, 2004). An alternative approach, that follows the
paradigm of the phylogenetic systematics, recognises
the whole lineage stemming from the divergence of the
birds and their closest living sister group (crocodiles) as
the necessary setting for a complete interpretation of the
bird biology. The history of the avian branch thus starts

last common ancestor of birds and crocodiles), and not
      Archaeopteryx.
This approach recognises that the most informative part
of the avian branch of  Cope, 1869, is the
least inclusive group including all living birds (
sensu stricto: the avian crown group, Gauthier, 1986;
but see Padian & Chiappe, 1998, for the use of a more
). The much larger
yet relatively less known part of the avian branch is formed
by all fossil forms not included in  s.s. but closer to
living birds than to the other living reptiles (the avian
stem-group). Following this approach, the subset of the
avian stem group formed by the series of branches that
leads to the origin of the crown group forms the Avian
Stem-Lineage (ASL). The series of evolutionary novelties
gained along the ASL describes the progressive assembly
of the avian body plan during over 160 million years, from
the origin of archosaurs (Early-Middle Triassic; Nesbitt
et al., 2017) to the root of the avian crown group (latest
Cretaceous or earliest Palaeogene; Clarke et al., 2005;
Lee at al., 2014a).
Although the close relationships between birds
       
late XIX Century (Huxley, 1868), for most of the XX
Century this hypothesis received secondary attention,
with dinosaurs and birds usually regarded as unrelated
lineages of the archosaurian radiation, rooted by distinct
“thecodontians” of the Triassic (see historical review in
Ostrom, 1976). The modern concept of the direct dinosaur-
bird relationships was introduced by Ostrom (1976), who
demonstrated that among all fossil reptiles, the small-
bodied theropod dinosaurs are those with the greatest
morphological similarity with Archaeopteryx. Under that
   
theropod dinosaurs. Barsbold (1983) further elaborated
the concept of a close evolutionary linkage between birds
      
combinations of bird-like features present in the various
groups of theropods demonstrate a general “ornithisation”
trend among these taxa, which culminated in the particular
lineage including Archaeopteryx. During the same decade,
the distribution of the avian-like features among the

   
scenario (Gauthier, 1986). The “ornithisation” of Barsbold
(1983) is thus a complex pattern that combines those
avian synapomorphies distributed along the ASL with the
numerous avian-like features independently gained by the
sister-taxa of birds (Holtz, 2001). The dinosaurian heritage

feathers and feather-like integumentary structures among
unambiguous dinosaurian taxa (Ji & Ji, 1996; Chen et al.,
1998; Ji et al., 1998). During the last 25 years, a growing

of the “reptile-bird discontinuity”. Along the crown-ward
side, dozens of new Mesozoic birds have revealed some
of the morphological, ecological and behavioural stages
between the grade of Archaeopteryx and the modern
birds; along the other side, several biological features,
traditionally restricted to birds among living vertebrates,
have been documented in many dinosaurian clades (see
reviews in Makovicky & Zanno, 2011; Xu et al., 2014).
       
phylogenetic systematics, Gauthier (1986) is one of the
      
      
and taxonomic system introduced by Gauthier (1986)
has inspired most of the theropod and stem-avian works
of the last three decades. In particular, Gauthier (1986)
Fig. 2 - All clades are monophyletic, but some clades are more monophyletic than others. Frequency of association of the “non-avian”
adjective with the most commonly-mentioned vernacular names of avian-including clades in technical publications (source, Google Scholar,
retrieved 7 March, 2018). While this paraphyletic use is marginally frequent with most names (< 5% of mentions), it represents about 13%
of the mentions of both “maniraptorans” and “theropods”.
Bollettino della Società Paleontologica Italiana, 57 (1), 2018
4
introduced a series of scions (i.e., monophyletic groups
including the crown clade and a subset of the stem
branches; Budd, 2001) or re-defined previous non-
monophyletic groups as avian scions. For example,

avian scions not including, respectively, ornithischians
and sauropodomorphs; tetanurans and maniraptorans
were introduced as the pan-avian scions excluding,
respectively, ceratosaurians and ornithomimosaurs.
During the last 30 years, dozens of analyses have
progressively expanded the sampling and improved
both completeness and resolution in the phylogenetic
investigation of the stem avians (e.g., Sereno et al., 1996,
1998; Sereno, 1999; Holtz, 2000; Norell et al., 2001;
Carrano et al., 2002; Rauhut, 2003; Holtz et al., 2004,
Sereno et al., 2004; Makovichy et al., 2005; Carrano
& Sampson, 2008; Smith et al., 2008; Xu et al., 2009;
Benson et al., 2010; Carrano et al., 2012; Naish et al.,
2012; Godefroit et al., 2013b; Brusatte et al., 2014; Lee
et al., 2014b; Cau et al., 2017; Lefèvre et al., 2017). The
present contribution focuses on the assembly of the body
plan of birds along the whole ASL, and reconstructs
the large-scale patterns during the “ornithisation”, here

distinguishing living birds from the other living reptiles.
MATERIAL AND METHODS
The phylogenetic data set includes 132 operational
taxonomic units scored for 1781 morphological character
statements (Supplementary Online Material). Character
     
Sereno (2007) and Brazeau (2011), and were in their large

in previous studies on theropod phylogeny (see Cau et
al., 2017, supplementary information, for the character
  
on a large-scale phylogenetic analysis of the pan-avian
clade (Cau, in prep.) and includes representatives of all
main pan-avian groups, each represented by two or more
species/genus-level taxa. Although pterosaurs are usually
placed among the basalmost members of the pan-avian
clade (e.g., Nesbitt et al., 2017), the ancestral condition
       
(Dalla Vecchia, 2013): pending a large-scale analysis
of pterosaur relationships that accurately samples the
Triassic disparity of the clade, they are provisionally
excluded from the analysis of the ASL. Taxa included
in this version of the analysis were chosen based on a
balanced series of criteria, such as amount of skeletal
completeness (preferring most complete taxa and those
sampling poorly known anatomical regions instead of
fragmentary taxa or those having character combinations
redundant with other better preserved taxa), inferred
phylogenetic position relative to other members of the
same subclade (i.e., using a consensus among recently
published phylogenies as reference, the earliest-diverging
members were preferred over members of late-diverging
    
oldest taxa of a clade to the youngest members). The data
set was analysed using maximum parsimony and Bayesian
inference integrating stratigraphic information as tree
search strategies. Parsimony analyses were performed
    
large size of the data set, the search strategy involved
100 “New Technology” search analyses using the default
setting, followed by a series of “New Technology” search
analyses exploring the tree islands found during the
      
recovered during the “New Technology” analysis rounds,
using “Traditional Search” analysis and saving up to
99.999 shortest trees (default maximum storage in TNT).
Nodal support was calculated saving all trees up to ten
steps longer than the shortest topologies found and using
the “Bremer Supports” function of TNT. Bayesian analysis
integrated the morphological data used for the parsimony
analysis with the absolute age (in million years before
the present, Mya) of each terminal taxon. The combined
morphological and stratigraphic data set was analysed
following the inference method discussed by Lee et al.
(2014a), using implementations discussed by Lee et
al. (2014b) and the Fossilised Birth-Death tree model
sampling ancestors (FBDSA) introduced by Gavryushkina
et al. (2014). Bayesian inference analyses were performed
in BEAST 2.4.4. (Drummond et al., 2012; Bouckaert et
al., 2014), implemented with the packages for the analysis
of morphological characters, using the model of Lewis
(2001), and for sampling potential ancestors among the
ingroup (Gavryushkina et al., 2014). Since the character
matrix includes autapomorphies of the sampled taxa, the
      
characters only. Stratigraphic information was taken from
the literature, and converted to mean geochronological
ages of the most inclusive known range of each taxon (see
Lee et al., 2014b). In this analysis, rate variation across
traits was modelled using the multi-gamma parameter
(default model and unique implemented for the analysis of
morphological data in BEAST 2). The rate variation across
branches was modelled using the relaxed log-normal
clock model, with the number of discrete rate categories
that approximates the rate distribution set as n-1 (with
n the number of branches), the mean clock rate using
default setting, and not setting to normalise the average
rate. Only root age constraint was enforced (the age of the
last common ancestor of all included taxonomic units),
conservatively set as a uniform range older than the age
of the oldest included taxa and centred on the Permian-
Triassic boundary (~ 252 Mya). The Bayesian analysis
performed a run of 40 million generations, sampling
every 1000 generations, with burnin set at 20%, and the
Maximum Clade Credibility Tree (MCCT) was used as
framework for phyletic reconstruction.
In all analyses, the Triassic archosauriform Euparkeria
Broom, 1913, was used as root of the trees. A detailed
description of the results and the diagnosis of all clades
recovered is beyond the aim of this study: here, I will focus
on the series of internodes along the lineage leading to
the extant birds (represented in the data set by Meleagris
Linnaeus, 1758), based on the strict consensus of all
shortest trees found.
The strict consensus topology of the shortest trees
found was used as framework for character transition
optimisation. Only unambiguous synapomorphies inferred
along the ASL internodes (the trajectory linking all pan-
avian scions) were considered. Although alternative
5
A. Cau - Avian Body Plan
options for ambiguous character optimisation are
available (e.g., accelerated transformation optimisation,
that minimises convergences and maximises reversals,
or delayed transformation optimisation, that maximises
convergences and minimises reversals) they may lead
to spurious character combinations and an unbalanced
distribution of character transition events along the
evolutionary sequence. The morphological characters
included in the analysis were grouped into six anatomical
regions: skull (including mandible and dentition),
presacral vertebral column (all vertebrae and ribs from
atlas to the posteriormost dorsal vertebra, and gastralia),
caudosacral vertebral column (all vertebrae and ribs
      
pectoral limb (pectoral girdle and forelimb, and including
sternum and clavicles), pelvic limb (including pelvis and
hindlimb) and integument (osteoderms and feathers). For
each node along the ASL, the total number of inferred
synapomorphies, and the particular number for each
anatomical region, were counted and compared to the
overall amount of changes along the entire lineage. The
relative amount of characters gained for each anatomical
region along the lineage was estimated and compared
with the overall amount and those in the other anatomical
regions. The resulted pattern formed the basis for a
quantitative analysis of modular evolution during the
avian body plan assemblage. Here, for “Ornithisation

of characters gained at a particular node of the ASL
and the corresponding amount gained at the  node
(the least inclusive node containing Meleagris in this
analysis). The OG may refer to the whole skeleton (as a
“whole OG”, OGw) or to a particular anatomical region
 
extant bird clade (the crown group ) is 100, whereas
that of  (or of any other more inclusive
bird-bearing clade) is 0. To avoid any misinterpretation
of the Ornithisation Grade as a “ranking” of the pan-
avian clades, note that the OG is exclusively a relative
measurement of the internodes along the ASL, and is
not a measure of “evolutionary level” for particular
terminal branches of the avian total group (e.g., although
the maniraptoriform node has a particular OG because
it is part of the ASL, the terminal members of the same
maniraptoriform node, that are not along the ASL [for
example, the ornithomimids] cannot be scored for the
OG).
The nodes of the “core topology” which is not biased
by the search strategy used (i.e., those shared by both
results of the parsimony-based and Bayesian-based

a chronologically progressive series of avian ancestors.
For each of these ancestors, the cladogenetic age (the
median age of the node inferred in the Bayesian analysis)
and the character state combination at that node (using
the parsimony-based topology) were inferred. The taxon-
character matrix of these ancestors was converted to an
Euclidean distance matrix and subjected to Principal
Coordinate Analysis (PCoA), in order to determine the

by the phylogenetically informative characters.
In this study, the name “ refers to the avian
crown-group, the least inclusive clade including the
living species (for a discussion on the use and alternative
    ”, see Gauthier, 1986);
accordingly, the term “avian” refers exclusively to the
modern birds. The taxonomic equivalent of the vernacular

1998). Here, the term is used conservatively for the taxa
that result members of Gauthier, 1986.
A note on the meaning of “assembly of body plan”:
although, under a typological paradigm, “plan” refers

the Darwinian paradigm explicitly recognises the plan as
the causal product of an assembly process. This means
that the avian body plan refers to the actual set of features
that describe the avians, and its assembly refers to the
historical process that produced that set.
RESULTS
Of the 1781 characters included in the analysis,
1431 resulted phylogenetically informative for the taxon
sample used. The phylogenetic analysis using TNT found
3072 shortest tree of 6790 steps each (Consistency Index
excluding uninformative characters = 0.2181, Retention
Index = 0.5634). The strict consensus of the 3072 shortest
trees found is well resolved, and is used as framework for
character evolution along the ASL (Figs 3-5). The result
of the Bayesian analysis is broadly consistent with that of
the parsimony analysis, and is visually summarised by the
stratigraphically calibrated Maximum Clade Credibility
Tree in Figs 6 and 7.
The sequence of character acquisitions along the ASL
The ASL is formed by a series of 38 internodes including
the extant bird Meleagris, here listed progressively from
the most inclusive node (Tab. 1). For each node, it is
reported the whole OG value (approximate to the nearest
integer), and the list of the unambiguous synapomorphies
inferred.
: (Teleocrater Nesbitt et al., 2017
+  Benton, 1985). The basalmost node

dinosauromorphs and aphanosaurians, represented here by
Teleocrater (Nesbitt et al., 2017). This node is diagnosed
by the following unambiguous synapomorphies: absence
of the subnarial fenestra, the relatively more acute
anterodorsal margin of maxilla, the relatively more
extensive ventral margin of the antorbital fossa, the

Taxon name Internal specier External specier Type
Dracohors (new) Megalosaurus bucklandii Marasuchus lilloensis Branch based
Maniraptoromorpha (new) Vultur gryphus Tyrannosaurus rex Branch based
Bollettino della Società Paleontologica Italiana, 57 (1), 2018
6
Fig. 3 - Result of the parsimony analysis. Strict consensus topology of the most parsimonious trees found. Numbers adjacent to nodes indicate nodal support (Decay Index) values > 1.
7
A. Cau - Avian Body Plan
extension of the supratemporal fossa onto the frontal,
the steeply inclined scapular acromion, the relatively
slender scapula, the relatively slender ischial shaft, and
the absence of paramedian osteoderms.
  (
Arcucci, 1986 +  Novas, 1992). Derived
features shared by lagerpetids and dinosauriforms but
absent in Teleocrater include a prominent supracetabular
shelf in the ilium and the upturning of the preacetabular
process (producing a concave dorsal margin of the
ilium), the ventral expansion of the femoral head and the
development of the anteromedial tuber in the femur. All
these features document the earliest re-organisation of
the hip-joint, from a primitive “reptile-like” sprawling
posture toward a more “bird-like” parasagittal posture
(Hutchinson, 2001a, b).
  (Marasuchus
Sereno & Arcucci, 1994 +  new clade).
Earliest dinosauriforms acquired additional muscle
attachment sites in the lateral surface of the femoral head,
absent in more basal pan-avians (a prominent trochanteric
shelf and a distinct anterior trochanter), and a more “bird-
like” foot, characterised by a distinct fossa in the distal
end of tibia for accommodating the astragalus ascending
process (indicating a more tight connection between
the proximal elements of the mesotarsal ankle joint),
a relatively more elongate metatarsus, and a relatively
shorter fourth toe compared to the third (producing a more
symmetrical foot).
:  (new clade).
Etymology - From draco (Latin, dragon) and cohors
(Latin, cohort, circle).
Definition - The most inclusive clade containing
Megalosaurus bucklandii Mantell, 1827, but excluding
Marasuchus lilloensis (Romer, 1971).
Remarks - Under all published topologies, 
includes silesaurids and all taxa universally recognised
as dinosaurs. Although the mutual relationships of the
main dracohorsian subclades (silesaurids, herrerasaurs,
sauropodomorphs, neotheropods and ornithischians)
are controversial (e.g., Sereno, 1999; Langer & Benton,
2006; Langer et al., 2010; Baron et al., 2017), this
lineage of dinosauriforms is universally recognised
by all authors, and its monophyly has never been
questioned by numerical analyses. Dracohorsian
synapomorphies include the anterior tympanic recess,
the axial epipophyses, the centrodiapophyseal laminae
in the presacral vertebrae, the relative size enlargement
of the postacetabular process of ilium, the elongation
of the pubis, the proximal sulcus and the reduction of
the ligament tuber in the femoral head, and the further
reduction in length of the fourth metatarsal and toe
compared to the third.

the recent re-evaluation of Pisanosaurus Casamiquela,
1967 among silesaurids and not as the basalmost
ornithischian (Agnolín & Rozadilla, 2017).
    :  Owen, 1842
(Eodromaeus Martinez et al., 2011, 
Benedetto, 1973,  Huene, 1932,
 Huxley, 1870). The analysis found
an unresolved polytomy including all dracohorsians
traditionally considered as “true” dinosaurs, but failed to
resolve the relationships of herrerasaur-grade forms relative
to sauropodomorphs and ornithoscelidans. In the Bayesian
analysis, herrerasaurs are found as non-dinosaurian
dracohorsians, although the support for this topology is
relatively weak. The numerous synapomorphies supporting
 (containing herrerasaurs) include the narial
fossa in the premaxilla, the posterolateral processes on
nasal, the reduction in height of the postorbital process of
jugal, the elongation of the dorsal quadratojugal process
of jugal, the posterodorsal process of dentary, the posterior
displacement of the axial neural spine, the elongation of the
anterior postaxial cervical vertebrae, the humerus not longer
than 60% of femur and with a distinction between head
and deltopectoral crest, the straight dorsal margin of ilium
(reversal to the plesiomorphic dinosauromorph condition),
the relative proximal placement of the obturator process
of ischium, a sharp fourth trochanter, the reduction of the
        
between metatarsal III and IV (suggesting a foot relatively
broader than in other dinosauromorphs).
:  (
Seeley, 1887 +  Marsh, 1881). This study
supports the recent hypothesis of a neotheropod-
ornithischian clade excluding sauropodomorphs and
herrerasaur-grade dinosaurs (Baron et al., 2017).
Ornithoscelidan synapomorphies (using the topology
inferred by the Bayesian analysis, i.e., sauropodomorphs
as sister-taxon of  relative to Eodromaeus
and herrerasaurs) are the interparietal median fusion, the
ventral expansion of the pterygoid ramus of quadrate,
the reduction of the anterior processes on cervical ribs,

gentle sloping of the acromial process relative to scapular
dorsal margin (reversal to the plesiomorphic pan-avian
condition), the elongation of the preacetabular process of
ilium, the relative narrowing of the intrapubic space, the
loss of the proximal sulcus of femoral head (reversal to
the plesiomorphic dracohorsian condition), the extensive
separation of the anterior trochanter from femoral shaft,
the transversal expansion of the medial malleous of

mediolateral constriction of the calcaneum with loss of
the posterolateral process.
:  (
[Nopcsa, 1928] +  Paul, 2002). The
ornithoscelidan hypothesis supported here excludes
most Triassic dinosaurs, otherwise considered as basal
theropods (e.g.,  see Sereno, 1999)
from , and restricts the latter to the two
neotheropod lineages, coelophysoids and averostrans.
The numerous (neo)theropod synapomorphies include
the medial subnarial foramen in premaxilla, the relatively
narrow snout with subparallel maxillae in ventral view, the
absence of a distinct rim along the margin of the antorbital
fossa, a subvertical orientation of the lacrimal ventral
Bollettino della Società Paleontologica Italiana, 57 (1), 2018
8
bar, the elongation of the postorbital process of jugal
(reversal to the plesiomorphic dinosaurian condition), the
posterior tympanic recess, the loss of the posterodorsal
process of dentary (reversal to the plesiomorphic
dinosaurian condition), the anteroventral foramen/
notch in the splenial, the pleurocoels in the anterior
presacral centra, the closely placed diapophyses and
parapophyses in anterior and middle cervical vertebrae,
the prezygodiapophyseal laminae in mid-cervical
vertebrae, the elongation of the posterior cervical centra,
the proximal caudal vertebrae with pre/postspinal laminae
and hyposphene-hypantra, the proximal chevrons with
anterior proximal processes, the fusion of the clavicles
(furcula), the relative elongation of the deltopectoral

length of metacarpal I, the anteroposteriorly elongate
pubic peduncle of ilium, the distal cleft in the obturator
process, the femoral head lacking the anteromedial tuber
and bearing a deep ligamental sulcus, the mediodistal
crest on femur, the tibia with a distinct cnemial crest and

tibia reaching the proximal end of bone, the oblique ridge

the anterior horizontal groove, the anterior platform and
a posterior ascending process, the very short metatarsal
I failing to contact the proximal end of the metatarsus,

      (
Marsh, 1884 +  Gauthier, 1986). The members
of 
premaxillary teeth with an asymmetrical cross section
     
the X cranial nerve foramen in the occipital surface, a
mediolaterally enlarged retroarticular process, a lateral
surangular shelf, an enlarged axial intercentrum, axial
pleurocoels, cervical pleurocoels inside fossae, the
scapular blade not expanding distally, the absence of the
ulnare and of distal carpals 3 and 4, a reduced ischial
peduncle of ilium, the lateral margin of the femoral
head that is squared in proximal view, a rounded medial
      

crest, a semilunate fossa at the base of the astragalar
     
face of astragalus.
:  (Zuolong Choiniere
et al., 2010 + [Chilesaurus Novas et al., 2015 +
 Sereno et al., 1994]). This analysis found
Zuolong and the enigmatic Chilesaurus as the most basal
members of the tetanuran lineage. Zuolong has been
considered among the basal coelurosaurs, although on
the basis of analyses relatively less-sampled among non-
coelurosaurs, or rooted on allosauroids (e.g., Choiniere et
al., 2010; Brusatte et al., 2014). The latter interpretation
is supported by the Bayesian analysis, where Zuolong
is recovered (although with weak support) among the
basalmost coelurosaurs. A coelurosaurian placement for
Chilesaurus is also supported in the Bayesian analysis.
Tetanuran synapomorphies in the parsimony-based
topology include the loss of the lacrimal shelf over the
antorbital fossa, the contact between the lateral ridge
and the lateral condyle in the quadrate, the dorsoventral
compression of the anterior cervical centra, the reduction
of the supracetabular shelf covering the anterodorsal
corner of acetabulum, the medial perforation of the pubic
apron, a medially-directed femoral head, the reduction of
 
trochlea of femur.
 (Chilesaurus + .
      
    Chilesaurus and dismisses
ornithischian relationships suggested by Baron & Barrett
(2017). This node is diagnosed by two unambiguous
synapomorphies: the extensor sulcus on femur and the
absence (due to secondary loss) of the femoral mediodistal
crest.
:  (
Huene, 1914 +  Huene, 1914). In the
parsimony-based scenario, neotetanuran synapomorphies
absent in Chilesaurus and Zuolong are the anterior
placement of the narial margin of premaxilla, anterior
presacral vertebrae with convex anterior facet, ventral
placement of metacarpal III relative to II, more gracile and



oriented condyles of astragalus, proximodistally longer
ascending process of astragalus, relatively stouter
metatarsal I, median constriction of proximal surface of
metatarsal III.
: . The analysis
recovered a series of “compsognathid-grade” forms
along a paraphyletic series leading to 
Sereno, 1999. Aorun Choiniere et al., 2013, resulted the
basalmost coelurosaur (with the possible exception of
Chilesaurus and Zuolong, see above). The basalmost
node of  under this topology is
supported unambiguously by four apomorphies: distinct
posteroventral process of lacrimal, distal surface of
pubic foot subrectangular, posterior part of pubic foot
elongate, distal half of metatarsal IV shaft contacting
metatarsal III.
    : (“compsognathid grade”
+ ). The parsimony analysis found
a paraphyletic series of small-bodied coelurosaurs
(“compsognathid-like” forms) as forming a pectinate
series leading to tyrannoraptorans. On the contrary, these
forms are united in a clade (Compsognathidae) in the
topology found by the Bayesian analysis. Coelurosaurs
with the exclusion of Aorun are diagnosed by the medially
opened maxillary recess, the elongation of the cervical
centra beyond the posterior level of the neural arch, fan-

the proximal end of tibia (reversal to the plesiomorphic
theropodan condition), and the absence of the anterior
distal fossa in the tibia (reversal to the plesiomorphic
dinosauriform condition).
    : (Sinocalliopteryx Ji et al.,
2007 + ). This clade of coelurosaurs
9
A. Cau - Avian Body Plan
is characterised by an elongate external naris, a straight
antorbital fossa margin on the ventral ramus of lacrimal,
the enclosed anteroventral foramen of the splenial, and a

    : .
Tyrannosauroids and maniraptoromorphs are diagnosed
by elongate posterolateral processes of nasal, the
prefrontal participating in the anterodorsal margin of
orbit, sub-rectangular dorsal neural spines (reversal to
the plesiomorphic coelurosaurian condition), cervical
centra not extended beyond the neural arches (reversal to
the plesiomorphic coelurosaurian condition), “T”-shaped
middle chevrons, steeply-inclined acromion on scapula,
elongate posterior process of coracoid, and the elongate
humeral diaphysis.
  :  (new
clade).
Etymology - From Maniraptora (Gauthier, 1986) and
-morpha (Greek, shaped like).
Denition - The most inclusive clade containing Vultur
gryphus Linnaeus, 1758, and excluding Tyrannosaurus
rex Osborn, 1905.
Remarks - All theropod phylogenies published in
the last 15 years have supported a clade of coelurosaurs
excluding tyrannosauroids and including maniraptorans
and ornithomimosaurs. Although most studies restricted
this clade to the node-based 
Holtz, 1994, the here-named 
represents a more inclusive, branch-based clade that may
include also non-maniraptoriform coelurosaurs (e.g.,
Ornitholestes Osborn, 1903, coelurids, and eventually
some compsognathid-grade forms). Maniraptoromorph
synapomorphies include keel or carinae in the postaxial
cervical centra, absence of hyposphene-hypantra
Fig. 4 - The assembly of the avian body plan is described by about 1500 morphological transitions. One among the shortest trees found by
the parsimony analysis. Branch length based on character optimisation (ambiguous apomorphies optimised using accelerated transformation).
Silhouettes based on artworks by Marco Auditore, Davide Bonadonna, Lukas Panzarin and the author.
Bollettino della Società Paleontologica Italiana, 57 (1), 2018
10
in caudal vertebrae (reversal to the plesiomorphic
theropodan condition), a prominent dorsomedial process
on the semilunate carpal, a convex ventral margin of
the pubic foot, a subrectangular distal end of tibia, and
a sulcus along the posterior margin of the proximal end

    : (Ornitholestes +
). This clade of coelurosaurs is

prominent anterior dorsal hypapophyses, a distinct
pronator muscle scar on radius, proximal pubic shaft that
is wider than deep, an enclosed pubic apron (reversal to
the plesiomorphic tetanuran condition), and a reduced
fourth trochanter.
    : 
( Barsbold, 1976 + 
Gauthier, 1986). Maniraptoriforms differ from other
tyrannoraptorans in having a relatively shallower snout,
the shorter external naris (reversal to the plesiomorphic
coelurosaurian condition), the anteroposteriorly shorter
        
frontal, the absence of distinct paradental laminae, a
straight anterior margin of the pubic peduncle of ilium
(reversal to the plesiomorphic coelurosaurian condition),
and a reduced proximal surface of metatarsal III that is not
constricted at mid-length (reversal to the plesiomorphic
neotetanuran condition).
    : 
( Bonaparte, 1991 + 
Foth et al., 2014). The results of the parsimony
and Bayesian analyses differ in the placement of
therizinosaurs relative to the other maniraptoriforms. In
the parsimony analysis, therizinosaurs are sister-taxon
of the oviraptorosaurs, whereas in the Bayesian analysis
they are the earliest diverging branch of Maniraptora.
The maniraptoran clade is diagnosed by the elongate
dorsal postzygapophyses, the shorter middle caudal
prezygapophyses, the narrower coracoid facet on the
scapular acromion, the proximodistal elongation of
the ventral tuber of the humerus, the relative lateral
extension of the semilunate carpal over metacarpal II,
Fig. 5 - Modularity in the assembly of the avian body. Absolute (a) and relative (b) amount of change based on the unambiguous apomorphies
Archaeopteryx and modern
birds.
11
A. Cau - Avian Body Plan
the longitudinal reduction of the supracetabular crest,
the reduction of the brevis shelf on ilium, the less convex
shape of the ventral margin of pubic foot (reversal to
the plesiomorphic maniraptoromorph condition), the
enlargement of the proximal obturator notch in the
ischium, and the absence of the ischial symphysis.
    : 
( Barsbold, 1976 + 
Sereno, 1997). Pennaraptora is a well-supported
clade of maniraptorans, including paravians and
oviraptorosaurs. Although the parsimony analysis
recovered therizinosauroids among  (as
sister-taxon of “core” oviraptorosaurs), another analysis
using a larger taxon sample and the same character
sample used here found therizinosauroids as sister-
group of pennaraptorans (Cau et al., 2017). Among

in the placement of the enigmatic scansoriopterygids,
found, in the former, among basal avialans, whereas
they are placed as the basalmost oviraptorosaurs in the
Bayesian analysis.
Unambiguous synapomorphies of  (or
of the pennaraptoran-therizinosauroid clade) include the
reduction of the orbital margin of prefrontal (reversal to
the plesiomorphic tyrannoraptoran condition), the ventral
displacement of the base of the paroccipital processes
relative to the occipital condyle, the development of
the surangular lateral shelf (convergent with more
basal averostrans), the short cervical neural spines,
the pleurocoels extended back to the anterior dorsal
vertebrae, the presence of a distinct median ridge on the
ulnar cotyle, the laterally-bowed ulna, the presence of a
distinct third distal carpal (reversal to the plesiomorphic
averostran condition), the postacetabular blades that
diverge posteriorly, the blade-like ischial shaft (reversal
to the plesiomorphic dinosauriform condition), the
absence of the ischial foot (reversal to the plesiomorphic
dinosaurian condition), and barely-bowed metatarsal V.
: . This analysis found
an unresolved polytomy at the paravian root, including
Fukuivenator Azuma et al., 2016, two dromaeosaurid
lineages and the troodontid-avialan clade. Exploration
of the results shows that the unresolved polytomy is
Fukuivenator (found,
alternatively, as a dromaeosaurid or as the basalmost
paravian). A posteriori pruning of Fukuivenator
dromaeosaurid monophyly. The latter topology is
used for character optimisation. An intriguing result
of this analysis is the unenlagiine-halszkaraptorine
sister-group relationships: although an analysis using
the same character sample with a larger taxon sample
among paravians does not support this hypothesis (Cau
et al., 2017), these two dromaeosaurid subclades show
adaptations related to a piscivorous diet (Gianechini et
al., 2011; Cau et al., 2017). Paravian synapomorphies
include relatively smaller infratemporal fenestra, carotid
processes in cervical vertebrae, fusion of the sacral neural
spines, elongation of the middle-caudal centra, loss of the
middle-caudal neural spines, reduction of the number of
caudal ribs, relatively lower scapular acromion, lateral
  
the posterior process of coracoid, inclusion of distal carpal

the manual unguals, shallower cuppedicus fossa on ilium,
development of the processus supratrochantericus on
ilium, absence of the anterior process of the pubic foot,
relatively shorter ischium, absence of the posteroventral
process of the calcaneum, development of a posterolateral
 
digit, and development of pennaceous feathers on ulna
and metatarsus.
      Agnolín &
Novas, 2013 ( Gilmore, 1924 + 
Gauthier, 1986). Troodontids, anchiornithids, and birds
(eventually including scansoriopterygids, but see result
of the Bayesian analysis) share a common ancestry
excluding dromaeosaurids. The troodontid-avialan node
is based on several apomorphies: premaxillary teeth with
round to elliptical cross section, anterodorsally inclined
lacrimal, a medially inset ventral ramus of lacrimal,
vaulted frontals and parietals, reduced supratemporal
fossae not extended onto the frontals, the absence of
the squamosal-quadratojugal contact, a depressed crista
interfenestralis in the middle ear, a dorsoventrally elongate
foramen magnum, absence of the fossa housing cranial
nerves X and XII, a posteriorly deepening lateral groove
of dentary, a vestigial coronoid, marked reduction of
middle caudal postzygapophyses, the absence of contact
between scapular acromion and coracoid, the reduced
bicipital scar in the deltopectoral crest, a relatively short
ilium, a median dorsal process of ischium, a proximally
narrowing femoral diaphysis, a more lateral placement of

tubercles in pedal unguals III and IV.
    : . In the parsimony
analysis, the recently-established anchiornithid and
scansoriopterygid groups result closer to birds than
any “traditional” maniraptoran clade. The unresolved
basal avialan tricotomy is unambiguously supported
by relatively shortened nasals, a marked reduction in
number and size of the proximal caudal neural spines,
the humerus shaft subequal in thickess to the femur, the
posteriorly concave ischium, a reduced cnemial crest, the
penultimate phalanx in the third toe not shorter than the
preceding phalanges.
: (Archaeopteryx + more crown-
ward avialans). Archaeopteryx is recovered as closer to
modern birds than anchiornithids and scansoriopterygids.
The “traditional” basal node of birds is supported by the
posterior elongation of the nasal process of premaxilla, the
participation of the maxilla in the margin of the external
naris, the subvertical ventral ramus of lacrimal (reversal
to the plesiomorphic averaptoran condition), the absence
of the surangular lateral shelf, the anterior projection of
the scapular acromion, the relatively more robust furcula,
          
manual digit II, the pubic peduncle of ilium longer than the
acetabulum, the posteriorly extended cuppedicus fossa, the

IV, and the absence of pennaceous feathers on metatarsus
(reversal to the plesiomorphic paravian condition).
Bollettino della Società Paleontologica Italiana, 57 (1), 2018
12
Fig. 6 - Maximum Clade Credibility resulted in the Bayesian analysis. Colour of branches according to posterior probability.
13
A. Cau - Avian Body Plan
 (Rahonavis Forster et al., 1998
+ more crown-ward avialans). The enigmatic paravian
Rahonavis is here recovered as a basal avialan, not related
to unenlagiines (Makovicky et al., 2005). Supports for
this placement are the lateral excavations and pneumatic
foramina on the dorsal centra, the presence of six sacral
vertebrae, the marked brachial scar in the ulna, the
thickened anterior proximal ridge on ulna, the relative
elongation of the ilium (reversal to the plesiomorphic
paravian condition), the marked proximal compression of
the pubic shaft, the posteriorly concave pubis, a straight
ischium (reversal to the plesiomorphic averaptoran
condition), the fusion between anterior and greater
trochanters, and the relatively short metatarsus.
: (Balaur Csiki et al., 2010 +
more crown-ward avialans). Another enigmatic paravian,
Balaur, is found as an avialan closer to short-tailed birds
than Archaeopteryx (see discussion in Cau et al., 2015).
This placement is based on the presence of seven sacral
vertebrae, the reduction of the supracetabular crest on
ilium, the marked posteroventral direction of the pubis,
the presence of a lateral longitudinal ridge on ischium, the
extension of the distal articular surface of the tibiotarsus
along the posterior surface, and the proximal fusion of
the metatarsals.
    : ( Zhou &
Zhang, 2006 +  Chatterjee, 1997). The long-
tailed jeholornithids are the closest relatives of short-tailed
avialans. This relationships is based on several derived
features, including the absence of the subglenoid fossa of
coracoid, the prominent humeral ectepicondyle, the ulna
more robust than the tibiotarsus, the proximally expanded

the shortened penultimate phalanx in manual digit II
(reversal to the plesiomorphic neotetanuran condition),

(reversal to the plesiomorphic paravian condition), and the

phalanx of the second toe (reversal to the plesiomorphic
paravian condition).
    : . This clade
includes all short-tailed birds, and is diagnosed by the
relatively elongate preantorbital ramus of maxilla, the
elongate posterodorsal process of lacrimal, acuminate
dentary tip, widely-spaced dentition, markedly backturned
posterior sacral ribs, short mid-caudal vertebrae, fusion
of the distalmost caudal vertebrae, anteriorly-restricted
humeral condyles, reduced pubic apron, obliteration
of the suture between tibia and astragalar ascending
process, distinctly ginglymoidal distal end of metatarsal
II, pedal toes II and IV subequal in length (reversal to
the plesiomorphic paravian condition), pedal ungual
II not larger than pedal unguals III and IV (reversal to
the plesiomorphic paravian condition), and the elongate
penultimate phalanx in pedal digit IV.
    : (Confuciusornis Hou et al.,
1995 +  Chiappe & Calvo, 1994).
Pygostylians with the exclusion of sapeornithids share an
elongate premaxillary facet on the anteromedial margin
of nasal, the absence of the ascending process of jugal, a
grooved lateral surface of furcula, a constricted coracoid
neck, a well-developed bicipital tubercle on ulna, a narrow
       
metatarsal II, a well-developed musculus tibialis cranialis
insertion tuber on metatarsal III.
    : .
Ornithothoracine birds include enantiornithines and
ornithuromorphs. The phylogenetic placement of
Protopteryx Zhang & Zhou, 2000, relative to other
ornithothoracines is ambiguous in the parsimony
analysis, being it found alternatively as the basalmost
enantiornithine or as sister-taxon of 
The Bayesian analysis supports the former alternative. This
clade is diagnosed by the shallow snout (reversal to the
condition evolved in ), the rod-like
suborbital bar of jugal, the upturned retroarticular process,
less than 12 dorsal vertebrae, a narrow interclavicular
angle, the carinate sternum, the dorsal placement of the
scapular blades on the ribcage, a mobile scapulocoracoid
joint, the straight/convex sternal margin of coracoid,
the relatively distal placement of the scapular facet on
coracoid, the absence of a distinct dorsomedial process
of the semilunate carpal (reversal to the plesiomorphic
maniraptoromorph condition), the fusion of metacarpal I
with the semilunate carpal, the reduced second phalanx in
the third manual digit, the distally narrowing penultimate

pubic peduncle of ilium (reversal to the plesiomorphic
avialan condition), the comparable anterior projection
of both tibiotarsal condyles, the elongate medial dorsal
process on ischium, and the alula.
:  Chiappe
et al., 1999. All avialans closer to extant birds than
enantiornithines belong to the ornithuromorphan clade.
The analysis found an unresolved polytomy among
basalmost ornithuromorphs and a clade comprising
hongshanornithids and more crown-ward birds. The
unambiguous synapomorphies of 
include the relatively enlarged premaxillary body,
the loss of the hypocleidum, the relatively elongate
sternum, a posteriorly extended sternal carina, the
elongate posteromedial processes of sternum, reduced
flexor processes on manual unguals (reversal to the
plesiomorphic paravian condition), the absence of
the median dorsal process of ischium (reversal to the
plesiomorphic averaptoran condition), and a relatively
short penultimate phalanx in pedal digit IV (reversal to
the plesiomorphic pygostylian condition).
    : (
       
This clade of ornithuromorphs is diagnosed by the
      
presence of teeth in the anterior end of dentary, the
       
the intercotylar eminence of the metatarsus, the latero-
plantar displacement of distal end of metatarsal II
relative to metatarsal III, the ventral placement of the
proximal end of metatarsal III relative to metatarsals II
and IV (convergently acquired by several coelurosaurian
Bollettino della Società Paleontologica Italiana, 57 (1), 2018
14
15
A. Cau - Avian Body Plan
lineages), the distal end of metatarsal I that does not reach
the level of the distal end of metatarsal II, the fusion

the distal vascular foramen.
    : ( Zhou &
Zhang, 2001 + more crown-ward avialans). This clade is
diagnosed by a relatively blunt anterior tip of the dentary
(reversal to the plesiomorphic pygostylian condition),
more than ten dentary teeth, a pair of fenestrae in the
posterior end of the sternum, the presence of distinct
plantar ridges on the distal tibiotarsal shaft, the distal
fusion between metacarpals II and III.
    : (Piscivoravis Zhou et al.,
2013 + more crown-ward avialans). Unambiguous
synapomorphies of this clade are the heterocoelous
condition in all cervical vertebrae, and the development
of the extensor (patellar) sulcus on the distal end of the
femur.
    : (Gansus yumenensis Hou
& Liu, 1984, Iteravis Zhou et al., 2014, 
Merrem, 1813). The clade including all ornithuromorphs
closer to moder birds than Piscivoravis is diagnosed
        
the absence of a lateral sulcus on the clavicular rami
(reversal to the plesiomorphic pygostylian condition),
the unexpanded distal end of the posterolateral processes
of sternum, the prominent and upturned lateral cnemial
        
fourth pedal digit.
 (Ichthyornis Marsh,
1872,  Agnolín & Martinelli,
2009,  Haeckel, 1866). The clade of
ornithuromorphs closer to (extant) avians than Gansus-
like forms is diagnosed by the intermetacarpal process on
metacarpal II, the posteroventral orientation of the pubic
peduncle of ilium, the relatively enlarged ischial peduncle
of ilium, the absence of the pubic symphysis, the presence
 
subequal in length to pedal digit II, and the relatively small
pedal ungual IV compared to pedal ungual III.
    :  (Hesperornis
Marsh, 1872 + ). Ornithurine birds are diagnosed
by heterocoelous dorsal vertebrae, a relatively low
deltopectoral crest on humerus, a marked anterior
(ambiens) expansion in the proximal end of pubis, and a
proximally projected femoral neck.
 :  (Vegavis Clarke et al.,
2005 + Meleagris). Unambiguous synapomorphies of
modern birds, here represented by the extant Meleagris,
and also present in the Cretaceous Vegavis include eleven
sacral vertebrae, the supratendineal bridge on distal
tibiotarsus, and distinct sulci on the hypotarsus.
Tempo and mode in the Assembly of the Avian Body Plan
Using the strict consensus of the shortest trees found
by the parsimony analysis, the sequence of character
acquisition along the 38 nodes of the ASL includes
348 unambiguous morphological state transitions. This
number of events represents the minimal value of the
actual sequence of changes, as it is based solely on the
unambiguously optimised apomorphies. Nevertheless,
as stated above, I have refrained from including
additional morphological transitions based on character
optimisations (i.e., accelerated or delayed optimisations),
because these approaches may spuriously include in the
sequence some evolutionary events that instead occurred
along other branches of the total avian group, not in the

in some parts of the sequence (see Figs 4-5). The amount
of changes per internode ranges between 2 and 38, with
a median value of 8. Among the 348 evolutionary events
that minimally describe the assembly of the avian body
plan, 75 (22% of the total) pertain to the skull, 28 (8%
of the total) to the presacral vertebral column, 20 (6%
of the total) to the caudosacral vertebral column, 71
(20% of the total) to the pectoral limb, 149 (43% of the

integumentary system. The number of events inferred
for the integumentary system is very small, and is not
included in the analysis of the modular evolution. Among
the 348 unambiguously optimised transitions, 32 events
(9% of the whole sequence) are interpreted as reversals
to the states lost in more inclusive nodes. The amount of
reversals in the internodes is not related to the amount
 S = 0.30, p = 0.06). The
amount of changes per node in each anatomical region is
not correlated to that in other regions, with the possible
exception of the caudosacral and pelvic limb regions
couple, that shows a moderately positive correlation (S =
0.42, p = 0.009). The latter result may indicate that these
two modules represent sub-units of a larger module.
The rate of character acquisition per node along the
ASL is relatively uniform. The slope of the incremental
curve plotting the total amount of changes gained does not
show particular variations, with only the Ornithoscelida-
Theropoda transition showing a higher increase relative
to the rest of the lineage (Fig. 5a). Similar trends are
evidenced comparing the change gain trends in the distinct
anatomical regions. Yet, the incremental curves of the
       
which may indicate a modular evolutionary pattern (Fig.
5b). Modularity in the evolution of the avian body plan is
expressed here in term of heterogeneity in the OG values
of the distinct anatomical regions, compared to the whole


(from the tree root to the theropod root), they diverge
between node 7 and node 21 (from the theropod root to
the paravian root), then converge progressively, with a
complete overlap along the terminal eight nodes (from
the ornithutomorph node to the avian crown group). This
Fig. 7 - Tempo and mode of the ASL. Same tree as in Fig. 6, with colour of branches according to median rate of divergence (probability to
observe one state transition per million year).
Bollettino della Società Paleontologica Italiana, 57 (1), 2018
16

caudosacral vertebral and pectoral limb regions relative
to the other regions: for most of the central part of the
series, these two regions show OG values consistently
lower than those of the other regions.
The analysis of the tempo and mode of character
evolution (i.e., inferring the timing of cladogenesis
and estimating the rate of change along the geological
time) using a topology based on parsimony analysis
may be biased by the particular method used for the a
posteriori stratigraphic calibration of the phylogenetic
tree (see discussion in Lee et al., 2014a); accordingly, the
discussion of the evolutionary dynamics is here based on
the results of the Bayesian analysis, which simultaneously
co-estimated relationships and timing of cladogenesis (Lee
et al., 2014a, b; Figs 6-7). Although the topology resulted

topology used for character optimisation (derived from
the parsimony analysis), the most strongly robust areas
in the two alternative topologies broadly overlap, and will
be the focus of the discussion.
Using those nodes shared by the parsimony-based
and Bayesian-based analyses, a minimum series of well-

matrix of these ancestors was converted to an Euclidean
distance matrix and subjected to Principal Coordinate
       
PCoA were used to describe the trajectory of the ASL in
the morphospace during the Mesozoic. The 3-dimensional

       

coelurosaurian nodes, with the lowest Axis 1 values,
the second from the maniraptoromorph node to the
ornithothoracine root, characterised by the progressive
increase of both Axis 1 and 2 values, and the third phase,
from the ornithuromorph ancestry to the origin of the
crown group, showing a substantial decrease of the Axis
2 values (Fig. 8).
DISCUSSION
The investigation of bird origins has often focused
on a few “key features” (e.g., feathers and the
musculoskeletal
those ecological scenarios that may drive the evolution
         
ground-dwelling vs. an arboreal lifestyle; Ostrom, 1976;
Padian & Chiappe, 1998; Dececchi & Larsson, 2013; Xu
et al., 2014). This approach has emphasised the role of a
subset of features and taxa over other factors, biasing our

leading to the modern avian bauplan. Such bias has
been exacerbated by the contingent nature of the fossil
record. For over a century, almost all discussions on
bird ancestry have focused on Archaeopteryx alone.
Even after the recognition of the dinosaurian ancestry of
birds, most of the discussion on the evolutionary patterns
      
iconic Urvogel and a limited set of “Archaeopteryx-like”
taxa (i.e., deinonychosaurian theropods; e.g., Ostrom,
1976). The application of a less restrictive paradigm of
bird evolution, including the whole stem lineage in the
analysis of the avian-like novelties (e.g., Gatesy & Dial,
1996a, b; Hutchinson, 2001a, b; Dececchi & Larsson,
2013), has represented the most productive innovation
in the study of bird evolution (Prum, 2002; Xu et al.,
2014). From both phylogenetic and palaeontological
perspectives, the particular lineage reconstructed here
is one among the several evolutionary trajectories that
form the Mesozoic history of the avian total group.

perspective is that this lineage is the only one of that
clade that survived the Cretaceous-Palaeogene boundary

any particular subset of this lineage is shared with
other Mesozoic pan-avians, the complete sequence
reconstructed here leads exclusively to extant birds,
and, retrospectively, it describes the unique sequence
of evolutionary events that assembled the modern bird
bauplan. It is therefore legit to “extract” that particular
trajectory from the branching topology of the pan-
avian clade and to discuss its properties as a linear,
historical process, focusing on those elements pivotal
in interpreting the modern avian biology.
Fig. 8 - Morphospace distribution of the avian ancestors. Selected
series of avian ancestors inferred by both parsimony and Bayesian
analyses. In a), ancestors plotted according to inferred age of nodes
  

disparity. Bubble size proportional to divergence rate. Note that
the ancestors are distributed along three relatively narrow regions.
17
A. Cau - Avian Body Plan
The amount of character state transitions along the
ASL is relatively uniform. In none of the internal branches
the number of inferred novelties is unusually high, and this
suggests that the character sample adequately describes
the whole disparity in the taxonomic sample used. On the
contrary, the evolutionary rate (amount of changes per
million year per branch) inferred along the ASL is not
homogeneous. In particular, the analysis inferred two main
phases when the rate of morphological divergence along

of the whole tree (i.e., higher than in 95% of all branches
in all sampled trees, a rate estimated in this analysis as >
4.78% of changes per million year): along the basalmost
internodes of the stem lineage (from the pan-avian root to
the node ) and in most of the coelurosaurian
internodes of the avian stem, with the exclusion of the least
Fig. 9 - The three main phases in the assembly of the avian body plan. Above, plot of the avian ancestors relative to the rate of morphological
divergence and cladogenetic median age, both inferred by the Bayesian analysis. Horizontal red bar indicates upper limit of background
rate of divergence from all branches of all sampled trees (4.87%). Roman numbers indicate the phases. Below, historical series indicating
the origin and evolution of the main features of the avian bauplan documented in the fossil record. Single-headed arrows indicate possible
causal relationships, double-headed arrows denote possible co-evolving features. Silhouettes based on artworks by Davide Bonadonna and
Lukas Panzarin.
Bollettino della Società Paleontologica Italiana, 57 (1), 2018
18
inclusive nodes including, respectively, Archaeopteryx
and Piscivoravis, and the scion including the four most
crown-ward nodes (Fig. 9). An unusually high rate of
divergence along part of the ASL was obtained in the
analysis of Lee et al. (2014b), although that analysis found
the highest values among the basal tetanuran internodes.
It should be remarked that in their analysis, Lee et al.
(2014b) enforced the age of the neotetanuran node at
about 175 Mya, and that such constraint may have biased

radiated from the enforced node. In the current analysis,
only the age of the root of the tree was enforced. In this
analysis, the age of the root of the tree was constrained
as not older than the Permian-Triassic boundary: this age
setting may explain the higher values inferred along the
basalmost nodes. No age constraints were enforced for
the tetanuran branches; thus the high evolutionary rate
inferred along the coelurosaurian internodes of the avian
stem is interpreted as a genuine evolutionary phenomenon.
Morphospace occupation along the ASL further
supports a non-homogeneous evolutionary history of the
bird ancestors. As described above, the trajectory linking
      


occurred among the ancestral coelurosaurs toward the
maniraptoromorph origin, while the second transition
occurred at the origin of the ornithothoracines. These
transitions in morphospace occupation closely match the


coelurosaurs and the other among basal ornithothoracines).

distinct phases along the ASL evolutionary trajectory, each

ecologies, locomotory modules and rates of morphological
divergence.
Pan-avian phase I: the Huxleyian stage (~ 245-185 Mya)
The last common ancestor of birds and crocodiles
was probably a predator with ziphodont dentition, a
quadrupedal semi-erect posture, and a scaly integument
including osteoderms, as in most Triassic archosauriforms
(Nesbitt et al., 2017): these have to be assumed as the
main features of the ancestral body plan at the beginning

the avian body plan spans about 60 million years, from
the Early Triassic to the Middle Jurassic. This phase, here
named “Huxleyian” in honor of Thomas Henry Huxley
(1825-1895), includes the pan-avian series leading to the
last common ancestor of the maniraptoromorph theropods.
About 40-45% of the key avian features, still shared by
modern birds, were acquired during this phase. In the
integumentary system, the osteoderms of the ancestral
archosaurs were lost relatively early among the basalmost

was acquired before the origin of the last ornithoscelidan
common ancestor (Godefroit et al., 2014), and then
elaborated among the earliest coelurosaurian ancestors
in basally branching feather-like structures (Ji et al.,
1998, 2007). During this phase, the presacral vertebral
series is progressively pneumatised by diverticula of the
respiratory system, following a pattern that is repeated
during the ontogeny of living birds (King, 1957). In the
tail skeleton, bird ancestors acquired a distinct “transition
point”, abruptly marking the regionalisation between
a proximal and more mobile region and a distal and
more rigid region: this morphology also characterises,
to an extreme level, the tail of modern birds. In the
locomotory system, the most distinctive bird adaptation
among living tetrapods (obligate bipedalism) is acquired
during this phase (Hutchinson, 2001a, b). Both hindlimb
and pelvis progressively developed an erect parasagittal
posture and a digitigrade and functionally-tridactyl pes
before the origin of the last common ancestor of all
theropods. The acquisition of a fully bipedal stance in
these hypercarnivorous forms allowed the hand to evolve
a grasping function: this predatory function selected
      
      
and V: at the end of this stage the tridactyl hand, another
key feature of the birds among modern vertebrates, had
acquired its fundamental structure.
Pan-avian phase II: the Ostromian stage (~ 185-145 Mya)
The second main phase in the evolution of the avian
body plan spans 40 million years, during the second half
of the Jurassic. This phase, here named “Ostromian” in
honor of John Harnold Ostrom (1928-2005), includes all
avian ancestors from the origin of maniraptoromorphs to
the last common ancestor of the pygostylian birds. This
       
rate of morphological divergence than in the rest of the
ASL, and a dramatic increase in the number of avian-like
features acquired (OG raises from 40 to 90) corresponding
to about half of the whole set of apomorphies evolved

trend observed during the whole phase, and presumably
started at the end of the previous phase, is a sustained
body miniaturisation which drove the accumulation
of paedomorphic features (Lee et al., 2014b). It is
particularly intriguing that many of the most successful
theropod clades of the Cretaceous (i.e., ornithomimosaurs,
alvarezsauroids, oviraptorosaurs, dromaeosaurids and
troodontids) are sister taxa of subsets of the lineage
evolved during the Ostromian stage. All these lineages are
inferred to originate during the Middle Jurassic (Lee et
al., 2014b). This relatively rapid morphological radiation
may be explained as the result of the “exploration” of
novel ecological regions, previously not occupied by
theropods (Zanno & Makovicky, 2011; Lautenschlager
et al., 2013; Lautenschlager, 2014). In particular, most
of the avian ancestors along the Ostromian stage are
inferred to lack the majority of the mandibular and tooth
features related to hypercarnivory and macrophagy,
and instead widespread among non-maniraptoriform
theropods (Zanno & Makovicky, 2011). This supports the
hypothesis that the ancestral hypercarnivorous ecology of
most archosaurs, retained during the Huxleyian stage, was
replaced in this second phase by an omnivorous ecology.
Furthermore, both encephalisation ratios and braincase
anatomy support an expansion and re-organisation of
the central nervous system during the Ostromian stage

In the appendicular system, a sustained elongation of the
forelimb is documented along the entire phase, being it
19
A. Cau - Avian Body Plan
particularly dramatic in the terminal internodes, when
the forelimb exceeds the hindlimb in both length and
robustness (Dececchi & Larsson, 2013). It has been shown

body miniaturisation (Dececchi & Larsson, 2013; Lee
et al., 2014b). In the hindlimb, the most important trend
observed during this phase is the progressive reduction
of the hindlimb retraction muscles, which represent the
main locomotory module in all limbed (non-avian) reptiles
(Gatesy & Dial, 1996a; Hutchinson, 2001a, b). This trend
culminated in the extreme reduction of the size of the tail at
the end of this phase, reduced to a very short element at the
root of pygostilians (Chiappe et al., 1999). Although often

the pygostyle is merely the extreme stage of a general
     
the distal half of the tail (Gatesy & Dial, 1996a). The
multiple evolution of pygostyle-like structures among the
non-volant maniraptoriforms and the absence of modern

2017) suggest that the origin of the pygostyle in the last
phase of the Ostromian series was merely a by-product
 
one possible explanation of the dramatic atrophy of the
caudofemoral musculature at the end of the Ostromian
stage may be non-adaptive: the combination of allometric
and structural factors in a miniaturised theropod with
     
most impressive reduction in the size of the pelvis bones
(in particular, in the area of the postacetabular part of
the ilium and the length of the ischium) is observed in
anchiornithids and some basal avialans (Godefroit et al.,
2013a, b; Lefèvre et al., 2017): these theropods probably
had, compared to body size, the smallest surface areas
for the origin of the retractor hindlimb muscles among
all dinosaurs (Hutchinson, 2001a, b), a peculiar condition
that requires further scrutiny.
What factors drove the modular re-organisation of the
appendicular system? While allometry may explain the

2013), the significantly enlarged forelimb and the

Ostromian stage appear as unambiguously related to the

1998; Chiappe et al., 1999). The precise optimisation
of these features along the sequence is complicated by
    
jeholornithids, sapeornithids and confuciusornithids
(Zhou & Zhang, 2006). Although an arboreal ecology
in the early internodes of the Ostromian stage is not
supported by morphometric analysis (Dececchi &
Larsson, 2011), it may have played an important role at the
end of this stage (among the internodes more crownward
than those shared with Archaeopteryx and Balaur).
Body size miniaturisation and reduction of the hindlimb
retractor muscles may indicate that during the last part of
the Ostromian stage, the avian ancestors adapted to more
densely vegetated ecotones, including arboreal settings,
that did not require the cursorial adaptations widespread
along most of the preceding internodes. This scenario is
supported by the unambiguous scansorial adaptations

hallux; Chiappe et al., 1999). Both the overall reduction in
adult body size and the possible exploration of scansorial/
arboreal ecologies during this phase co-evolved with
(or co-opted) a progressive elaboration of the plumage.
During the Ostromian stage, the avian ancestors acquired
and then elaborated the pennaceous feathers (Ji et al.,
1998; Prum, 1999; Lefévre et al., 2017). It is useful to
compare feather complexity and distribution, on one side,
and locomotory adaptations, on the other side, along the
series of the avian ancestors inferred in this phase. Based
on the known distribution of the pennaceous plumage
among the maniraptorans, this novel type of feather
appeared initially only on the distal end of the forelimb
and on the distal end of the tail in cursorial/ground-
dwelling forms (Ji et al., 1998; Zelenitsky et al., 2012).
Distinct lines of evidence suggest that sexual selection and
reproductive functions may have driven the origin of the
pennaceous structures in the forelimb and tail (Zelenitsky
et al., 2012; Persons et al., 2014). A more extensive
distribution of pennaceous feathers, along the whole
forearm, most of the tail, and the hindlimb, is inferred
exclusively in more crown-ward avian ancestors (among
the paravians), characterised by a smaller adult body size,
a relative reduction of the hindlimb musculature, and
incipient scansorial adaptations (e.g., unguals in both
fore- and hindlimb showing a marked falciform shape,
and relatively longer forelimbs). Assuming that these
small-bodied theropods were able, even incipiently, to
exploit arboreal environments, the selection of plumage
elaboration due to its passive aerodynamic function
(i.e., parachuting) cannot be ruled out. This stage may
precede the evolution of a fully-developed wing with
asymmetric feathers, inferred in the last nodes of this
phase (Chiappe, 1995; Chiappe et al., 1999; Zhou &
Zhang, 2006). Following this scenario, the progressive
adult size miniaturisation, the elaboration of sexually-

acquisition of scansorial habits, all co-evolved through
a positive feedback along the avialan internodes of the
Ostromian stage.
The lineage leading exclusively to Archaeopteryx
is the sister-taxon of one of the internodes along
the Ostromian stage. Thus, the evolutionary point
traditionally considered the boundary between birds and
“non-birds” is placed along the Ostromian stage. Yet,
the internode represented by the last common ancestor
of Archaeopteryx and birds (node that is often used to

divergence in morphospace occupation, compared to the
adjacent nodes along the ASL. Its historical meaning
aside, once analysed using a large-scale morphological
and taxonomic sampling, Archaeopteryx does not mark
any peculiar evolutionary shift toward the origin of
       
the first unambiguous flight-related adaptations are
inferred along the last nodes of the Ostromian stage,
after the divergence of Archaeopteryx-grade avialans
(Padian & Chiappe, 1998; Dececchi & Larsson, 2011).
        
unambiguous flight-adaptations (e.g., Jeholornis,
Sapeornis, confuciusornithids; Padian & Chiappe, 1998;
Chiappe et al., 1999; Senter, 2006; Zhou & Zhang, 2006)
are controversial. It should be remarked that even if
potentially adapted to scansorial or arboreal ecologies,
Bollettino della Società Paleontologica Italiana, 57 (1), 2018
20
the most advanced members of the Ostromian stage
(at the root of Pygostylia) lack unambiguous features
      
(Senter, 2006). Although the ongoing debate on the basal
     
(Padian & Chiappe, 1998; Senter, 2006; Dececchi &
Larsson, 2011, 2013), all alternative scenarios agree
in considering the last node of the Ostromian stage as
represented by miniaturised theropods (body mass < 1
kg; Lee et al., 2014b) with an omnivorous ecology and
a series of appendicular and integumentary adaptations
       
dinosaurs, the arboreal environment.
Pan-avian phase III: the Marshian stage (~ 145-85 Mya)
The last main phase in the assembly of the avian
body plan is exclusively Cretaceous in age, and spans
from the origin of ornithothoracines to the last common
ancestor of all living birds (inferred to be mid- or Late
Cretaceous in age; Lee et al., 2014a). This phase is
named “Marshian”, honoring Othniel Carles Marsh
    
       
(Marsh, 1872, 1880). During about 60 million years,
the bird ancestors acquired 10-15% of the features
     bauplan. This relatively
smaller amount of apomorphies, compared in particular
to the evolutionary transformation occurred during
the shorter Ostromian stage, is also expressed by the

of this phase in the Bayesian analysis. In particular,

morphological evolution is inferred to drop from the high
values estimated during the Ostromian stage to values
comparable to the background rate of the rest of the tree
(Fig. 9). Morphospace occupation in this phase mirrors
the trend in the divergence rates, showing that the avian
ancestors along the most basal ornithuromorph nodes
were subjected to a remarkable shift along the second
main axis inferred in the PCoA.
 
       
(in particular, during the earliest part of this phase),
and a general simplification of the skeleton due to
loss or fusion of elements. The largest majority of the

the very beginning of this phase, once the avian ancestors
had acquired the full set of adaptations allowing them,
     
(Chiappe & Calvo, 1994; Chiappe, 1995; Zhang & Zhou,
2000). These include, among others, the expansion of
the sternum and the development of a midline keel,
the radical transformation of the coracoid in a strut-
like bar, the re-location of the scapulae on the dorsal
surface of the ribcage, paralleling the dorsal vertebral

       
of any predatory or grasping function in the hand, and
the development of the alula (Chiappe & Calvo, 1994;
Chiappe, 1995; Padian & Chiappe, 1998; Zhang &

this radiation, the atrophied tail was inherited relatively
        


the tail is co-opted to its modern function, as a third
locomotory module that is independent from the
musculoskeletal system of the hindlimb (Gatesy & Dial,
1996b).
      
the Cretaceous phase is the general co-ossification
or loss of many skeletal elements, in particular in the
skull (including the complete loss of dentition), in the
thoraco-sacral vertebrae, and in the metapodial elements
(formation of the carpometacarpus and tarsometatarsus)
(Padian & Chiappe, 1998; Clarke & Chiappe, 2001;
Clarke & Norell, 2002; Clarke, 2004). Although a causal
 

cannot be ruled out, the fusion of previously-distinct
elements is reported also in non-volant theropod lineages
(e.g., Carrano & Sampson, 2008; Cau et al., 2015). It is
       
is documented along the mammalian stem lineage, and
may represent a general trend in tetrapod evolution
(Sidor, 2001).
Innovation, reduction and exaptation
Several key features that characterise the modern
avian bauplan are modification of innovations that
evolved before the Marshian phase. These exaptations
    
under the Huxleyian and Ostromian stages, presumably

those exploited for the actual functions. For example,
although the tail was part of the locomotory module in the
ancestral pan-avians, there is not unambiguous evidence
that the shortened tail of the earliest pygostylians (at the
end of the Ostromian stage) retained some locomotory
function: later, during the Marshian phase, this organ
was co-opted to a novel locomotory module, among
the ornithuromorphs (Gatesy & Dial, 1996a; Wang &
   
articulation that provides mobility to the alular feathers
is the exaptation of the hyper-extendable articulation
(related to a predatory use of the forelimb) acquired
during the earliest internodes of the Huxleyian stage,
after the evolution of the fully-bipedal posture (Galton,
1971; Zhang & Zhou, 2000; Senter & Robins, 2006).
       
relatively early during the Huxleyian stage, when the
toes acquired a symmetrical and a functionally-tridactyl
  
posture: in that state of “latency” for over 60 million
years, the hallux was then subjected to a limited set of
changes in order to acquire a novel function, grasping
related to arboreality, during the last internodes of the
Ostromian stage (Galton, 1971; Middleton, 2001; Hattori,
2016). Finally, feathers probably evolved as structures
with no aerodynamic function along the early internodes
of the Huxleyian stage (Godefroit et al., 2014), were
then elaborated as complex appendages probably under
a sexual selection regime along the earliest internodes of
the Ostromian phase (Zelenitsky et al., 2012), and then
co-opted as aerodynamically-functional organs at the end
of the same phase (Padian & Chiappe, 1998).
21
A. Cau - Avian Body Plan
CONCLUSION
The evolution of the avian body plan is a 160-million-
year long macroevolutionary process that cannot be

mid-Jurassic, and could not be resolved exclusively to the

bauplan     


scale, that of the morphological features usually targeted
by phylogenetic analyses, the ornithisation was a gradual
and probably stochastic process, as evidenced by the
overall uniformity in the amount of novelties gained
during the assembly of each module. Body size reduction
is a large-scale macroevolutionary trend that is inferred

drove several innovations related to paedomorphosis and
heterochrony, progressively gained along the Ostromian
stage. A by-product of miniaturisation, allometric
        
in producing non-adaptive novelties, subsequently co-
opted as exaptations in later internodes. At a higher scale,
       
each ancestor along the stem lineage, the ornithisation
was a more heterogeneous pattern, as indicated by the

the Middle Jurassic-earliest Cretaceous interval. Such
rate heterogeneity is paralleled by the trajectory linking
the series of ancestors in the morphospace analysis. Both
patterns support a tripartite sequence for the avian history,
into the Huxleyian, Ostromian and Marshian stages: these
   
explored by the ASL during its history.
Phylogenetic taxonomy aims to describe evolutionary
events (the origin of taxa), and thus must be strictly
monophyletic (Gauthier & Padian, 1985). The terminology
  
taxonomy of the avian branch, because it describes the
ordered stages of a process (i.e., the assembly of an
avian body plan). This tripartite subdivision of the avian
evolution is novel in its formulation, and is based on the
explicit rejection of a privileged explicative role to some
particular taxa (e.g., Archaeopteryx, or basal paravians)

adapted forelimbs). This alternative paradigm could not
be properly recognised until we look at bird evolution
under the traditional dichotomy between “birds/avians/
avialans” vs. “non-birds/avians/avialans”, anchored for
over a century to Archaeopteryx    
the arbitrary root of the clade Avialae. The importance
of Archaeopteryx in our understanding of bird evolution

morphological disparity differentiates it from “non-
avian dinosaurs”. While it is now widely recognised that
Archaeopteryx is not the “missing link” between two


“bird clade” to the Jurassic Urvogel. Yet, bird evolution
   
name, and focusing on the evolutionary events placed
around the “Archaeopteryx
of a few internodes over the rest of the assembly process.
Distinct lines of evidence have shown that the debate on
the ecomorphology of the “Archaeopteryx grade” is not
just overrated, it is probably misleading. Placed along the
 
by the last common ancestor of Archaeopteryx and
avians was not a “key discontinuity” in what is indeed
a longer and more complex branching continuum. More
significant ecomorphological transitions occurred in
distinct moments, before and after the Archaeopteryx-
bearing internode of the ornithisation. As a couple of
examples, both obligate bipedalism and feathers, the most

from the other living amniotes, are key innovations of
the Huxleyian stage and evolved under a regime distinct


     
origin of the Archaeopteryx-like forms.
Being it the framework of the “reptile-to-bird”
transition, the pan-avian (avemetatarsalian) radiation
has been polarised into two opposite narrations. One,

pattern among the lineages not leading to birds, that
have been reduced to a series of steps along the avian
body assembly. The other, focusing almost exclusively
on “non-avian dinosaurs”, has implicitly perpetuated the
use of arbitrary grades based on paraphyletic groups, and
  
pan-avians. The terminology introduced here is explicitly
- and exclusively - devoted to bird ancestry (i.e., to the
process that assembled the avian bauplan), but, at the same
time, it avoids to use the same systematic terminology
that refers to branches not involved in the avian body
assembly. Such terminological distinction prevents the
various sister taxa of birds to be inappropriately mentioned
as examples of “stages” along the avian body assembly
(e.g., eudromaeosaurs mentioned as examples of the
“proto-bird” bauplan).
The recognition of a hierarchical structure linking the
various factors involved in bird evolution reinforces, at the
lowest anatomical scale, the unity and continuity between
the “reptilian” and “avian” body plans, and, at the highest
ecological and functional scale, helps in identifying and
interpreting the complex concert of historical factors that
shaped this unique and successful bauplan.
SUPPLEMENTARY ONLINE MATERIAL
All the Supplementary data of this work are available
on the BSPI website at http://paleoitalia.org/archives/
bollettino-spi/
ACKNOWLEDGEMENTS
This idiosyncratic review is a very synthetic resume of
almost 20 years of elaborations on theropod macroevolution and
palaeontological phylogenetics. I thank the Editorial Board of the
Bollettino della Società Paleontologica Italiana for the invitation
to write this paper. Simone Maganuco is thanked for the critical
review of an early phase of this manuscript, that improved the
         
bibliographic material, photographs of specimens and copies of
Bollettino della Società Paleontologica Italiana, 57 (1), 2018
22
their works, used for the much larger phylogenetic analysis that
is the basis of the one published here, I thank S. Maganuco, M.
Auditore, S. Brusatte, M. Carrano, A. Chiarenza, C. Dal Sasso, P.
Godefroit, J. Harris, T. Holtz, C. Hendrickx, U. Lefèvre, D. Madzia,


D. Bonadonna and L. Panzarin.
REFERENCES
Agnolín F.L. & Martinelli A.G. (2009). Fossil birds from the Late
Cretaceous Los Alamitos Formation, Río Negro Province,
Argentina. Journal of South American Earth Sciences, 27:
42-49.
Agnolín F.L. & Novas F.E. (2013). Avian ancestors. A review of
the phylogenetic relationships of the theropods Unenlagiidae,
Microraptoria, Anchiornis and Scansoriopterygidae. 96 pp.
Springer Briefs in Earth System Sciences, Dordrecht Heidelberg
New York London.
Agnolín F.L. & Rozadilla S. (2017). Phylogenetic reassessment of
Pisanosaurus mertii Casamiquela, 1967, a basal dinosauriform
from the Late Triassic of Argentina. Journal of Systematic
Palaeontology, 1-27.
Arcucci A.B. (1986). Nuevos materiales y reinterpretacion de
Lagerpeton chanarensis Romer (Thecodontia, Lagerpetonidae
nov.) del Triassic medio de La Rioja, Argentina. Ameghiniana,
23: 233-242.
Azuma Y., Xu X., Shibata M., Kawabe S., Miyata K. & Imai T.
(2016). A bizarre theropod from the Early Cretaceous of Japan
highlighting mosaic evolution among coelurosaurians. Scientic
Reports, 6: 20478.
        
Evolutionary origins of the avian brain. Nature, 501: 93-96.
Baron M.G. & Barrett P.M. (2017). A dinosaur missing-link?
Chilesaurus and the early evolution of ornithischian dinosaurs.
Biology Letters, 13: 20170220.
Baron M.G., Norman D.B. & Barrett P.M. (2017). A new hypothesis
of dinosaur relationships and early dinosaur evolution. Nature,
543: 501-506.
Barsbold R. (1976). On the evolution and systematics of the late
Mesozoic carnivorous dinosaurs. In Devâtkin E.V. & Ânovskaâ
      Trudy,
Sovmestnaâ Sovetsko - Mongolskaâ paleontologičeskaâ
èkspediciâ, 3: 68-75. [in Russian]
Barsbold R. (1983). “Avian” features in the morphology of
predatory dinosaurs. Transactions of the Joint Soviet Mongolian
Paleontological Expedition, 24: 96-103. [in Russian]
Benedetto J.L. (1973). Herrerasauridae, nueva familia de saurisquios
triasicos. Ameghiniana, 10: 89-102.
Benson R.B.J., Carrano M.T. & Brusatte S.L. (2010). A
new clade of archaic large-bodied predatory dinosaurs
(Theropoda: Allosauroidea) that survived to the latest Mesozoic.
Naturwissenschaften, 97: 71-78.
    
reptiles. Zoological Journal of the Linnean Society, 84: 97-164.
Bonaparte J.F. (1991). Los vertebrados fosiles de la formacion Rio
Colorado, de la ciudad de Neuquen y Cercanias, Cretacico
Superior, Argentina. Revista del Museo Argentino de Ciencias
Naturales “Bernardino Rivadavia”, Paleontologia, 4: 16-123.
Bouckaert R.R., Heled J., Kuehnert D., Vaughan T.G., Wu C.-H.,
Xie D., Suchard M.A., Rambaut A. & Drummond A.J. (2014).
BEAST 2: a software platform for Bayesian evolutionary
analysis. PLOS Computational Biology, 10: e1003537.
Brazeau M.D. (2011). Problematic character coding methods in
Biological Journal of the Linnean
Society, 104: 489-498.
Broom R. (1913). Note on Mesosuchus browni, Watson and on a new
South African Triassic pseudosuchian (Euparkeria capensis).
Records of the Albany Museum, 2: 394-396.
Brusatte S., Lloyd G., Wang S. & Norell M.A. (2014). Gradual
assembly of avian body plan culminated in rapid rates of
evolution across the dinosaur–bird transition. Current Biology,
24: 2386-2392.
Budd G.E. (2001). Tardigrades as “stem-group arthropods”: the
evidence from the Cambrian fauna. Zoologischer Anzeiger,
240: 265-279.
Carrano M.T., Benson R.B.J. & Sampson S.D. (2012). The
phylogeny of Tetanurae (Dinosauria: Theropoda). Journal of
Systematic Palaeontology, 10: 211-300.
Carrano M.T. & Sampson S.D. (2008). The phylogeny of
Ceratosauria (Dinosauria: Theropoda). Journal of Systematic
Palaeontology, 6: 183-236.
Carrano M.T., Sampson S.D. & Forster C.A. (2002). The osteology
of Masiakasaurus knoperi, a small abelisauroid (Dinosauria:
Theropoda) from the Late Cretaceous of Madagascar. Journal
of Vertebrate Paleontology, 22: 510-534.
Casamiquela R.M. (1967). Un nuevo dinosaurio ornitisquio
triásico (Pisanosaurus mertii; Ornithopoda) de la Formación
Ischigualasto, Argentina. Ameghiniana, 4: 47-64.
     
P., Stein K., Barsbold R., Tsogtbaatar K., Currie P.J. &
Godefroit P. (2017). Synchrotron scanning reveals amphibious
ecomorphology in a new clade of bird-like dinosaurs. Nature,
552: 395-399.

of the bizarre Late Cretaceous Romanian theropod Balaur
bondoc
bird? PeerJ, 3: e1032.
Chatterjee S. (1997). The Rise of Birds. 312 pp. John Hopkins
University Press, Baltimore.
Chen P.J., Dong Z. & Zhen S.N. (1998). An exceptionally preserved
theropod dinosaur from the Yixian Formation of China. Nature,
391: 147-152.

Nature, 378: 349-355.
Chiappe L.M. & Calvo J.O. (1994). Neuquenornis volans, a New
Late Cretaceous Bird (Enantiornithes: Avisauridae) from
Patagonia, Argentina. Journal of Vertebrate Paleontology, 14:
230-246.
Chiappe, L.M., Ji S., Ji Q. & Norell M.A. (1999). Anatomy and
systematics of the Confuciusornithidae (Theropoda: Aves) from
the Late Mesozoic of Northeastern China. Bulletin of American
Museum of Natural History, 242: 1-89.
Choiniere J.N., Clark J.M., Forster C.M., Norell M.A., Eberth D.A.,
Erickson G.M., Chu H. & Xu X. (2013). A juvenile specimen of
a new coelurosaur (Dinosauria: Theropoda) from the Middle–
      
Republic of China. Journal of Systematic Palaeontology, 12:
177-215.
Choiniere J.N., Clark J.M., Forster C.A. & Xu X. (2010). A basal
coelurosaur (Dinosauria: Theropoda) from the Late Jurassic

Republic of China. Journal of Vertebrate Paleontology, 30:
1773-1796.
Clarke J.A. (2004). Morphology, phylogenetic taxonomy, and
systematics of Ichthyornis and Apatornis (Avialae: Ornithurae).
Bulletin of the American Museum of Natural History, 286: 1-179.
Clarke J.A. & Chiappe L.M. (2001). A new carinate bird from the
Late Cretaceous of Patagonia. American Museum Novitates,
3323: 1-23.
Clarke J.A. & Norell M.A. (2002). The morphology and phylogenetic
position of Apsaravis ukhaana from the Late Cretaceous of
Mongolia. American Museum Novitates, 3387: 1-46.
Clarke J.A., Tambussi C.P., Noriega J.I., Erickson G.M. & Ketcham
        
radiation in the Cretaceous. Nature, 433: 305-308.
Cope E.D. (1869). Synopsis of the extinct Batrachia, Reptilia
and Aves of North America. Transactions of the America
Philosophical Society, 14: 1-252.
23
A. Cau - Avian Body Plan
Csiki V.M., Brusatte S.L. & Norell M.A. (2010). An aberrant
island-dwelling theropod dinosaur from the Late Cretaceous
of Romania. Proceedings of the National Academy of Sciences
of the United States of America, 107: 15357-15361.
Currie P.J. (1985). Cranial anatomy of Stenonychosaurus inequalis
(Saurischia, Theropoda) and its bearing on the origin of birds.
Canadian Journal of Earth Sciences, 22: 1643-1658.
Dalla Vecchia F.M. (2013). Triassic pterosaurs. Geological Society,
London, Special Publications, 379: 119-156.
Dececchi T.A. & Larsson H.C.E. (2011). Assessing Arboreal
Adaptations of Bird Antecedents: Testing the Ecological Setting
of the Origin of the Avian Flight Stroke. PLoS ONE, 6: e22292.
Dececchi T.A. & Larsson H.C.E. (2013). Body and limb size
dissociation at the origin of birds: uncoupling allometric
constraints across a macroevolutionary transition. Evolution,
67: 2741-2752.
Drummond A.J., Suchard M.A., Xie D. & Rambaut A. (2012).
Bayesian phylogenetics with BEAUti and the BEAST 1.7.
Molecular Biology and Evolution, 29: 1969-1973.

Annual Review of Ecology and Systematics, 15: 1-24.
Forster C.A., Sampson S.D., Chiappe L.M. & Krause D.W. (1998).
Genus correction. Science, 280: 179.
Foth C., Tischlinger H. & Rauhut O.W.M. (2014). New specimen
of Archaeopteryx provides insights into the evolution of
pennaceous feathers. Nature, 511: 79-82.
Galton P.M. (1971). Manus movements of the coelurosaurian
dinosaur Syntarsus and opposability of the theropod hallux.
Arnoldia, 15: 1-8.
Gatesy S.M. & Dial K.P. (1996a). Locomotor modules and the
Evolution, 50: 331-340.
Gatesy S.M. & Dial K.P. (1996b). From frond to fan: Archaeopteryx
and the evolution of short-tailed birds. Evolution, 50: 2037-2048.
Gauthier J.A. (1986). Saurischian monophyly and the origin of birds.
In Padian K. (ed.), The Origin of Birds and the Evolution of
Flight. Memoirs of the California Academy of Sciences: 1-47.
Gauthier J. & Padian K. (1985). Phylogenetic, functional, and
In
Hecht M.K., Ostrom J.H., Viohl G. & Wellnhofer P. (eds), The
Beginnings of Birds, Freunde des Jura-Museums, Eichstatt:
185-197.
Gavryushkina A., Welch D., Stadler T. & Drummond A.J. (2014).
Bayesian inference of sampled ancestor trees for epidemiology
and fossil calibration. PLOS Computational Biology, 10:
e1003919.
Gianechini F.A., Makovicky P.J. & Apesteguía S. (2011). The teeth
of the unenlagiine theropod Buitreraptor from the Cretaceous
of Patagonia, Argentina, and the unusual dentition of the
Gondwanan dromaeosaurids. Acta Palaeontologica Polonica,
56: 279-290.
Gilmore C.W. (1924). A new coelurid dinosaur from the Belly River
Cretaceous of Alberta. Canada Geological Survey Bulletin,
38: 1-12.
Godefroit P., Cau A., Hu D., Escuillié F., Wenhao W. & Dyke G.J.
(2013b). A Jurassic avialan dinosaur from China resolves the
early phylogenetic history if birds. Nature, 498: 359-362.
Godefroit P., Demuynck H., Dyke G.J., Hu D., Escuillié F. & Claeys

paravian theropod from China. Nature Communications, 4:
1394.
Godefroit P., Sinitsa S.M., Dhouailly D., Bolotsky Y.L., Sizov A.V.,
McNamara M.E., Benton M.J. & Spagna P. (2014). A Jurassic
ornithischian dinosaur from Siberia with both feathers and
scales. Science, 345: 451-455.
 
for phylogenetic analysis. Cladistics, 24: 774-786.
Gould S.J. & Vrba E.S. (1982). Exaptation, a missing term in the
science of form. Paleobiology, 8: 4-15.
Haeckel E. (1866). Generelle Morphologie der Organismen. 462
pp. Georg Reimer, Berlin.
Harris J.D. (2004). Confusing dinosaurs with mammals: tetrapod
phylogenetics and anatomical terminology in the world of
homology. The Anatomical Record Part A, 281: 1240-1246.
Hattori S. (2016). Evolution of the hallux in non-avian theropod
dinosaurs. Journal of Vertebrate Paleontology, 36: e1116995.

A Reply to Ernst Mayr. Systematic Zoology, 24: 244-256.
Hicks R.D. (1925). Lives of the Eminent Philosophers. By Laërtius,
Diogenes. 956 pp. Translated by Hicks, Robert Drew, Loeb
Classical Library.
Holtz T.R. Jr. (1994). The phylogenetic position of the
Tyrannosauridae. Implications for theropod systematics.
Journal of Paleontology, 68: 1100-1117.
Holtz T.R. Jr. (2000). A new phylogeny of the carnivorous dinosaurs.
Gaia, 15: 5-61.
Holtz T.R. Jr. (2001). Arctometatarsalia Revisited: The Problem of
Homoplasy in Reconstructing Theropod Phylogeny. In Gauthier
J. & Gall L.F. (eds), New Perspectives on the Origin and Early
Evolution of Birds, Proceedings of the International Symposium
in Honor of John H. Ostrom: 99-122.
Holtz T.R. Jr., Molnar R.E. & Currie P.J. (2004). Basal Tetanurae.
In Weishampel D.B., Dodson P. & Osmolska H. (eds). The
Dinosauria. Second edition: 71-110. University of California
Press, Berkeley.
Hou L. & Liu Z. (1984). A new fossil bird from Lower Cretaceous
of Gansu and early evolution of birds. Scientia Sinica Series
B, 27: 1296-1302.
Hou L., Zhou Z., Gu Y. & Zhang H. (1995). Description of
Confuciusornis sanctus. Chinese Science Bulletin, 10: 61-63.
Huene F. (1914). Das natürliche System der Saurischia. Centralblatt
fur Mineralogie, Geologie und Paläontologie, 1914: 154-158.
Huene F. (1932). Die fossile Reptil-Ordnung Saurischia, ihre
Entwicklung und Geschichte. Monographien zur Geologie und
Paläontologie, 4: 1-361.
Hutchinson J.R. (2001a). The evolution of pelvic osteology and
soft tissues on the line to extant birds (Neornithes). Zoological
Journal of the Linnean Society, 131: 123-168.
Hutchinson J.R. (2001b). The evolution of femoral osteology and
soft tissues on the line to extant birds (Neornithes). Zoological
Journal of the Linnean Society, 131: 169-197.
Huxley T.H. (1868). On the animals which are most nearly
intermediate between birds and reptiles. Annals and Magazine
of Natural History, 4: 66-75.
   
observations on the Dinosauria of the Trias. Quarterly Journal
of the Geological Society, 26: 32-51..
Ji Q., Currie P.J., Norell M.A. & Ji S.A. (1998). Two feathered
dinosaurs from northeastern China. Nature, 393: 753-761.
Ji Q. & Ji S.A. (1996). On the discovery of the earliest bird fossil
in China and the origin of birds. Chinese Geology, 233: 30-33
Ji S., Ji Q., Lu J. & Yuan C. (2007). A new giant compsognathid
dinosaur with long filamentous integuments from Lower
Cretaceous of Northeastern China. Acta Geologica Sinica,
81: 8-15.
King A.S. (1957). The aerated bones of Gallus domesticus. Acta
Anatomica, 31: 220-230.
Langer M.C. & Benton M.J. (2006). Early dinosaurs: A phylogenetic
study. Journal of Systematic Palaeontology, 4: 309-358.
Langer M.C., Ezcurra M.D., Bittencourt J.S. & Novas F.E. (2010).
The origin and early evolution of dinosaurs. Biological Reviews,
85: 55-110.
Larsson H.C.E., Sereno P.C. & Wilson J.A. (2000). Forebrain
enlargement among nonavian theropod dinosaurs. Journal of
Vertebrate Paleontology, 20: 615-618.
Lautenschlager S. (2014). Morphological and functional diversity
in therizinosaur claws and the implications for theropod
claw evolution. Proceedings of the Royal Society B, 281:
20140497.
       
(2013). Edentulism, beaks, and biomechanical innovations in
Bollettino della Società Paleontologica Italiana, 57 (1), 2018
24
the evolution of theropod dinosaurs. Proceedings of the National
Academy of Sciences of the USA, 110: 20657-20662.
Lee M.S.Y., Cau A., Naish D. & Dyke G.J. (2014a). Morphological
clocks in palaeontology, and a mid-Cretaceous origin of crown
Aves. Systematic Biology, 63: 442-449.
Lee M.S.Y., Cau A., Naish D. & Dyke G.J. (2014b). Sustained
miniaturization and anatomical innovation in the dinosaurian
ancestors of birds. Science, 34: 562-566.
Lefèvre U., Cau A., Cincotta A., Hu D., Chinsamy A., Escuillie F.
& Godefroit P. (2017). A new Jurassic theropod from China
documents a transitional step in the macrostructure of feathers.
The Science of Nature, 104: 74.
Lewis P.O. (2001). A likelihood approach to estimating phylogeny
from discrete morphological character data. Systematic Biology,
50: 913-925.
Linnaeus C. (1758). Systema naturæ per regna tria naturæ, secundum

synonymis, locis. 1 (10th ed.). 824 pp. Laurentius Salvius,
Stockholm.

trees: application to mosasauroid nomenclature. PeerJ, 5: e3782.
Makovicky P.J., Apesteguia S. & Agnolín F.L. (2005). The earliest
dromaeosaurid theropod from South America. Nature, 437:
1007-1011.
Makovicky P.J. & Zanno L.E. (2011). Theropod Diversity and the
In Dyke G. & Kaiser G.
(eds), Living Dinosaurs: The Evolutionary History of Modern
Birds, First Edition: 9-29.
Mantell G. (1827). Illustrations of the geology of Sussex: a general
view of the geological relations of the southeastern part of

Forest. 92 pp. Fellow of the Royal College of Surgeons, London.
Marsh O.C. (1872). Preliminary description of Hesperornis regalis,
with notices of four other new species of Cretaceous birds.
American Journal of Science, 3rd ser., 3: 359-365.
Marsh O.C. (1880). Odontornithes: a monograph on the extinct
toothed birds of North America. United States Geological
Exploration of the 40th Parallel. 201 pp. U.S. Government

Marsh O.C. (1881). Principal characters of American Jurassic
dinosaurs. American Journal of Science, 21: 417-423.

reptiles. Nature, 31: 68-69.
Martínez R.N., Sereno P.C., Alcober O.A., Colombi C.E., Renne
P.R., Montañez I.P. & Currie B.S. (2011). A basal dinosaur
from the dawn of the dinosaur era in southwestern Pangaea.
Science, 331: 206-210.
Merrem B. (1813). Tentamen systematis naturalis avium.
Abhandlungen Konigel. (Preussische) Akademie der
Wissenschaften Berlin, 1812-1813 (1816) (Physikal): 237-259.
Meyer von H. (1861). Vogel-Federn und Palpipes priscus von
Solenhofen. Neues Jahrbuch für Mineralogie, Geognosie
Geologie und Petrefakten-Kunde, 1861: 561.
Middleton KM. (2001). The morphological basis of hallucal
orientation in extant birds. Journal of Morphology, 250: 51-60.
Naish D., Dyke G., Cau A., Escuillié F. & Godefroit P. (2012).
A gigantic bird from the Upper Cretaceous of Central Asia.
Biology Letters, 8: 97-100.
Nesbitt S.J., Butler R.J., Ezcurra M.D., Barrett P.M., Stocker M.R.,
      
G., Sennikov A.G. & Charig A.J. (2017). The earliest bird-line
archosaurs and the assembly of the dinosaur body plan. Nature,
544: 484-487.
Nopcsa B.F. (1928). The genera of reptiles. Palaeobiologica, 1:
163-188.
Norell M.A., Clark J.M. & Makovicky P.J. (2001). Phylogenetic
relationships among coelurosaurian theropods. In Gauthier J.
& Gall L.F. (eds), New Perspectives on the Origin and Early
Evolution of Birds, Proceedings of the International Symposium
in Honor of John H. Ostrom: 49-68.
Novas F.E. (1992). Phylogenetic relationships of basal dinosaurs,
the Herrerasauridae. Palaeontology, 63: 51-62.
Novas F.E., Salgado L., Suárez M., Agnolín F.L., Ezcurra M.N.D.,
Chimento N.S.R., de la Cruz R., Isasi M.P., Vargas A.O. &
Rubilar-Rogers D. (2015). An enigmatic plant-eating theropod
from the Late Jurassic period of Chile. Nature, 522: 331-334.
        
ornithuromorph (Aves: Ornithothoraces) bird from the Jehol
Group indicative of higher-level diversity. Journal of Vertebrate
Paleontology, 30: 311-321.
Osborn H.F. (1903). Ornitholestes hermanni, a new compsognathoid
dinosaur from the Upper Jurassic. Bulletin of the American
Museum of Natural History, 19: 459-464.
Osborn H.F. (1905). Tyrannosaurus and other Cretaceous
carnivorous dinosaurs. Bulletin of the American Museum of
Natural History, 21: 259-265.
Ostrom J.H. (1976). Archaeopteryx and the origin of birds.
Biological Journal of the Linnean Society, 8: 91-182.
Owen R. (1842). Report on British fossil reptiles. Report of the
British Association of Advanced Sciences, 9: 60-204.
Padian K. & Chiappe L.M. (1998). The origin and early evolution
of birds. Biological Reviews, 73: 1-42.
Paul G.S. (2002). Dinosaurs of the Air. 460 pp. The Johns Hopkins
University Press, Baltimore and London.
Persons W.S. IV, Currie P.J. & Norell M.A. (2014). Oviraptorosaur
tail forms and functions. Acta Palaeontologica Polonica, 59:
553-567.
Prum R.O. (1999). Development and evolutionary origin of feathers.
Journal of Experimental Zoology, 285: 291-306.
Prum R.O. (2002). Why ornithologists should care about the
theropod origin of birds. Auk, 119: 1-17.
Rauhut O.W.M. (2003). The interrelationships and evolution of basal
theropod dinosaurs. Special Papers in Palaeontology, 69: 1-213.
Rieppel O. (2005). Proper names in twin worlds: Monophyly,
paraphyly, and the world around us. Organisms Diversity and
Evolution, 5: 89-100.
Romer A.S. (1971). The Chanares (Argentina) Triassic reptile
fauna. X. Two new but incompletely known long-limbed
pseudosuchians. Breviora, 378: 1-10.
         
commonly named Dinosauria. Proceedings of the Royal Society
of London, 43: 165-171.
Senter P. (2006). Scapular orientation in theropods and basal birds,
Acta Palaeontologica Polonica,
51: 305-313.
Senter P. & Robins J.H. (2006). Range of motion in the forelimb
of the theropod dinosaur Acrocanthosaurus atokensis, and
implications for predatory behaviour. Journal of Zoology,
266: 307-318.
Sereno P.C. (1997). The origin and evolution of dinosaurs. Annual
Review of Earth & Planetary Sciences, 25: 435-489.
Sereno P.C. (1999). The evolution of dinosaurs. Science, 284:
2137-2147.
Sereno P.C. (2007). Logical basis for morphological characters in
phylogenetics. Cladistics, 23: 565-587.
Sereno P.C. & Arcucci A.B. (1994). Dinosaurian precursors from
the Middle Triassic of Argentina: Marasuchus lilloensis, gen.
nov. Journal of Vertebrate Paleontology, 14: 53-73.
Sereno P.C., Beck A.L., Dutheil D.B., Gado B., Larsson H.C.E.,
Lyon G.H., Marcot J.D., Rauhut O.W.M., Sadleir R.W., Sidor
C.A., Varricchio D.D., Wilson G.P. & Wilson J.A. (1998). A
long–snouted predatory dinosaur from Africa and the evolution
of spinosaurids. Science, 282: 1298-1302.
Sereno P.C., Dutheil D.B., Iarochene M., Larsson H.C.E., Lyon
G.H., Magwene P.M., Sidor C.A., Varricchio D.J. & Wilson
J.A. (1996). Predatory dinosaurs from the Sahara and Late
Science, 272: 986-991.
Sereno P.C., Forster C.A., Larsson H.C.E., Dutheil D.B. & Sues
H.-D. (1994). Early Cretaceous dinosaurs from the Sahara.
Science, 266: 267-271.
25
A. Cau - Avian Body Plan
Sereno P.C., Wilson J.A. & Conrad J.L. (2004). New dinosaurs
link southern landmasses in the Mid-Cretaceous. Proceedings:
Biological Sciences, 271: 1325-1330.
   
evolution. Evolution, 55: 1419-1442.
Smith N.D., Makovicky P.J., Agnolín F., Ezcurra M., Pais D. &
Salisbury S. (2008). A Megaraptor-like theropod (Dinosauria:
Tetanurae) in Australia; support for faunal exchange across
eastern and western Gondwana in the mid-Cretaceous.
Proceedings of the Royal Society, Series B, 275: 2085-2093.

of the pygostyle and tail feathers in Early Cretaceous birds.
Vertebrata PalAsiatica, 10: 1-26.
Xu X., Qi Z., Norell M., Sullivan C., Hone D., Erickson G.,Wang
X., Han F. & Guo Y. (2009). A new feathered maniraptoran

Chinese Science Bulletin, 54: 430-435.
Xu X., Zhou Z., Dudley R., Mackem S., Chuong C.M., Erickson
G.M. & Varricchio D.J. (2014). An integrative approach to
understanding bird origins. Science, 346: 1253-1293.
Zanno L.E. & Makovicky P.J. (2011). Herbivorous ecomorphology
and specialization patterns in theropod dinosaur evolution.
Proceedings of the National Academy of the United States of
America, 108: 232-237.
Zelenitsky D.K., Therrien F., Erickson G.M., Debuhr C.L.,
  
Non-Avian Dinosaurs from North America Provide Insight
into Wing Origins. Science, 338: 510-514.
Zhang F. & Zhou Z. (2000). A Primitive Enantiornithine Bird and
the Origin of Feathers. Science, 290: 1955-1959.

an ornithuromorph dominated locality of the Jehol Group.
Chinese Science Bulletin, 59: 5366-5378.
Zhou Z. & Zhang F. (2001). Two new ornithurine birds from the
Early Cretaceous of western Liaoning, China. Chinese Science
Bulletin, 46: 1258-1264.
Zhou Z. & Zhang F. (2006). Mesozoic birds of China – a synoptic
review. Vertebrata Palasiatica, 44: 74-98.
         
ornithuromorph from the Jehol Biota. Historical Biology,
26: 608-618.
Manuscript received 19 March 2018
Revised manuscript accepted 4 April 2018
Published online 3 May 2018
Editor Annalisa Ferretti
Andrea Cau is a vertebrate palaeontologist. After his MSc
in Natural Sciences, in 2003, he has collaborated with various
palaeontological institutions, including the Natural History
Museum in Milan, the “Giovanni Capellini” Museum in Bologna,
and the Royal Belgian Institute of Natural Sciences in Bruxelles. In
2013, he returned to the Academia and obtained his PhD in Earth,
Life and Environmental Sciences, in 2017. He is author of over
40 publications, and coauthored the institution of 17 fossil taxa,
from foraminiferans to birds. His research has focused mainly on
Mesozoic archosaurs, Italian fossil reptiles, the origin of birds,
theropod phylogenetics, and the integration of stratigraphic and
morphological information in evolutionary palaeontology.
... Previous studies (e.g. [13,15,19,25,[39][40][41][42][43][44][45][46][47][48]) described the evolutionary transition of the femoral head orientation as follows. The femoral head of plesiomorphic saurians (grade 'I' in figure 4) had a small angle with the distal end reference (the intercondylar line). ...
Full-text available
Article
Significant evolutionary shifts in locomotor behaviour often involve comparatively subtle anatomical transitions. For dinosaurian and avian evolution, medial overhang of the proximal femur has been central to discussions. However, there is an apparent conflict with regard to the evolutionary origin of the dinosaurian femoral head, with neontological and palaeontological data suggesting seemingly incongruent hypotheses. To reconcile this, we reconstructed the evolutionary history of morphogenesis of the proximal end of the femur from early archosaurs to crown birds. Embryological comparison of living archosaurs (crocodylians and birds) suggests the acquisition of the greater overhang of the femoral head in dinosaurs results from additional growth of the proximal end in the medial-ward direction. On the other hand, the fossil record suggests that this overhang was acquired by torsion of the proximal end, which projected in a more rostral direction ancestrally. We reconcile this apparent conflict by inferring that the medial overhang of the dinosaur femoral head was initially acquired by torsion, which was then superseded by mediad growth. Details of anatomical shifts in fossil forms support this hypothesis, and their biomechanical implications are congruent with the general consensus regarding broader morpho-functional evolution on the avian stem.
... The White Rock spinosaurid was included in a comprehensive phylogenetic matrix derived from Cau (2018) and implemented in Barker et al. (2021) Scores for five character statements concerning the caudal vertebrae of the two operational taxonomic units (OTUs) Baryonyx (NHMUK PV R 9951) and Riparovenator (IWCMS 2020.447.1, 2) were changed relative to the analysis in Barker et al. (2021). For Baryonyx, these changes related to the caudal neural arch characters (Ch.) 358, 359, 868 and 1576. ...
Full-text available
Article
Postcranial elements (cervical, sacral and caudal vertebrae, as well as ilium, rib and limb bone fragments) belonging to a gigantic tetanuran theropod were recovered from the basal unit (the White Rock Sandstone equivalent) of the Vectis Formation near Compton Chine, on the southwest coast of the Isle of Wight. These remains appear to pertain to the same individual, with enormous dimensions similar to those of the Spinosaurus holotype and exceeding those of the largest European theropods previously reported. A combination of features—including the presence of spinodiapophyseal webbing on an anterior caudal vertebra—suggest that this is a member of Spinosauridae, though a lack of convincing autapomorphies precludes the identification of a new taxon. Phylogenetic analysis supports spinosaurid affinities but we were unable to determine a more precise position within the clade weak support for a position within Spinosaurinae or an early-diverging position within Spinosauridae were found in some data runs. Bioerosion in the form of curved tubes is evident on several pieces, potentially related to harvesting behaviour by coleopteran bioeroders. This is the first spinosaurid reported from the Vectis Formation and the youngest British material referred to the clade. This Vectis Formation spinosaurid is unusual in that the majority of dinosaurs from the Lower Cretaceous units of the Wealden Supergroup are from the fluviolacustrine deposits of the underlying Barremian Wessex Formation. In contrast, the lagoonal facies of the upper Barremian–lower Aptian Vectis Formation only rarely yield dinosaur material. Our conclusions are in keeping with previous studies that emphasise western Europe as a pivotal region within spinosaurid origination and diversification.
... In its original description, Bu. schultzi was placed as the sister taxon to all other sauropodomorphs (Cabreira et al. 2016). This hypothesis has been corroborated by following studies that employed modified versions of that dataset (Müller et al. 2017b Müller and Garcia 2020), as well as different datasets (Cau 2018;2018c;Baron and Williams 2018;Ezcurra et al. 2020a;Müller 2020). In the case of Pol et al. (2021), that position is shared in a polytomy by P. protos. ...
Chapter
Carnian (Late Triassic) deposits of South America provide the oldest unequivocal dinosaur records worldwide, most of which has been assigned to the sauropodomorph lineage. This includes Eoraptor lunensis, Panphagia protos, and Chromogisaurus novasi, from the Ischigualasto Formation, Argentina, and Saturnalia tupiniquim, Pampadromaeus barberenai, Buriolestes schultzi, and Bagualosaurus agudoensis, from the Santa Maria Formation, Brazil. Here, we demonstrate that their holotypes anatomically differ from one another, supporting the taxonomic validity of the species. In addition, a morphological disparity analysis, with significant statistical support, clustered some of the better-known specimens of E. lunensis, Sat. tupiniquim, and Bu. schultzi, with the respective holotypes. For the latter two taxa, this was corroborated by a specimen-level phylogenetic analysis that also found Ba. agudoensis as the sister taxon to post-Carnian sauropodomorphs. Our results also suggest that Bu. schultzi and E. lunensis are the earliest branching sauropodomorphs and that Sa. tupiniquim and Pam. barberenai are closer to Bagualosauria. A species-level phylogenetic analysis further suggests that Bu. schultzi and E. lunensis form a clade, that Sa. tupiniquim is the sister taxon to Bagualosauria, and that Pan. protos and Ch. novasi are also more highly nested, forming a clade with Pam. barberenai.
... Feathers and filaments were first introduced as character states by Sereno (1999), and further characters have been elaborated on by subsequent authors (e.g. O'Connor, 2009;Cau et al., 2017;Cau, 2018). These characters are particularly poorly represented in such data sets [e.g. ...
Article
Modern birds are typified by the presence of feathers, complex evolutionary innovations that were already widespread in the group of theropod dinosaurs (Maniraptoriformes) that include crown Aves. Squamous or scaly reptilian-like skin is, however, considered the plesiomorphic condition for theropods and dinosaurs more broadly. Here, we review the morphology and distribution of non-feathered integumentary structures in non-avialan theropods, covering squamous skin and naked skin as well as dermal ossifications. The integumentary record of non-averostran theropods is limited to tracks, which ubiquitously show a covering of tiny reticulate scales on the plantar surface of the pes. This is consistent also with younger averostran body fossils, which confirm an arthral arrangement of the digital pads. Among averostrans, squamous skin is confirmed in Ceratosauria (Carnotaurus), Allosauroidea (Allosaurus, Concavenator, Lourinhanosaurus), Compsognathidae (Juravenator), and Tyrannosauroidea (Santanaraptor, Albertosaurus, Daspletosaurus, Gorgosaurus, Tarbosaurus, Tyrannosaurus), whereas dermal ossifications consisting of sagittate and mosaic osteoderms are restricted to Ceratosaurus. Naked, non-scale bearing skin is found in the contentious tetanuran Sciurumimus, possibly ornithomimosaurians (Pelecanimimus) and tyrannosauroids (Santanaraptor), and also on the patagia of scansoriopterygids (Ambopteryx, Yi). Scales are surprisingly conservative among non-avialan theropods compared to some dinosaurian groups (e.g. hadrosaurids); however, the limited preservation of tegument on most specimens hinders further interrogation. Scale patterns vary among and/or within body regions in Carnotaurus, Concavenator and Juravenator, and include polarised, snake-like ventral scales on the tail of the latter two genera. Unusual but more uniformly distributed patterning also occurs in Tyrannosaurus, whereas feature scales are present only in Albertosaurus and Carnotaurus. Few theropods currently show compelling evidence for the co-occurrence of scales and feathers (e.g. Juravenator, Sinornithosaurus), although reticulate scales were probably retained on the mani and pedes of many theropods with a heavy plumage. Feathers and filamentous structures appear to have replaced widespread scaly integuments in maniraptorans. Theropod skin, and that of dinosaurs more broadly, remains a virtually untapped area of study and the appropriation of commonly used techniques in other palaeontological fields to the study of skin holds great promise for future insights into the biology, taphonomy and relationships of these extinct animals.
Article
The systematic position of the unusual Late Jurassic dinosaur Chilesaurus diegosaurezi remains an area of real uncertainty. Originally thought to be an unusual theropod, Chilesaurus has, since its first discovery, been suggested to either be a sauropodomorph dinosaur, an early diverging, ‘transitional’ ornithischian, or a more ‘derived’ neornithischian dinosaur, close to the heterodontosaurs. However, despite these recent fluctuations, the original placement within Theropoda remains the most prevalent phylogenetic hypothesis. This study looks to critically assess these various competing hypotheses regarding Chilesaurus and to analyse various aspects of the anatomical datasets that have placed it within Theropoda, Sauropodomorpha and Ornithischia. The choice of anatomical characters in past studies is considered and, crucially, so too is the choice of operational taxonomic units. By revising the datasets used in the initial analyses of Chilesaurus, one cause of this phylogenetic uncertainty has been clearly identified – outgroup choice. The total absence of members of Ornithischia from the predominantly saurischian-taxon-based datasets used in the original phylogenetic analyses of Chilesaurus is here shown to have had a substantial effect on the systematic position that was first recovered for this taxon. By scoring a selection of basal ornithischian taxa into the primarily saurischian datasets previously used, this study has recovered a different position for the Chilesaurus within Dinosauria, different internal topologies within Theropoda, and even a different arrangement of three major dinosaurian groups. The inclusion of an ornithischian outgroup in an otherwise predominantly saurischian focused dataset has produced results that again suggest that Chilesaurus could be an early diverging member of Ornithischia. What is more, this result was achieved without the addition of any classical ornithischian-like characters to the data. Whilst the results of this analysis do not by any means end the debate on the phylogenetic position of Chilesaurus, or on the fundamental relationships within Dinosauria, they do serve to highlight the current instability of such cryptic taxa in phylogenetic analyses, as well as the need for more comprehensive taxon sampling in such analyses generally. The results further highlight the potentially adverse effects of the omission of certain outgroup taxa on the overall topologies recovered within clades.
Full-text available
Article
A predominantly fish-eating diet was envisioned for the sail-backed theropod dinosaur Spinosaurus aegyptiacus when its elongate jaws with subconical teeth were unearthed a century ago in Egypt. Recent discovery of the high-spined tail of that skeleton, however, led to a bolder conjecture that S. aegyptiacus was the first fully aquatic dinosaur. The 'aquatic hypothesis' posits that S. aegyptiacus was a slow quadruped on land but a capable pursuit predator in coastal waters, powered by an expanded tail. We test these functional claims with skeletal and flesh models of S. aegyptiacus. We assembled a CT-based skeletal reconstruction based on the fossils, to which we added internal air and muscle to create a posable flesh model. That model shows that on land S. aegyptiacus was bipedal and in deep water was an unstable, slow-surface swimmer (<1 m/s) too buoyant to dive. Living reptiles with similar spine-supported sails over trunk and tail are used for display rather than aquatic propulsion, and nearly all extant secondary swimmers have reduced limbs and fleshy tail flukes. New fossils also show that Spinosaurus ranged far inland. Two stages are clarified in the evolution of Spinosaurus, which is best understood as a semiaquatic bipedal ambush piscivore that frequented the margins of coastal and inland waterways. Editor's evaluation This article evaluates the hypothesis that Spinosaurus was a specialized aquatic dinosaur, by developing a CT-based skeletal restoration and examining its hydrodynamic properties. In this reappraisal of the "aquatic hypothesis", new results support the alternative "semi-aquatic hypothesis". This article will be of interest to vertebrate paleontologists and functional morphologists, as well as wider academic and non-academic audiences.
Full-text available
Article
The origin and evolutionary relationships of ornithischian dinosaurs are topics that have undergone a series of substantial revisions. At present there are several competing hypotheses concerning the relationship between Ornithischia and the other principal clades of Dinosauria. Some hypotheses have posited a tree topology within Dinosauria that imply a ‘ghost-lineage’ for Ornithischia (whose representatives make their first unambiguous appearance in the Hettangian) that extends through a substantial portion of Triassic time. In contrast, other hypotheses have placed conventionally Triassic dinosauromorph (stem-lineage Dinosauria) taxa within the clade Ornithischia. Recently, a large-scale phylogenetic analysis recovered an array of taxa, known as ‘silesaurids’, as a paraphyletic assemblage of taxa (referred to in this article using the informal terms silesaurs or silesaurians) on the branch leading to the clade Ornithischia. This latter hypothesis of relationships would account for the apparent absence of Triassic ornithischians, because stem-lineage ornithischians (silesaurs in this article) are exclusively Triassic. However, the analysis that produced this novel topology used a dataset that, in its original form, did not include all early representatives of Ornithischia (sensu lato), and did not incorporate all the anatomical characters that have been suggested to unite Ornithischia with other dinosaurian clades (Theropoda and Sauropodomorpha). Nor did the initial study go on to expand upon some important taxonomic, palaeobiological and evolutionary implications of a topology that links a paraphyletic array of silesaurs to the clade Ornithischia. The present article addresses these latter issues by expansion and re-analysis of the original dataset. The results find further support for the hypothesis that silesaurs comprise a paraphyletic grouping of taxa on the stem of Ornithischia and that successive silesaur taxa acquire anatomical characters anagenetically in a process that culminates in the assembly of what may be described as a ‘traditional’ ornithischian. The overall topology of the consensus tree remains but little changed from the original analysis, despite the addition of new taxa and characters. To provide stability to this area of the tree and to preserve the most important of the relevant taxonomic names, we suggest a revised taxonomic framework for ornithischians that is consistent with this new topology. We retain the name Ornithischia for the total-group (traditional Ornithischia and its stem-lineage), while we resuscitate a name originally proposed by Richard Owen, Prionodontia (= ‘coarse edged teeth’) for the clade containing only the so-called traditional ornithischian (= ‘bird-hipped’) dinosaurs. We also erect Parapredentata as a more exclusive subclade in Ornithischia. This novel taxonomic framework is intended to provide phylogenetic clarity and a degree of stability in Ornithischia and Dinosauria as further analyses and new data continue to refine and re-shape the tree. The data presented in this study represent a stage in our attempt to establish an early dinosaur dataset in which character definitions and character scores are agreed upon and used consistently.
Full-text available
Article
Living birds (Aves) have bodies substantially modified from the ancestral reptilian condition. The avian pelvis in particular experienced major changes during the transition from early archosaurs to living birds1,2. This stepwise transformation is well documented by an excellent fossil record2–4; however, the ontogenetic alterations that underly it are less well understood. We used embryological imaging techniques to examine the morphogenesis of avian pelvic tissues in three dimensions, allowing direct comparison with the fossil record. Many ancestral dinosaurian features2 (for example, a forward-facing pubis, short ilium and pubic ‘boot’) are transiently present in the early morphogenesis of birds and arrive at their typical ‘avian’ form after transitioning through a prenatal developmental sequence that mirrors the phylogenetic sequence of character acquisition. We demonstrate quantitatively that avian pelvic ontogeny parallels the non-avian dinosaur-to-bird transition and provide evidence for phenotypic covariance within the pelvis that is conserved across Archosauria. The presence of ancestral states in avian embryos may stem from this conserved covariant relationship. In sum, our data provide evidence that the avian pelvis, whose early development has been little studied5–7, evolved through terminal addition—a mechanism8–10 whereby new apomorphic states are added to the end of a developmental sequence, resulting in expression8,11 of ancestral character states earlier in that sequence. The phenotypic integration we detected suggests a previously unrecognized mechanism for terminal addition and hints that retention of ancestral states in development is common during evolutionary transitions. The developing pelvis in birds revisits its dinosaurian state before transitioning to the characteristic avian form, providing evidence of terminal addition during evolution.
Full-text available
Article
A new enantiornithine, Musivavis amabilis n. gen. n. sp., is reported from the Lower Cretaceous Jehol Biota in western Liaoning, China. The new taxon is similar to the bohaiornithids in the robust subconical teeth, bluntly expanded omal ends of the furcula, caudolaterally oriented lateral trabeculae with triangular distal ends of the sternum, and a robust second pedal digit. Yet it differs from members of Bohaiornithidae in several features recalling other enantiornithine lineages, such as the acuminate rostral ramus of maxilla, the shape of the coracoid lateral margin, the presence of craniolateral processes on the sternum, the proportions of the manual phalanges, and the unspecialized third pedal ungual phalanx. A comprehensive phylogenetic analysis of Mesozoic birds shows that homoplasy significantly affects the reconstruction of enantiornithine relationships. When all phylogenetic characters are considered of equal weight, Musivavis is reconstructed in a lineage related to a radiation of large-bodied enantiornithines including Bohaiornithidae and Pengornithidae. Alternative scenarios based on progressive downweighting of the homoplastic characters support more basal placements of the pengornithids among Enantiornithes, but do not alter the affinity of Musivavis as a member of the “bohaiornithid-grade” group. UUID: http://zoobank.org/617c7062-21ab-4d33-ae80-4edf5a129683
Preprint
Ichthyornis has long been recognized as a pivotally important fossil taxon for understanding the latest stages of the dinosaur-bird transition, but little significant new postcranial material has been brought to light since initial descriptions of partial skeletons in the 19th Century. Here, we present new information on the postcranial morphology of Ichthyornis from 40 previously undescribed specimens, providing the most detailed morphological assessment of Ichthyornis to date. The new material includes four partially complete skeletons and numerous well-preserved isolated elements, enabling new anatomical observations such as muscle attachments previously undescribed for Mesozoic euornitheans. Among the elements that were previously unknown or poorly represented for Ichthyornis , the new specimens include an almost-complete axial series, a hypocleideum-bearing furcula, radial carpal bones, fibulae, a complete tarsometatarsus bearing a rudimentary hypotarsus, and one of the first-known nearly complete three-dimensional sterna from a Mesozoic avialan. Several pedal phalanges are preserved, revealing a remarkably enlarged pes presumably related to foot-propelled swimming. Although diagnosable as Ichthyornis , the new specimens exhibit a substantial degree of morphological variation, some of which may relate to ontogenetic changes. Phylogenetic analyses incorporating our new data and employing alternative morphological datasets recover Ichthyornis stemward of Hesperornithes and Iaceornis , in line with some recent hypotheses regarding the topology of the crownward-most portion of the avian stem group, and we establish phylogenetically-defined clade names for relevant avialan subclades to help facilitate consistent discourse in future work. The new information provided by these specimens improves our understanding of morphological evolution among the crownward-most non-neornithine avialans immediately preceding the origin of crown group birds.
Full-text available
Article
Maniraptora includes birds and their closest relatives among theropod dinosaurs. During the Cretaceous period, several maniraptoran lineages diverged from the ancestral coelurosaurian bauplan and evolved novel ecomorphologies, including active flight, gigantism, cursoriality and herbivory. Propagation X-ray phase-contrast synchrotron microtomography of a well-preserved maniraptoran from Mongolia, still partially embedded in the rock matrix, revealed a mosaic of features, most of them absent among non-avian maniraptorans but shared by reptilian and avian groups with aquatic or semiaquatic ecologies. This new theropod, Halszkaraptor escuilliei gen. et sp. nov., is related to other enigmatic Late Cretaceous maniraptorans from Mongolia in a novel clade at the root of Dromaeosauridae. This lineage adds an amphibious ecomorphology to those evolved by maniraptorans: it acquired a predatory mode that relied mainly on neck hyperelongation for food procurement, it coupled the obligatory bipedalism of theropods with forelimb proportions that may support a swimming function, and it developed postural adaptations convergent with short-tailed birds.
Full-text available
Article
The transformation from a long reptilian tail to a shortened tail ending in a pygostyle and accompanied by aerodynamic fanning rectrices is one of the most remarkable adaptations of early avian evolution. However, no fossils directly capture this transition, and information regarding the structural morphology and the early evolution of the pygostyle in Mesozoic birds and their integuments is relatively limited. Here we provide a review of the pygostyle morphology of Early Cretaceous birds with comparison to the structure in living birds. This study emphasizes the convergent evolution of distally co-ossified caudal vertebrae in non-avian maniraptorans and early birds. There further exist distinct differences in pygostyle morphology between Sapeornithiformes, Confuciusornithiformes, Enantiornithes, and Ornithuromorpha. The morphology of the pygostyle and rectrices in early ornithuromorphs appear similar to that of extant birds, whereas the pygostyle in more primitive birds does not appear morphologically capable of supporting the rectricial bulbs and musculature necessary to control an aerodynamic fan-shaped tail. The rectricial bulbs and rectricial fan appear to have coevolved with the plough-shaped pygostyle early in the evolution of the Ornithuromorpha. This study also shows that the confuciusornithiform pygostyle was more similar to that of enantiornithines than previously recognized, consistent with the presence of nearly identical ornamental tail feathers in both groups.
Full-text available
Article
Mosasauroid squamates represented the apex predators within the Late Cretaceous marine and occasionally also freshwater ecosystems. Proper understanding of the origin of their ecological adaptations or paleobiogeographic dispersals requires adequate knowledge of their phylogeny. The studies assessing the position of mosasauroids on the squamate evolutionary tree and their origins have long given conflicting results. The phylogenetic relationships within Mosasauroidea, however, have experienced only little changes throughout the last decades. Considering the substantial improvements in the development of phylogenetic methodology that have undergone in recent years, resulting, among others, in numerous alterations in the phylogenetic hypotheses of other fossil amniotes, we test the robustness in our understanding of mosasauroid beginnings and their evolutionary history. We reexamined a data set that results from modifications assembled in the course of the last 20 years and performed multiple parsimony analyses and Bayesian tip-dating analysis. Following the inferred topologies and the 'weak spots' in the phylogeny of mosasauroids, we revise the nomenclature of the 'traditionally' recognized mosasauroid clades, to acknowledge the overall weakness among branches and the alternative topologies suggested previously, and discuss several factors that might have an impact on the differing phylogenetic hypotheses and their statistical support.
Full-text available
Article
Genuine fossils with exquisitely preserved plumage from the Late Jurassic and Early Cretaceous of northeastern China have recently revealed that bird-like theropod dinosaurs had long pennaceous feathers along their hindlimbs and may have used their four wings to glide or fly. Thus, it has been postulated that early bird flight might initially have involved four wings (Xu et al. Nature 421:335–340, 2003; Hu et al. Nature 461:640–643, 2009; Han et al. Nat Commun 5:4382, 2014). Here, we describe Serikornis sungei gen. et sp. nov., a new feathered theropod from the Tiaojishan Fm (Late Jurassic) of Liaoning Province, China. Its skeletal morphology suggests a ground-dwelling ecology with no flying adaptations. Our phylogenetic analysis places Serikornis, together with other Late Jurassic paravians from China, as a basal paravians, outside the Eumaniraptora clade. The tail of Serikornis is covered proximally by filaments and distally by slender rectrices. Thin symmetrical remiges lacking barbules are attached along its forelimbs and elongate hindlimb feathers extend up to its toes, suggesting that hindlimb remiges evolved in ground-dwelling maniraptorans before being co-opted to an arboreal lifestyle or flight.
Full-text available
Article
Pisanosaurus mertii was originally described on the basis of an incomplete skeleton from the early Late Triassic (Carnian) of northern Argentina. It is consistently regarded by most authors as a very basal ornithischian, the sister group of remaining members of the clade. The referral to Ornithischia is based mainly on tooth-bearing bones and tooth morphology. On the other hand, the postcranium is recognized as strikingly plesiomorphic for ornithischians, and even for dinosaurs. The recent description of non-dinosaurian dinosauriforms of the clade Silesauridae having ornithischian-like dentition invites a review of the phylogenetic affinities of Pisanosaurus. In this regard, an overview of the holotype specimen allows a reanalysis of previous anatomical interpretations of this taxon. The phylogenetic analysis presented here suggests that Pisanosaurus may be better interpreted as a member of the non-dinosaurian Silesauridae. It shares with silesaurids reduced denticles on the teeth, teeth fused to maxilla and dentary bone, sacral ribs shared between two sacral vertebrae, lateral side of proximal tibia with a fibular flange, and dorsoventrally flattened pedal ungual phalanges. The present analysis indicates that Pisanosaurus should be removed from the base of the Ornithischia and should no longer be considered the oldest representative of this dinosaurian clade.
Full-text available
Article
The relationship between dinosaurs and other reptiles is well established, but the sequence of acquisition of dinosaurian features has been obscured by the scarcity of fossils with transitional morphologies. The closest extinct relatives of dinosaurs either have highly derived morphologies or are known from poorly preserved or incomplete material. Here we describe one of the stratigraphically lowest and phylogenetically earliest members of the avian stem lineage (Avemetatarsalia), Teleocrater rhadinus gen. et sp. nov., from the Middle Triassic epoch. The anatomy of T. rhadinus provides key information that unites several enigmatic taxa from across Pangaea into a previously unrecognized clade, Aphanosauria. This clade is the sister taxon of Ornithodira (pterosaurs and birds) and shortens the ghost lineage inferred at the base of Avemetatarsalia. We demonstrate that several anatomical features long thought to characterize Dinosauria and dinosauriforms evolved much earlier, soon after the bird–crocodylian split, and that the earliest avemetatarsalians retained the crocodylian-like ankle morphology and hindlimb proportions of stem archosaurs and early pseudosuchians. Early avemetatarsalians were substantially more species-rich, widely geographically distributed and morphologically diverse than previously recognized. Moreover, several early dinosauromorphs that were previously used as models to understand dinosaur origins may represent specialized forms rather than the ancestral avemetatarsalian morphology.
Article
Birds evolved from and are phylogenetically recognized as members of the theropod dinosaurs; their first known member is the Late Jurassic Archaeopteryx, now represented by seven skeletons and a feather, and their closest known non-avian relatives are the dromaeosaurid theropods such as Deinonychus. Bird flight is widely thought to have evolved from the trees down, but Archaeopteryx and its outgroups show no obvious arboreal or tree-climbing characters, and its wing planform and wing loading do not resemble those of gliders. The ancestors of birds were bipedal, terrestrial, agile, cursorial and carnivorous or omnivorous. Apart from a perching foot and some skeletal fusions, a great many characters that are usually considered ‘avian’ (e.g. the furcula, the elongated forearm, the laterally flexing wrist and apparently feathers) evolved in non-avian theropods for reasons unrelated to birds or to flight. Soon after Archaeopteryx, avian features such as the pygostyle, fusion of the carpometacarpus, and elongated curved pedal claws with a reversed, fully descended and opposable hallux, indicate improved flying ability and arboreal habits. In the further evolution of birds, characters related to the flight apparatus phylogenetically preceded those related to the rest of the skeleton and skull. Mesozoic birds are more diverse and numerous than thought previously and the most diverse known group of Cretaceous birds, the Enantiornithes, was not even recognized until 1981. The vast majority of Mesozoic bird groups have no Tertiary records: Enantiornithes, Hesperornithiformes, Ichthyornithiformes and several other lineages disappeared by the end of the Cretaceous. By that time, a few Linnean ‘Orders’ of extant birds had appeared, but none of these taxa belongs to extant ‘families’, and it is not until the Paleocene or (in most cases) the Eocene that the majority of extant bird ‘Orders’ are known in the fossil record. There is no evidence for a major or mass extinction of birds at the end of the Cretaceous, nor for a sudden ‘bottleneck’ in diversity that fostered the early Tertiary origination of living bird ‘Orders’.
Article
The enigmatic dinosaur taxon Chilesaurus diegosuarezi was originally described as a tetanuran theropod, but this species possesses a highly unusual combination of features that could provide evidence of alternative phylogenetic positions within the clade. In order to test the relationships of Chilesaurus, we added it to a new dataset of early dinosaurs and other dinosauromorphs. Our analyses recover Chilesaurus in a novel position, as the earliest diverging member of Ornithischia, rather than a tetanuran theropod. The basal position of Chilesaurus within the clade and its suite of anatomical characters suggest that it might represent a ‘transitional’ taxon, bridging the morphological gap between Theropoda and Ornithischia, thereby offering potential insights into the earliest stages of ornithischian evolution, which were previously obscure. For example, our results suggest that pubic retroversion occurred prior to some of the craniodental and post-cranial modifications that previously diagnosed the clade (e.g. the presence of a predentary bone and ossified tendons). © 2017 The Author(s) Published by the Royal Society. All rights reserved.
Article
For 130 years, dinosaurs have been divided into two distinct clades—Ornithischia and Saurischia. Here we present a hypothesis for the phylogenetic relationships of the major dinosaurian groups that challenges the current consensus concerning early dinosaur evolution and highlights problematic aspects of current cladistic definitions. Our study has found a sister-group relationship between Ornithischia and Theropoda (united in the new clade Ornithoscelida), with Sauropodomorpha and Herrerasauridae (as the redefined Saurischia) forming its monophyletic outgroup. This new tree topology requires redefinition and rediagnosis of Dinosauria and the subsidiary dinosaurian clades. In addition, it forces re-evaluations of early dinosaur cladogenesis and character evolution, suggests that hypercarnivory was acquired independently in herrerasaurids and theropods, and offers an explanation for many of the anatomical features previously regarded as notable convergences between theropods and early ornithischians.