Download Form and function of the shoulder girdle in sauropod dinosaurs : a

Form and Function
of the Shoulder Girdle in Sauropod Dinosaurs
A Biomechanical Investigation with the Aid
of Finite Elements
Dissertation to obtain the degree
Doctor Rerum Naturalium (Dr. rer. nat)
at the Faculty of Biology and Biotechnology
International Graduate School of Biosciences
Ruhr-University Bochum
Department of Zoology and Neurobiology
submitted by
Bianca Hohn-Schulte
from Dortmund
Bochum, April 2010
Form und Funktion
des Schultergürtels bei sauropoden Dinosauriern
Eine biomechanische Untersuchung
mit Hilfe finiter Elemente
Dissertation zur Erlangung des Grades
eines Doktors der Naturwissenschaften
der Fakultät für Biologie und Biotechnologie
an der Internationalen Graduiertenschule Biowissenschaften
der Ruhr-Universität Bochum
angefertigt am
Lehrstuhl für Zoologie und Neurobiologie
von
Bianca Hohn-Schulte
aus Dortmund
Bochum, im April 2010
Tag der Einreichung:
15. April 2010
Tag der mündlichen Prüfung
07. Juni 2010
1. Betreuer
PD Dr. Claudia Distler-Hoffmann
Lehrstuhl für Zoologie und Neurobiologie
Fakultät für Biologie und Biotechnologie
Ruhr-Universität Bochum
2. Betreuer:
Prof. Dr. Wolfgang H. Kirchner
AG Verhaltensbiologie und Didaktik der Biologie
Fakultät für Biologie und Biotechnologie
Ruhr-Universität Bochum
Für Hannah in Liebe
"An anatomist who ignores engineering mechanics
is missing one of the most powerful resources
of human knowledge"
-R. McNeill Alexander-
Contents
CONTENTS ................................................................................................................................................... I
CHAPTER 1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
INTRODUCTION .................................................................................................................... 1
EVOLUTION OF DINOSAUR LOCOMOTION ..................................................................................................... 2
SAUROPOD LOCOMOTOR HABIT AND BODY POSTURE...................................................................................... 4
SHOULDER GIRDLE RECONSTRUCTION IN SAUROPOD DINOSAURS ...................................................................... 6
SHOULDER GIRDLE EVOLUTION AND LOCOMOTION ........................................................................................ 9
FORM AND FUNCTION IN BONES .............................................................................................................. 14
FINITE-ELEMENT METHOD (FEM) ............................................................................................................ 19
AIMS AND OBJECTIVES ........................................................................................................................... 20
CHAPTER 2
MATERIAL .......................................................................................................................... 23
CHAPTER 3
METHODS .......................................................................................................................... 25
3.1
MORPHOLOGICAL DATA ......................................................................................................................... 25
3.1.1
Relative scapular position in reptiles and mammals ............................................................... 25
3.1.2
Coracosternal joint in extant reptiles ....................................................................................... 25
3.2
MUSCLES OF THE SHOULDER GIRDLE IN ALLIGATOR MISSISSIPIENSIS ................................................................ 25
3.3
BASIC MECHANICAL PRINCIPLES ............................................................................................................... 26
3.4
FINITE-ELEMENT METHOD (FEM) ............................................................................................................ 27
3.4.1
Generation of the model .......................................................................................................... 28
3.4.2
Link elements ........................................................................................................................... 28
3.4.3
Vectorial forces ........................................................................................................................ 29
3.4.4
Bearings ................................................................................................................................... 29
3.4.5
Stress distribution .................................................................................................................... 29
3.4.6
Contact elements ..................................................................................................................... 30
3.4.7
Reduction of the bauraum ....................................................................................................... 30
3.5
3-D FE MODELS OF TWO GENERALIZED TETRAPODS ..................................................................................... 31
3.6
2-D FE MODEL OF A CROCODILIAN SHOULDER GIRDLE .................................................................................. 34
3.7
3-D FESS OF A CROCODILIAN SHOULDER GIRDLE......................................................................................... 37
3.8
3-D FESS DIPLODOCUS LONGUS ............................................................................................................. 38
3.8.1
Mechanical considerations for building the bauraum ............................................................. 39
3.8.2
Generation of the model .......................................................................................................... 41
3.8.3
Estimation and distribution of body weight ............................................................................ 42
3.8.4
Muscle reconstruction ............................................................................................................. 43
3.8.5
Calculation of static equilibrium .............................................................................................. 43
CHAPTER 4
RESULTS ............................................................................................................................. 45
4.1
MORPHOLOGY OF THE SHOULDER GIRDLE .................................................................................................. 45
4.1.1
Comparison of extant reptiles and mammals .......................................................................... 45
4.1.2
Histology of the coracosternal joint in Caiman spec. ............................................................... 49
4.2
3-D FE MODELS OF SOLID TETRAPOD BODIES ............................................................................................. 51
4.2.1
Early tetrapod .......................................................................................................................... 52
4.2.1.1
4.2.1.2
4.2.1.3
4.2.2
4.2.2.1
4.2.2.2
4.2.2.3
4.2.3
4.2.3.1
Symmetrical stance compressive stresses ........................................................................................... 52
Symmetrical stance tensile stresses .................................................................................................... 54
Asymmetrical stance compressive stresses......................................................................................... 55
Cursorial mammal ................................................................................................................... 57
Symmetrical stance compressive stresses ........................................................................................... 57
Symmetrical stance tensile stresses .................................................................................................... 59
Asymmetrical stance compressive stresses......................................................................................... 60
2-D FE model of a crocodilian shoulder girdle ......................................................................... 62
Static equilibrium ................................................................................................................................ 63
i
4.3
3-D FESS OF A CROCODILIAN SHOULDER GIRDLE......................................................................................... 64
4.3.1
Static equilibrium ..................................................................................................................... 65
4.3.2
Muscle function ....................................................................................................................... 68
4.3.3
Synthesis of the scapulocoracoid ............................................................................................. 72
4.4
3-D FESS OF THE SCAPULOCORACOID OF DIPLODOCUS LONGUS .................................................................... 74
4.4.1
Forelimb position I. .................................................................................................................. 76
4.4.2
Forelimb position II. ................................................................................................................. 78
4.4.2.1
4.4.2.2
4.4.3
4.4.3.1
4.4.3.2
CHAPTER 5
Static equilibrium ................................................................................................................................ 79
Synthesis of the scapulocoracoid in Diplodocus longus ...................................................................... 84
Reconstruction of the shoulder girdle ...................................................................................... 89
Skeletal elements of the shoulder girdle ............................................................................................. 89
Muscles of the shoulder girdle ............................................................................................................ 92
DISCUSSION ..................................................................................................................... 103
5.1
FORM AND FUNCTION OF THE TETRAPOD BODY ......................................................................................... 103
5.2
FESS OF THE CROCODILIAN SHOULDER GIRDLE .......................................................................................... 110
5.3
3-D FESS OF THE SHOULDER GIRDLE OF DIPLODOCUS LONGUS .................................................................... 113
5.3.1
Reconstruction of the shoulder girdle of Diplodocus longus. ................................................. 113
5.3.1.1
5.3.1.2
5.3.1.3
5.3.2
CHAPTER 6
6.1
6.2
Skeletal elements .............................................................................................................................. 113
Muscles of the shoulder girdle .......................................................................................................... 119
Muscle forces and dimensions .......................................................................................................... 127
FESS method .......................................................................................................................... 129
GENERAL CONCLUSIONS AND FUTURE PERSPECTIVES...................................................... 133
GENERAL CONCLUSIONS ....................................................................................................................... 133
FUTURE PERSPECTIVES ......................................................................................................................... 136
CHAPTER 7
SUMMARY ....................................................................................................................... 137
CHAPTER 8
ZUSAMMENFASSUNG ...................................................................................................... 141
CHAPTER 9
BIBLIOGRAPHY ................................................................................................................. 145
CHAPTER 10
SUPPLEMENT................................................................................................................ 160
10.1 LABORATORY MANUAL FOR HISTOLOGICAL SECTIONS OF THE CORACOSTERNAL JOINT IN CAIMAN SPEC. ................ 160
10.2 MUSCLES OF THE CROCODILIAN SHOULDER GIRDLE .................................................................................... 162
10.3 3-D FESS OF A CROCODILIAN SHOULDER GIRDLE....................................................................................... 166
10.3.1
Adjustment of static equilibrium ........................................................................................... 166
10.3.2
Reduction of the bauraum ..................................................................................................... 171
10.3.2.1
10.3.2.2
First reduction step ....................................................................................................................... 171
Second reduction step .................................................................................................................. 172
10.3.3
Technical settings .................................................................................................................. 173
10.4 3-D FESS OF THE SHOULDER GIRDLE IN DIPLODOCUS LONGUS ..................................................................... 180
10.4.1
Adjustment of static equilibrium in forelimb position I. ........................................................ 180
10.4.2
Adjustment of static equilibrium in forelimb position II. ....................................................... 189
10.4.3
Reduction of the “bauraum”.................................................................................................. 199
10.4.4
Technical settings Forelimb position I. ................................................................................... 207
10.4.5
Technical settings forelimb position II. .................................................................................. 212
CURRICULUM VITAE................................................................................................................................ 223
ii
Chapter 1
Introduction
Sometime in the late Devonian (380 Myr) early tetrapodomorphs began to enlarge their
territory from water to land. With this evolutionary step new mechanical conditions challenge the
entire body and already the first tetrapodomorphs were preadapted for these demands (Coates &
Clack, 1995; Jarvik, 1996). One of the most important factors for terrestrial locomotion is the
ability of the body to sustain gravitational forces. The characteristic tetrapod bauplan evolved and
is basically still present in all living tetrapods. In general it consists of a head, neck, a dorsal
vertebral column, dorso-lateral ribs, a tail and two pairs of fore- and hindlimbs attached to the
trunk via the limb girdles. These elements are held together and moved with the aid of muscles,
tendons and ligaments. During terrestrial locomotion the most critical point is the transmission of
body weight from the trunk to the supporting limbs, which is always accomplished by the limb
girdles.
This holds true also for the biggest land living animals that ever walked on earth, the
sauropod dinosaurs. Estimations of body size in sauropods ranges from 10-100 t (Alexander, 1998;
Wedel, 2003; Seebacher, 2001) The biomechanical conditions and morphology of the shoulder
girdle are therefore of great importance to understand the biology of an animal as they contribute
to body weight support. There still is a lack of knowledge concerning the mechanics, morphology
and evolution of the shoulder girdle in both extinct and extant tetrapods, especially sauropods. In
contrast to the hindlimbs the shoulder girdle exhibits a unique feature, because it has no direct
connection to the vertebral column and is attached to the trunk only by muscles, tendons and
ligaments. Up to now research focused mainly on the hindlimbs, where osteological features and
arrangements at the pelvis are commonly used to reconstruct phylogenetic relationships. The
musculoskeletal system is assumed to fulfil the specific mechanical requirements by investing a
minimum of energy, as evolution seems to favour structures and patterns of movement
increasing the personal fitness of the animal (Alexander, 2002). A direct connection between the
posture of the limbs, their biomechanics, energy saving strategies, and costs of locomotion is
predicted (Reilly et al., 2007). Locomotor behaviour influences the biology of the animal as it plays
a major role in foraging activities, mating and in prevention of predatory contact.
1
1.1
Evolution of dinosaur locomotion
The age of dinosaurs covers almost the whole Mesozoic age from the late Triassic (230
Myr ago) until their extinction at the end of the Cretaceous (65 Myr ago). They dominated the
terrestrial fauna for about 165 Myr displaying a great variety in size and locomotor behaviour
(Carroll, 1987; Weishampel et al., 1990; Benton, 2005). Dinosaurs are part of the Archosauria
(Fig.1), which were a highly diverse group in the Mesozoic age and at present are still represented
by crown-group crocodiles and birds. In the Early Triassic archosaurs underwent a great
diversification splitting into two major groups, the Crurotarsi including the ancestors of the
crown-group crocodilians and the Ornithodira leading to dinosaurs and birds (Fig.1) (Sereno,
1991; Gower & Wilkinson, 1996; Benton, 1999).
The evolution of archosaur locomotion is subject of ongoing discussion. One of the most
important issues here is the “sprawling-to-erect-paradigm” i.e. the shift from a sprawling to an
extended limb posture and the development of parasagittal gait (Hutchinson, 2006). Sprawling is
the plesiomorphic limb position in reptiles, in which the femur is held more or less horizontally
and directed laterally. Thus, the trunk can be elevated from the ground only to a limited extent
and for a limited time. Whether the first Archosauria had an erect hindlimb position (Sereno,
1991) or were quadrupedal sprawlers (Charig, 1972; Hutchinson & Gatesy, 2000) is still debated.
Nonetheless, a fully erect hindlimb postures evolved several times during the Late Triassic, first, in
the crocodilian and the dinosaurian line including pterosaurs and leading to extant birds, and,
second, independently in the Cynodonts and their descendants the extant mammals (Parish,
1986, 1987; Benton, 2005). The sprawling or semi-erect hindlimb posture of crown-group
crocodiles therefore evolved secondarily in relation to an aquatic lifestyle (Sereno 1991; Reilly &
Elias, 1998; Hutchinson & Gatesy 2000; Hutchinson, 2006). Up to now no information is available
to evaluate evolution of the forelimbs in this context.
Dinosaurs can be separated into Ornithischians and Saurischians, mainly or initially based
on the architecture of the pelvis. In addition to morphological features of the skull, ornithischians
can be distinguished from other dinosaurs by their avian-like pelvic girdle with a postero-ventrally
pointing pubis. These herbivorous animals first appeared in the Upper Triassic and are
represented in the fossil record by several groups until the Late Cretaceous (Fig.1). Basal
ornithischians were obligate bipeds (e.g. Norman et al, 2004). With increasing body size,
ornithischians evolved a facultative bipedal or obligate quadrupedal locomotor posture (Galton
and Upchurch, 2004; Horner et al. 2004). The more derived Ceratopsians comprise two major
groups, the Cerapoda, with the horned quadrupedal ceratopsians and facultative bipedal
2
iguanodontids, and the Thyreophora with the obligate quadrupedal armored ankylosaurs and
stegosaurs (Carroll, 1987; Benton, 2005).
Figure 1 Cladogram showing the relationship of the Archosauria (modified from Remes, 2008; after Benton, 2005).
In contrast, saurischians have a reptile-like pelvic morphology with the pubic bone
pointing antero-ventrally. The saurischian clade is subdivided into the carnivorous bipedal
theropods, which comprise the ancestors of extant birds, and the quadrupedal herbivorous
sauropodomorphs (Carroll, 1987; Benton, 2005; Weishampel, 1990). The sauropodomorphs, and
here especially the sauropods, can be distinguished from other dinosaurs especially by their
bauplan. All sauropods possess rather long necks and tails, and columnar legs and were
quadrupedal herbivores (Fig.2) (Carroll, 1987; Benton, 2005; Weishampel, 1990).
The basal sauropodomorphs first appear in the Upper Triassic (Fig.2). The most abundant
sauropodomorph Plateosaurus is commonly known from excavation sites in southern Germany
and Switzerland (Yates, 2003). Like Plateosaurus, other sauropodomorphs, such as Riojasaurus
from Argentina and Melanorosaurus from South Africa, were herbivores with a body size of up to
7m long (Bonaparte, 1971; Benton, 2005). The anatomy of their forelimbs as well as the
occurrence of a furcula-like clavicle inhibiting free swinging of the forelimb necessary for
quadrupedal locomotion implicates that basal sauropodomorphs like Plateosaurus and
Massospondylus were not able to perform effective quadrupedal locomotion (Yates &
Vasconcelos, 2005; Bonnan & Senter, 2007). The just recently discovered basal sauropod
Antetonitrus from South Africa was recognized as the first “true” sauropod exhibiting
morphological adaptations for graviportal quadrupedalism and therefore serves as a link between
basal sauropodomorphs and sauropods (Fig.2) (Yates & Kitching, 2003).
3
Figure 2 Cladogram showing the evolution of the saurischian clade (modified from Remes, 2008: after Benton, 2005).
With few exceptions sauropod dinosaurs were larger than any extant tetrapod and even all contemporary dinosaurs.
Whether sauropods evolved from bipedal ancestors or habitual quadrupeds is currently
one of the most debated questions in sauropod locomotion research (Dzik, 2003; Fechner, 2005,
2006 a,b, 2009). Whereas theropods retained obligate bipedal locomotion and a carnivorous diet,
basal sauropodomorphs are thought to have been facultatively bipedal, with some taxa more
bipedal than others (e.g. Anchisaurus). The heavily-built forms such as Riojasaurus were
presumably more quadrupedal (Galton, 1990; Barrett & Upchurch, 2007). With increasing body
size and elongation of trunk and neck a transition to obligate quadrupedal locomotion can be
observed. These assumptions, however, are inconsistent with new findings, indicating that the
sauropods are the only group of saurischians with exclusively quadrupedal locomotion (Fechner
2009).
1.2
Sauropod locomotor habit and body posture
Historically, the body posture and locomotor habits of sauropods was reconstructed
based on observations in extant crocodiles, lizards, and mammals. A sprawling posture of the
limbs, similar to that of modern crocodiles was assumed by some researchers (Hay, 1908; Tornier,
1908), even though Owen (1841) concluded from features in the limb skeleton and fossil tracks
that these new species of reptiles had moved much more like mammals than living reptiles or
birds. Others agreed, but hypothesized at the same time a semi-aquatic lifestyle in which
buoyancy could support the animals’ enormous body weight (Marsh, 1878; Colbert, 1962; Romer,
1966; Swinton, 1970). Other researchers favoured a straight limb posture enabling swinging of the
4
limbs in the parasagittal plane during quadrupedal locomotion (Osborn, 1899; Hatcher, 1901;
Matthew, 1901; Abel, 1910; Gilmore, 1932), but it was not until the seventies of the last century,
that sauropods were interpreted as graviportal fully terrestrial animals (Bakker, 1971a; Coombs,
1975, 1978; Alexander, 1976).
Despite of morphological differences within the various groups, sauropods share common
morphological adaptations to quadrupedal locomotion in the limb skeleton and girdles (Coombs,
1978). In a first attempt dinosaurs were classified into locomotor types. Cursorial adaptations
observed in extant quadrupedal tetrapods, were correlated to dinosaur anatomy to evaluate their
running abilities (Coombs, 1975). Morphological evidence as the reduction of muscular
attachments in the distal limbs suggests that in sauropods forelimbs are exclusively used to
support the body weight. Furthermore, an extreme reduction of the phalangeals indicates a
columnar support in the limb skeleton (Coombs, 1978). Further, columnar legs, an increased limb
bone robusticity, and shortened limb segments in relation to an increase of body size were
determined as adaptations for slow graviportal locomotion (Carrano, 2000, 2005).
Presumably the gradual increase in body mass during sauropod evolution was responsible
for morphological changes best observed at the limbs and limb girdles. The strengthening of the
sacrum and the lengthening of the forelimbs can be viewed as adaptations to changes in
locomotor posture due to increasing body size (Coombs, 1978; Alexander, 1985; Fechner, 2009).
One of the advantages of an extended limb posture is the increase of stride length and thus
locomotor speed (Alexander, 1985). Another advantage is the ability to support higher body
weights with less energy. Since body weight increases, muscle forces can be kept low by holding
the limbs in a straight position, thus lowering the rotational moments acting on the joints
(Preuschoft & Christian, 1999). This is why animals with a certain body size do not display a
sprawling limb posture. The correlation between acting muscles forces and lever length in the
tetrapod limb further was termed the effective mechanical advantage (EMA). An increase of EMA
therefore permits a lower amount of muscles forces to keep the joints in equilibrium (Biewener,
1989, 2005; Reilly et al., 2007).
In case of tracks and trackways of sauropods, the problem arises to assign them to the
potential trackmaker. Nevertheless there exist reliable data about trackways assigned to
sauropods like Diplodocus or Brachiosaurus (Thulborn & Wade, 1984; Farlow, 1992; Lockley et al.,
1994). There are currently no footprints of running sauropod dinosaurs available. Speed
estimations based on the present data indicate only moderate walking speeds of about 3.96 km/h
in sauropods (Alexander, 1985). For comparison, the biggest extant terrestrial tetrapod, the
5
elephant, reaches walking speeds up to 7.2 km/h (Ren et al. 2009) and maximal running speeds of
about 25 km/h (Hutchinson et al., 2003). Sauropods are often compared to elephants according to
their columnar limb design. With regard to speed estimations and most recent investigations on
locomotor mechanics in elephants, it is suggested, that sauropods might have been more than
`stiff legged' walkers (Ren et. al., 2009), which would influence further speed estimations for
sauropods as well. .
One crucial issue is the relative position of the limbs during standing and locomotion.
Sauropods have been assumed to produce either wide or narrow gauge trackways. For example
Diplodocus produced narrow gauge tracks in contrast to Brachiosaurus as a wide gauge trackway
maker. The results were estimated based on the portion of total weight supported by the
respective limb pair (Henderson, 2006). The proportion of weight carried by the limbs differs
between dinosaur groups. Body mass estimations contribute to the reconstruction of the position
of the centre of mass (COM) in the body, and the different amount of body weight supported by
the hind- or forelimbs (Gunga et al., 1995, 1999, 2008; Henderson, 1999; Seebacher, 2001). In
most dinosaurs a greater proportion of weight was supported by the hindlimbs (Alexander, 1989;
Henderson, 1999, 2006). By contrast, Brachiosaurus and Titanosaurids like e.g. Saltasaurus and
Opisthocoelocaudia supposedly enlarged their amount of weight supported by the forelimbs
(Gunga et al., 1995, 1999, 2008; Henderson, 1999; Seebacher, 2001). However, the proportion of
body weight supported by the forelimbs, justifies further investigations on the forelimbs and
shoulder girdle to evaluate their relevance in body weight transmission.
According to the former paragraph it can be noted, that investigations on archosaur and
dinosaur evolution mainly refer to the hindlimbs, while the shoulder girdle and forelimbs are
mainly neglected. This is probably for historical reasons, as Romer, whose studies focused mainly
on the hindlimb girdle to evaluate their relevance in reconstructing phylogeny, did the first
studies concerning evolution of locomotion in vertebrates during the 20th century. In none of the
former described investigations on gaits and speeds of sauropod dinosaurs (Alexander, 1976,
1985, 1989; Thulborn, 1990) as well as on the posture and the evolution of the forelimbs
(Christian & Preuschoft, 1996; Christian et al. 1999a, b; Christiansen 1997; Wilson & Carrano 1999;
Christian, 2002; Bonnan, 2003, 2004; Bonnan & Yates, 2007; Carrano, 2005) the shoulder girdle
skeleton was taken into account.
1.3
Shoulder girdle reconstruction in sauropod dinosaurs
Considering the functional morphology of the shoulder girdle there still is a large gap of
knowledge especially in reconstructing the position of the scapulocoracoid, the placement of the
6
sternal plate, and the orientation of the glenoid fossa. The crucial point in shoulder girdle
reconstruction is the nature of the connection of these elements. In mammals no bony
connection from the trunk to the scapula exists. By contrast the scapulocoracoid of extant reptiles
is connected to the trunk via the coracoid to the sternal elements. Although the connection of the
elements in extant reptiles therefore should be far more comprehensible than in mammals, they
are rarely preserved articulated in fossils.
The shoulder girdle of sauropod dinosaurs has been described for a number of species.
However, few descriptions exist containing all elements. In Diplodocus (Hatcher, 1901) and
Camarasaurus (Gilmore, 1932), scapula, coracoids and sternal elements were present and
described in detail, as well as one pair of clavicles preserved in Diplodocus. Most findings lack at
least one element, so incomplete morphological descriptions on the shoulder girdle are available
for various diplodocids (McIntosh, 1988; Sereno et al. 1999; Harris & Dodson 2004; Remes 2006)
and titanosaurids (Gomani, 2005; Rose, 2007). The clavicles are often identified with reservations,
however a variety of investigators vote for their presence (Hatcher, 1901; Sereno et al. 1999;
Michelis, 2004). Except for Hatcher (1901), Gilmore (1932) and McIntosh (1988) the anatomical
position of the shoulder girdle elements in the skeleton has only been described in parts.
Reconstructions of the shoulder girdle skeleton are mainly based on phylogenetic and
morphological comparisons with extant archosaurs, like crocodiles and birds. As the scapular
position in these species varies from a position nearly parallel to the dorsal margin of the
vertebral column (birds), to a rather vertical orientation (crocodiles), present scapular
reconstructions in sauropods range between 10° and 60° towards the horizontal plane (Gilmore,
1932; Bakker, 1971a; Borsuk-Bialynicka, 1977; McIntosh 1988; McIntosh et al., 1997; Wilson und
Sereno, 1998; Parish & Stevens 2002; Wilhite, 2003; Bonnan, et al., 2005). The distal extent of the
scapula in living birds and crocodiles is roughly oriented parallel to the neural spines. It was
assumed that a similar condition in sauropods leads to a postero-dorsal inclination of scapula
(Stevens and Parrish, 2002). Further the presence of flattened areas on the external surface of the
dorsal ribs, termed facets, is used as an indicator for a constrained sub-vertical orientation of the
pectoral girdle in sauropods (Bonnan et al. 2005). Pectoral girdle position was inferred from either
death pose (Gilmore, 1932; McIntosh 1988), comparison with extant analogues (Bakker, 1971a,
Borsuk-Bialynicka, 1977; Wilson & Sereno, 1998; Wilhite, 2003) or its evolution in a phylogenetic
context (Bonnan et al. 2005), but none of the investigations take functional issues into
consideration.
7
Sternal plates are often missing in the fossil record or were not recognized correctly. In
cases in which the sternal plates are present, they were placed either far behind the coracoids,
vertically in front of the ribcage (reconstruction in the Dinosaur Museum Aahtal (CH); Wilson &
Sereno 1998; Wilhite, 2003) or were not placed on the mounted skeleton at all (McIntosh, 1988).
Schwarz et al., (2007) were the first to present a combined approach of morphological
comparison and functional considerations to reconstruct the shoulder girdle in Diplodocus,
Camarasaurus and Opisthocoelocaudia. The study reveals the following constraints for
scapulocoraocid reconstruction. In all amniotes the sternal plates are positioned parallel to the
distal ends of the thoracic ribs so as to maintain contact to the corresponding ribs and the
shoulder girdle via the coracoids. Following this, an inclination of the scapula of 60° to the
vertebral column is predicted (Schwarz et al., 2007). These implications were recently adopted by
Remes (2008) in his reconstruction of the shoulder girdle of sauropodomorphs. Nevertheless the
mechanical background for these conclusions was not determined.
In lacertilians (varanids, chameleon) the coracoid is moving translational relative to the
sternal apparatus (Peterson, 1973; Jenkins & Goslow, 1985; Lilje, 2007). The junction between
coracoid and sternum in these reptiles exhibits a bony” groove and tongue” mechanism, where
the coracoid is acting as the tongue and the sternal elements show a groove structure. The
junction is surrounded by a joint capsule embedded in connective tissue. There is currently no
information available about the condition within the sternum and the coracoids in sauropod
dinosaurs. The presence of a coracosternal joint influence the mobility in the shoulder girdle, as it
was predicted in former studies (Bakker 1971, 1987). Therefore this issue is included in the
present thesis.
It is assumed that transmission of body weight depends on the position of the limbs and
the direction in which they are articulated at the glenoid joint. Concerning their enormous body
weights some authors recommend an orthogonal force transmission from the humerus to the
scapula in sauropod dinosaurs. Assuming the humerus is positioned vertically to the glenoid joint,
the scapula will be oriented nearly parallel to the dorsal margin of the vertebral column (McIntosh
et al., 1997; Wilson & Sereno, 1998; Bonnan, 2001; Wilhite, 2003; Upchurch et al., 2004). These
assumptions again are disputed, as other investigators argued for a moderate flexion in the limbs
to reduce peak forces in the glenoid and elbow joints (Christian et al., 1999).
Few descriptions exist on the dinosaur forelimb musculature in general (Raath, 1977;
Coombs, 1978b; Santa Luca, 1980; Nicholls & Russell, 1985; Dilkes, 2000; Carpenter, 2002) and on
sauropod dinosaurs in particular. Reconstruction of the pectoral girdle musculature in Sauropoda
8
was performed for Opistocoelocaudia (Borsuk-Bialynicka, 1977; Schwarz et al. 2007), Diplodocus
(Wilhite, 2003; Schwarz et al. 2007) Camarasaurus (Wilhite, 2003; Schwarz et al. 2007) and
Apatasaurus (Wilhite, 2003). All approaches rely either on muscular analogues obtained from
extant lacertilians and crocodiles (Schwarz et al. 2007) or on crocodiles alone (Wilhite, 2003) and
differ in terminology and number of muscles taken into account. Remes (2008) applied the extant
phylogenetic bracketing (EPB) on various basal sauropodomorphs up to the basal sauropod
Patagosaurus. Soft tissues like muscles, tendons and ligaments of an extinct taxon is
reconstructed based on anatomical homologies observed in extant archosaurs bracketing
dinosaurs (crocodiles and birds) (Bryant & Russell, 1992; Witmer, 1995).
1.4
Shoulder girdle evolution and locomotion
The origin of the tetrapod shoulder girdle can best be observed in fossil stem tetrapods
from the late Devonian. Similar to fishes the pectoral girdle of the first tetrapodomorphs is of
dermal origin. The glenoid joint serves as the articulation site of the limbs to the pectoral girdle
and is formed by the dorso-lateral scapula and the ventral coracoid. In early tetrapodomorphs the
dorsal element is still covered by a pronounced dermal cleithrum. Morphological changes are
clearly visible in tetrapodomorphs like Eusthenopteron, Panderichthys, Tiktaalik, and
Acanthostega (Shubin et al., 2006). During the line of evolution a gradual reduction of the dermal
parts and an enlargement of the enchondral elements forming the scapula and the coracoid
occured. Together with the loss of the supracleithral series, connecting the skull to the limb girdle,
these changes are of great importance during the evolution of the tetrapod shoulder girdle. The
dermal elements except for the clavicles and interclavicles are lost in Ichthyostega, which is
assumed to be the first true tetrapod (Jarvik, 1996; Clack et al. 2003; Shubin, 2006). The evolution
of terrestrial activity in early tetrapods is part of an ongoing discussion. Tiktaalik roseae is
currently claimed to exhibit morphological features or at least preadaptation for terrestrial
locomotion in the limb girdle and cranial skeleton (Fig.3) (Shubin et al., 2006, Daeschler et al.,
2006, Ahlberg & Clack, 2006; Markey et al. 2008).
From the first carboniferous tetrapods to the early archosauromorphs the limbs and limb
girdles are basically identical. In early reptiliomorphs like Diadectes the shoulder girdle is built of a
massive bony scapulocoracoid, still retaining a cleithral part at the cranial edge, a clavicle and a
ventrally positioned interclavicula (Fig. 3). The pectoral girdle shows a large scapulocoracoid
bearing a screw shaped glenoid for the head of the humerus. From Diadectes to basal archosaurs
as Prestosuchus the cleithrum is first reduced then completely absent, while the shoulder girdle
typically retains a frontally positioned clavicle and a t-shaped interclavicula (Huene, 1935-42;
9
Benton, 2005) (Fig.3). Diadectes retained the ancestral sprawling limb posture, whereas for early
archosaurs as Prestosuchus a semi-erect position of the limbs can be assumed (Benton, 2005). Up
to the first archosaurs and amongst them the basal dinosaurs, the sternal elements are
cartilaginous, and ossification of the breastbone is first reported in pterosaurs and dinosauria
(Starck, 1979). In basal dinosauromorphs like Marasuchus some gradual changes in the pectoral
girdle become visible. The whole appearance of the skeleton becomes more slender. These first
dinosaurs were presumably obligate or at least facultative bipeds, adapted to fast running
locomotor habits (Benton, 2005). The main elements of the shoulder girdle, scapula and coracoid,
are reduced in size, the scapula becomes more slender (Fig.3). In both basal dinosauromorphs and
dinosaurs like e.g. Herrerasaurus there is no evidence for a clavicle or sternal elements (Sereno,
1993), whether due to preservational reasons or functional modifications remain unclear.
The pectoral girdle in “prosauropods” consists of a more or less slender scapula with
moderate expansion of the distal end and a medium sized coracoid (Huene, 1926; Young, 1941;
Bonaparte, 1971; Cooper, 1981; Zhang & Yang, 1995) (Fig.3). A furcula-like clavicle has been
described for Massospondylus (Yates & Vasconcelos, 2005) and Plateosaurus (Huene, 1926, 1932;
Galton, 2001) although a probable facet for the clavicle at the acromial process is visible in all
forms. In both cases it is likely, that this v-shaped element embraced the pectoral girdle, while
being attached to the frontal edge of the scapula (Yates & Vasconcelos, 2005). Ossified sternal
plates are known from Plateosaurus (Huene, 1926), Massospondylus (Cooper, 1981),
Lufengosaurus (Young, 1941), Jingshanosaurus (Zhang & Yang, 1995), and Yunnanosaurus (Young,
1942). Most of these sternal plates have a suboval shape with a thickened, rugose cranio-lateral
border, which may indicate a connection to the adjacent coracoid plates. As it was stated before a
facultative bipedal posture of the forelimbs has been assumed for basal sauropodomorph
dinosaurs.
In diplodoicid sauropods the scapula exhibits an enormous craniad enlargement of the
proximal part, termed the acromial region, and an elongated scapula with a variably expanded
distal end. The prominent acromion ridge divides a larger anterior from a smaller posterior fossa.
Scapular and coracoid elements are well ossified forming together the glenoid facet, in which the
scapula accounts for two thirds of the glenoid surface area (Wilhite, 2003) (Fig.3,4). The coracoids
in sauropods are oval shaped, with an anterior border showing a roughened edge, indicating that
it was covered by cartilage during life and may have been attached to the corresponding sternal
elements. In general all these features are present in all members of this subgroup e.g. Diplodocus
(Hatcher, 1901), Apatosaurus (Upchurch et al., 2004), Supersaurus (Lovelace et al., 2007),
Tornieria (Remes, 2006), Suuwassea (Harris & Dodson, 2004) and other more recently described
10
diplodocoids. The pectoral girdle morphology of macronarian sauropods differs in detail. In
Brachiosaurus for instance the distal end of the scapula is enlarged, while the middle part is more
slender than in Diplodocus. Together with the broad proximal part the scapula assumes a barbelllike appearance (Fig.4). The most obvious difference between the two sauropod groups lies in the
shape of the sternal elements. In diplodocoids the sternal plates are oval and dorso-ventrally
convex, whereas in macronarians the sternal plates are more slender and elongated as in
Camarasaurus or semilunar in all Titanosaurids (Fig.4).
Comparison of the shoulder girdle in non-sauropodan sauropodomorphs and sauropods
reveals some differences: i) the elongation of the scapula relative to the coracoid, ii) an expanded
acromial region, iii) a more ventrally positioned glenoid and iv) a relative enlargement of the
sternal plates (Fig.4). In contrast to the v-shaped elements described in some non-sauropodan
sauropodomorphs (Yates & Vasconcelos, 2005) the clavicles in all sauropods are bilateral arched
elements as described for e.g. Diplodocus (Hatcher, 1901; Holland, 1906). Further it appears that
the expansion of the distal end of the scapula and therefore the cartilaginous suprascapula is
reduced during sauropod evolution (Remes, 2008). As stated above a quadruped locomotor habit
is assumed for all sauropod dinosaurs.
In extant tetrapods the dermal elements interclavicula and clavicula persist in the
shoulder girdle to varying degrees (Fig.3). The interclavicula is still present in extant archosaurs
(birds, crocodilians) and reptiles, but retained in the mammalian line only in monotremes (Fig.3).
The clavicle on the other hand can be observed in anurans, lacertilians, Sphenodon and in a
number of mammals, while it is absent in extant crocodilians. The enchondral parts of the
shoulder girdle in anurans and reptiles are represented by the scapula and the procoracoid plate
(Fig.3) (Romer & Parsons 1977; Starck, 1979). During the evolution of the mammalian shoulder
girdle the chondral elements receive gradual reduction as well. In monotremes the basic
condition is still visible. It consists of a scapula, a procoracoid, and a metacoracoid, while the
derived mammals retain only the scapula and a metacoracoid process (Fig.3) (Romer & Parsons,
1977; Starck, 1979). The sternum or sternal plates are rarely included into considerations about
the shoulder girdle, even though they represent the connection between the pectoral girdle and
the ribcage in tetrapods. A bony sternum is first present in basal tetrapods (Fig.3). Its evolutionary
origin or possible homology remains unclear, although derivation from the ventral parts of the
ribs or the pectoral girdle has been hypothesized. Embryology in reptiles and mammals indicates
that the sternum develops independently and only in a second step becomes associated to the
ribs and pectoral girdle (Starck, 1979; Feduccia & McCrady, 1991). Among tetrapods the degree of
ossification, shape and dimensions of the sternal apparatus is extremely variable. In urodeles and
11
anurans the sternum consists of one or two cartilaginous plates, which are only connected to the
pectoral girdle (Romer & Parsons, 1977; Starck, 1979). The generalized sternum in extant reptiles
is represented by a ventral cartilaginous plate, which can be ossified to variable degrees. It
exhibits a connection to the coracoids via a coracosternal joint, while the ventral parts of the ribs
insert more or less laterally (Romer & Parsons, 1977; Starck, 1979). The mammalian sternum is
reduced to a rod-like element, which consists of a ventral series of separate bones fused together
to a variable degree (Fig.3) (Romer, 1977; Starck, 1979). In contrast to reptiles and birds the
mammalian shoulder girdle shows no bony connections to the trunk except formammals, which
retain a clavicle between the scapula and the sternum (Fig.3).
12
Figure 3 Evolution of the tetrapod shoulder girdle showing the main girdle elements in lateral view. Elements of dermal
origin (yellow); Chondral elements: scapula (light grey), coracoid (dark grey). Forelimb position is displayed in frontal
(left) and lateral (right) view.
13
Figure 4 Shoulder girdle elements of three major sauropod groups. Camarasauridae (Camarasaurus lentus), redrawn
after Gilmore (1932); Diplodocoidea (Diplodocus longus); Brachiosauridae (Brachiosaurus brancai). Scapulocoracoids are
shown in lateral view. Sternal plates are viewed ventrally, whereby the cranial border is pointing upwards.
1.5
Form and function in bones
It seems obvious that form and function are closely related, as appropriate structures
enable an animal to fulfill specific tasks. The basic principles of this interrelationship however are
not that apparent and still subject of many discussions (Hildebrand & Goslow, 2001). According to
the complexity of the vertebrate body one has to be aware to determine correlations based on
simple comparison. Interrelationships between form and function of a structure should always
take into account variable influences acting on the system, confirmed by empirical and
experimental data (Hildebrand & Goslow, 2001).
14
The musculoskeletal system as one part of the animal body complex provides the ability
to demonstrate form and function relationships in a well-arranged system, as it is "perhaps
nowhere as evident as here" (Carter & Beauprè, 2001). Structural organisation of the bones,
cartilage, tendons and ligaments is a result of a unique and complex phylogenetic and ontogenetic
history in which genes and mechanical forces provide critical control (Carter & Beauprè, 2001).
Focusing on bone, functional adaptation is not only visible in the status quo, but along the
evolutionary line and during ontogenesis. The evolution of marine mammals is representative for
a shift from a terrestrial to an aquatic mode of locomotion, accompanied by modification and
reduction or loss of whole skeletal elements (Hildebrand & Goslow, 2001). In the early
development of marsupials temporary functional elements like the metacoracoid, ontogenetically
retained from former ancestors, is secondarily reduced according to locomotor changes in very
early stages of postnatal development (Klima, 1987).
In a macroscopic view functional adaptation of bone can be detected as alterations of
bone dimensions, the development of rugosities at the insertion sites of muscles, tendons and
ligaments as well as the internal distribution of osseous structures, namely compact and
cancellous bone. These changes are determined as the remodelling of bone in mature stages,
compared to the modelling of bone during ontogenesis. A wide range of studies reveal the strong
influence of mechanical stimuli on bone formation. This contributes to reconstruction of life
history in hominid remains (Preuschoft, 1970; Runestadt et al., 1993; Trinkaus et al., 1994; Larson,
1997; Orr, 2005), in orthopaedics and biomechanical engineering (Currey, 1984; Carter, 1987;
Kaspar et al. 2000; Witzel 1985, 1996, 2000), as well as in zoology (Demes et al., 2001;
Liebermann et al 2003; Carlson & Patel, 2006; Patel & Carlson 2007; Polk et al. 2008). All studies
are based on apparent correlations between locomotor type or habitual use in the individual and
detecTable alterations such as bone density and formation. During ontogenesis ossification in the
developing embryo appears at about the same time as first muscle contractions are detecTable.
Inhibiting muscular activity during embryonic development by either muscular disease or genetic
modification results in incomplete bone formation indicated by a lower amount of bone
substance, decreased diaphyseal diameters and an overall miss-shaping of the single bony
elements (Lightfoot & German, 1998; Gomez et al., 2007; Jones et al., 2007).
The relevance of mechanical influence on the skeletogenesis and remodelling as an
epigenetic factor was recently pointed out by Newman & Müller (2005). Most of the presented
studies refer to mammalian bone, but there are studies indicating a strong influence of
mechanical stimuli e.g. in teleostians as well (Kranenbarg, 2005).
15
The current approach is based on two important mechanisms connected to bone
morphology; bone formation during development (bone modelling), and its ability to functional
adaptation according to everyday usage (bone remodelling). The relationship between function
and form in bones was first noted by Wolff (1892) in his “Law of bone remodelling” and Roux
(1893) predicting a causal relationship between mechanical forces and morphological
modifications during life. Moreover, Wolff (1892) argued that the design of an organism is a result
of both functional adaptation and natural selection according to Darwin’s evolutionary theory
(Carter & Beauprè, 2001). The functional adaptation of bone was described as an auto-regulatory
system responding to external stimuli, but did not explain the issue sufficiently (Roux, 1893).
Later on Pauwels established the term “causal morphogenesis” to describe the
functionally dependent auto-regulatory differentiation of the supporting tissue (Pauwels, 1965).
Mesenchymatic stem cell tissue here is only able to detect two kinds of mechanical stimuli,
elongation and compression. Elongation leads to connective tissue fibers directed to the tensional
traction line, as it appears in tendons, ligaments, disci and menisci (Kummer, 1985) (Fig. 5). Under
the influence of compression the cell would adapt its metabolism to produce proteoglycanes,
which are basic components of the cartilage matrix. According to Pauwels (1965) bone formation
is initiated by elastic deformation of the primary supporting tissue. Compression leads to
chondrale ossification via a cartilaginous intermediate stage, whereas tension is responsible for
desmal ossification out of connective tissue (Fig. 5) (Pauwels 1965; Kummer, 1980).
Figure 5 Scheme of Pauwels “causal morphogenesis”, illustrating the influence of compression and tension on the
mesenchymatic tissue. Both pathways towards lamellar bone, chondral as well as desmal ossification, are shown
(Pauwels, 1965) (redrawn after Kummer, 1980).
16
Once the skeleton is fully grown each bone has to maintain its function during everyday
use. Bone tissue has the ability to respond to mechanical stimuli to keep the resulting stresses
constant, thus adapting its strength to the actual mechanical condition (Fig. 6) (Kummer, 1985;
Roux, 1893). In an evolutionary context the ability to respond to physical stresses is determined
by the genome of the individual. This leads to the conclusion that organisms possessing a highly
adaptive potential could differentiate in continuous interaction with external stimuli (Pauwels,
1965).
Figure 6 Functional adaptation of bone. Bone density depends upon the mechanical stress value. Stress value below the
physiological level will lead to atrophic alterations of the bony. In contrast, a stress value above the physiological level
causes an abnormal increase of bone (redrawn after Witzel, pers. communication).
Nowadays it is widely accepted that specific mechanical stimuli have a strong influence on
the differentiation of bone, cartilage and connective tissue, but whether bone modelling and
remodelling is induced by compressive and tensile stresses (Biewener, 1991; Blob & Biewener,
1999, 2001; Currey 2002; Liebermann et al., 2003; Main & Biewener, 2004) or is dominated by
compressive stresses alone (Sverdlova & Witzel 2010; Witzel et al., 2010) is still disputed.
It has long been predicted that bone is mainly adapted to withstand high bending and
torsional moments, whereas compression does only plays a minor role (Biewener et al., 1991;
Carter & Beauprè, 2001; Currey 2002; Liebermann et al., 2003). The majority of these studies
investigate locomotor stresses in long bones, where the resulting stresses are supposed to be
mainly a result of bending. Strains hereby were determined by in vivo implanted strain gauges
applied to the periost of the bone (Biewener, 1991; Blob & Biewener, 1999, 2001; Main &
Biewener, 2004). Although bending induced strains are supposed to play the major role in bone
loading, it was noted, that there are "noTable exceptions to the predominance of bending
17
induced bone stress at the femora of certain rodents and the metapodials of the horse, which
experience primarily axial compression during steady speed locomotion" (Biewener, 1991). The
absence of correlation between regional histomorphometric patterns and the measured strain
environments during limb bone loading was observed in bone of goats during ontogeny. Here it
was mentioned that the potential effects of other physiological and mechanical factors, such as
skeletal metabolism and adjacent muscle insertions, could influence the gross and microstructural
morphology of the radius during ontogeny (Main, 2007).
However, there is strong evidence that in the acting musculoskeletal apparatus bending
moments are reduced to a certain degree and that bone is mainly loaded by compression.
According to Pauwels (1965) the femur is exposed to bending moments superimposed by
compressive stresses that are always higher than the occurring tensile stresses. Such a mechanism
was termed as “flexural neutralization” by Frost (1964). Kummer (1962) stated that bones are
functionally adapted to minimize the occurring bending moments, thus allowing for a reduction of
the required bony material. If less material can be used to gain maximal stability, this is a great
advantage for the living organism, as it reduces the amount of the invested energy (Rossmann et
al. 2001; Witzel & Preuschoft 2005). Bending moments can be reduced either by active structures
like muscles or passive elements like tendons and ligaments (Möser & Hein, 1992; Witzel 2007;
Hert, 1994; Rudman et al., 2006; Witzel & Preuschoft, 2005; Sverdlova & Witzel, 2010). The
optimization of bone as a light-weight structure was recently investigated by FE analysis of the
human femur. The results of this study indicate accounts for that `bone formation takes place on
the pathways of the compressive stresses` (Sverdlova & Witzel, 2010).
Further evidence supporting the large influence of compressive stresses is provided by
experimental studies focusing on the cellular development and genetically determined
mechanisms in modelling and remodelling. Recent analysis of human osteocytes reveals the
nature of cell proliferation after application of cyclic tensile strain using physiological amplitudes.
Cell activity associated with primary matrix formation is increased, while further differentiation of
osteoblasts and matrix maturation decrease. Therefore tensile stresses play a role in initiating
proliferation in osteocytes in early stages of development, but do not account for bone formation
itself, as it was postulated by Pauwels (1965; see Kaspar et al., 2000). There is a close relationship
between cyclic compressive strains applied to cells and the release of bone formation related
substances (Rath et al., 2008). Under static compression the chondrocytes produce growth factors
leading to an increase in chondrocyte differentiation as well as their maturation, leading to
ossification of the tissue (Takahashi et al., 1998). In contrast, tensile stresses inhibit the release of
such metabolic products and therefore inhibit proliferation and arrangement of chondrogenic
18
cells (Sato et al., 1999; Ikegame et al., 2001). These studies strongly indicate mechanisms in which
cartilage as a primary supporting tissue develops under the influence of compression loading,
according to Pauwels causal morphogenesis.
1.6
Finite-element method (FEM)
The relationship between bone formation and mechanical stresses can be investigated
using the Finite Element Method (FEM). FEM is an approach widely used in engineering design for
mechanical strength analysis of machine parts and structures. FEM is a technique used for the
calculation of the strain, stress, and deformation in solid and liquid structures (Zienkiewicz, 2005).
The mechanical response of a structure to external loading is described by a set of differential
equations. In order to solve them, the structure is divided into a finite number of geometrical
elements. Each element is connected to the neighbouring elements via nodes. The elements and
nodes together build a finite-element mesh of the structure. This process of subdividing a
structure is called discretization, pointing out that the whole calculation of stresses and
deformation occurs at a discrete number of locations instead of at every point within the original
structure. The solution of the differential equations within each of these elements then leads to a
highly accurate solution. The software implementation of the FEM additionally is able to visualize
the solution as stress distribution and deformation (Carter & Beauprè 2001; Zienkiewicz, 2005). It
has been commonly used for medical application in orthopaedics for many years (Witzel, 1985,
1996, 2000; Carter & Beauprè, 2001; Effenberger et al., 2001; Sverdlova & Witzel 2010) but is
recently applied in palaeontology and zoology as well (Daniel & McHenry, 2001; Rossmann et al.
2001; Rayfield, 2004, 2005; Dumont et al., 2005; Ross et al., 2005; Huber et al. 2005; Witzel &
Preuschoft, 2002, 2004, 2005; McHenry et al., 2006; 2007; Witzel & Gössling, 2006; Witzel, 2006,
2007; Wroe et al. 2008; Gössling et al. 2008; Moser et al. 2008; Fujiwara, 2009). A survey of the
application of FE-methods in biology and palaeontology was recently given by Rayfield (2007). The
current research using FEM can be divided into two different approaches using the method, in an
inductive and a deductive approach.
In the inductive approach (FEA) the structure is analyzed according to adaptation and
optimization of bone to a particular function (Rayfield, 2007). In FEA in-vivo stresses and strains
are calculated within a virtual structure under an applied load. The FE models are generated
based on CT scans of the object. FEA is mainly used for investigations of the skull, i.e whether
bone is adapted to resist bite forces and other functional forces (Daniel & McHenry, 2001;
Rayfield, 2003, 2004, 2005; Dumont et al., 2005; Ross et al., 2005; McHenry et al., 2006). Other
studies used the FE- method for estimation of the maximal occurring bite forces in the white
19
shark (Wroe et al., 2008) or for investigation of the bending strength of the ribcage in
quadrupedal animals (Fujiwara, 2009).
Studies using FE by the inductive approach reveal the stresses occurring in a priori existing
structures after application of mechanical forces but do not explain the development of bone
shape as a response to mechanical loading. To compensate this lack of information, a more
deductive approach was developed. The finite-element structure synthesis (FESS) is an interactive
procedure of the virtual synthesis of skeletal structures out of a homogeneous solid block after
applying muscle forces and calculation of the stress distribution in the structure. The bone hereby
is viewed as a compressive structure, in which bending loading is compensated to eliminate
tensional stresses, hence, the bony elements are synthesized according to the observed
compressive stress pattern. This approach is essentially based on Wolff`s law and clearly
demonstrates the direct correlation between the functional loading and the biological structure
and shape. It was successfully applied to the skulls of extant (Witzel & Preuschoft 2002, 2004;
Witzel et al. 2004) and extinct species. These were i.e. basal archosaurs as Proterosuchus
(Rossmann et al. 2001), hominids (Homo neanderthalensis) (Witzel, 2006); and the skulls of
sauropod dinosaurs (Witzel 2007; Witzel & Preuschoft, 2005). Ongoing studies apply the FESS to
the skull of Plateosaurus (Goessling et al. 2008) and Herrerasaurus and in an investigation of
sauropod vertebrae (Moser et al. 2008).
1.7
Aims and objectives
As it has been reported, reconstructions of sauropod shoulder girdles are rare and
mainly based on morphological homology with extant relatives or on functional analogies with
extant tetrapods. To provide ground for morphological and functional comparison a series of data
is collected. Mounted skeletons of the shoulder girdle of extant reptiles and mammals are
described according to their overall anatomy and relative position to the trunk. The morphology
of the coracosternal joint is described using mounted skeletons of extant lacertilians and
crocodiles as well as based on histological cross sections of one crocodile specimen.
The development of the shoulder girdle elements during both tetrapod and
dinosaurian evolution is assumed to be closely connected to the respective locomotor type. The
mechanical necessities of the single elements in the tetrapod body had not been investigated in
detail. With the aid of the finite-element method (FEM) the stress flow in two generalized threedimensional solid tetrapod bodies, representing either a sprawling or extended limb posture, will
be analyzed to determine the influence of gravity on the body. The occurring mechanical stresses
then will be correlated to the presence of structural elements of the tetrapod body.
20
The current approach is based on the prediction of Wolff´s law and Pauwels causal
morphogenesis that bone formation is mainly determined by compressive stresses resulting from
mechanical function of the respective element in the musculoskeletal system of the vertebrate
body. Following this assumption the mechanical function in each vertebrate musculoskeletal
system can be determined from the shape of its elements and vice versa. The finite element
structure synthesis (FESS) is a deductive approach to synthesize the shape of the original bone out
of a primary generated bauraum, and in this study was initially applied to the post-cranial
skeleton of vertebrates. The novel method FESS is first adapted to the shoulder girdle and the free
forelimb of an extant crocodile (Caiman crocodilus) in order to evaluate the method to assess the
mechanical principles of the shoulder girdle.
Sauropods are supposed to exhibit graviportal functional adaptations in the shoulder
girdle skeleton similar to those observed in large extant mammals e.g. in elephants. In contrast
they exhibit reptile-like osteological features and muscle equipment similar to their extant
relatives. Thus, reconstruction of the shoulder girdle function, only based on comparison with
extant tetrapods, is assumed to be insufficient. Sauropods are assumed to exhibit an intermediate
type of force transmission in the shoulder girdle, which has to be determined in the present
investigation. Then it will be compared to the conditions in extant archosaurs (crocodilians) and
mammals.
In this case only little information on either function or behaviour of the animal exists. The
only evidence is provided by the fossil record and the previous results of the FEM in tetrapod
bodies and FESS in the crocodilian shoulder girdle. This will lead to basic conclusions about the
arrangement of the shoulder girdle elements and limb posture in Diplodocus longus, according to
the predicted mechanical requirements for a successful force transmission from the trunk to the
forelimbs. With the aid of FESS the scapulocoracoid of Diplodocus longus is synthesized by, first,
reaching static equilibrium between the elements of the shoulder girdle, trunk and forelimbs. The
focus will be on the static condition in one limb stance to offer a precisely defined frame for the
investigation, with a minimum of boundary conditions. The resulting compressive stress patterns
then will serve as the blueprint for the final synthesis of the scapulocoracoid element. If the
formerly predicted boundary conditions, like positioning of the elements and applied muscle
forces required to maintain equilibrium in the shoulder-trunk-forelimb system, are valid the
synthesis will reflect the original shape of the scapulocoracoid of Diplodocus longus, thus
confirming the previously predicted reconstruction of the acting musculature as well as the
position of the pectoral girdle under a biomechanical point of view.
21
The functional conditions in the shoulder girdle of Diplodocus longus, which are revealed
in the present thesis, are discussed according to their relevance for reconstructions in scientific
expositions and sauropod locomotion in general. Further it will be evaluated if the shoulder girdle
of Diplodocus longus shows any body weight related adaptations that allow for an increase of
body size, and whether the study contributes to current discussions in context of sauropod
locomotor evolution.
This approach will provide insight into the biomechanical conditions of the tetrapod
shoulder girdle in general and the basic principles of the construction of the shoulder girdle in
sauropod dinosaurs. A mechanically plausible reconstruction of the shoulder girdle in its natural
arrangement is presented to reveal mechanically based functional adaptations for locomotor
habits in these extinct giants. The correlation between form and function in the musculoskeletal
system should be supported, by the determination of mechanical stresses as one major influence
for the formation of bone. And finally FESS should be confirmed as a powerful method to test
hypotheses regarding the relationship between structure and function during the evolution of the
vertebrate skeleton.
22
Chapter 2
Material
Table 1 The following Table shows the species examined, as well as their part investigated and the origin of the
specimen.
Species
Investigated part
Institution
Abbreviation
Alligator missisipiensis
(skeleton)
Overall anatomy of the
shoulder girdle skeleton;
Relative scapular position,
Anatomy of the
coracosternal joint.
Museum of Natural History
Dusseldorf/ Aquazoo (DE)
MNHD
Alligator mississipiensis
(dissected specimen)
Determination of the
shoulder girdle musculature;
Estimation of muscle forces;
Overall anatomy of the
pectoral girdle.
Crocodile farm „Renz“,
Friedberg
CFR
Basiliscus plumifrons
Relative scapular position;
Anatomy of the
coracosternal joint.
Museum of Natural History
Dusseldorf/ Aquazoo (DE)
MNHD
Caiman crocodilus
(skeleton)
Overall anatomy of shoulder
girdle skeleton used for
modelling the 3-D FESS
model.
Preparatory School Bochum
(DE)
IAS
Caiman crocodilus
Overall anatomy of the
shoulder girdle skeleton;
Relative scapular position,
Anatomy of the
coracosternal joint.
Institute of Zoology and
Neurobiology RuhrUniversity Bochum (DE)
MNHB
Chamaeleo jacksonii
Relative scapular position;
Anatomy of the
coracosternal joint.
Museum of Natural History
Dusseldorf/ Aquazoo (DE)
IZNB
Chamaeleo lateralis
Relative scapular position;
Anatomy of the
coracosternal joint.
Museum of Natural History
Dusseldorf/ Aquazoo (DE)
IZUB
Chamaeleo melleri
Relative scapular position;
Anatomy of the
coracosternal joint.
Museum of Natural History
Dusseldorf/ Aquazoo (DE)
PSB
Cordylus giganteus
Relative scapular position;
Anatomy of the
coracosternal joint.
Museum of Natural History
Dusseldorf/ Aquazoo (DE)
MNHD
Crocodylus niloticum
Relative scapular position;
Anatomy of the
Museum of Natural History
MNHD
Extant reptiles
23
coracosternal joint.
Dusseldorf/ Aquazoo (DE)
Cyclura nubilis
Relative scapular position;
Anatomy of the
coracosternal joint.
Museum of Natural History
Dusseldorf/ Aquazoo (DE)
MNHD
Diplodocus longus
Overall shoulder girdle
anatomy for modelling the
3-D FESS model
Senckenbergmuseum
Frankfurt (DE)
SMF
Diplodocus spec.
Shoulder girdle anatomy;
Morphology of the sternal
plates.
Sauriermuseum Aahtal (CH)
SMA
Tiliqua rugosa
Relative scapular position;
Anatomy of the
coracosternal joint.
Museum of Natural History
Dusseldorf/ Aquazoo (DE)
MNHD
Iguana iguana
Relative scapular position;
Anatomy of the
coracosternal joint.
Museum of Natural History
Dusseldorf/ Aquazoo (DE)
MNHD
MNHM
Early tetrapod
Tiktaalik roseae(NUFV 108)
Pectoral girdle morphology
and body measurements.
University of Chicago
UC
MNHD
Mammals
Bradypus spec.
Relative scapular position.
Institute of Zoology
University of Bonn
IZB
Camelus bactrianus
Relative scapular position.
Institute of Anatomy,
Université Louis Pasteur,
Strasbourg (F)
IAS
Equus caballus
Relative scapular position.
Institute of Anatomy at the
Université Louis Pasteur,
Strasbourg (F)
IAS
Leo panthera
Relative scapular position.
Institute of Zoology
University of Bonn
IZB
Rupicapra rupicapra
Relative scapular position.
Institute of Anatomy at the
Université Louis Pasteur,
Strasbourg (F)
IAS
Vicugna vicugna
Relative scapular position.
Institute of Anatomy at the
Université Louis Pasteur,
Strasbourg (F)
IAS
24
Chapter 3
3.1
Methods
Morphological data
All pictures of the skeletal material and muscular dissections were photographed using a
digital camera and processed using Adobe CS4. In this study it should be noted, that in the
following, despite its different homology, the term coracoid is used for all coracoid-, procoracoid
and metacoracoid elements, according to palaeontological conventions. Homologies concerning
this term will be pointed out in the text if required.
3.1.1 Relative scapular position in reptiles and mammals
For determination of the scapular position relative to the trunk in extant tetrapods,
pictures were taken from mounted skeletons of various reptiles and mammals. (Table 1)
3.1.2 Coracosternal joint in extant reptiles
Morphology of the coracosternal joint in extant reptiles (Table 1) was documented by
digital pictures. The coracosternal joint in one Caiman spec. was dissected, described and cross
sections were prepared. Selected cross sections were photographed and the distribution of
cartilage and bony tissue in this region was described.
3.2
Muscles of the shoulder girdle in Alligator mississipiensis
Pectoral girdle morphology was documented by dissection of one female Alligator
mississipiensis (CFR). Gross anatomy of the musculature, as position, origin and insertion sites of
each muscle as well as their principle function were documented, digitally photographed and
compared with former descriptions (Brinkmann 2000, Meers 2002). Descriptions focus mainly on
the muscular lines of action and function during transmission of body weight from the trunk to
the shoulder girdle and maintenance of equilibrium. Measurements include total area and crosssectional area of each muscle. Total area was estimated by drawing the muscle’s outline on
millimetre paper. Cross-sections were taken through the centre of mass, perpendicular to the
main fibre direction. Cross-sectional area then again was measured by drawing its outline on
millimetre paper. Maximum muscle force Fmax of each shoulder girdle muscle in the alligator
specimen was estimated by multiplying their anatomical cross-sectional area (A) by the specific
25
tension (Tsp) of reptiles. A specific tension of about 0.2 N/mm² in this case refers to the desert
iguana (Medler 2002).
1) Fmax = A * Tsp
3.3
Basic mechanical principles
Biomechanics can be viewed as the link between the engineering and biology,
investigating biological problems with the aid of the mechanical principles. Before discussing the
applied methods in detail, a short introduction into the basic mechanical principles is presented
providing the basis for the following investigation.
Force F is a measure of the intensity of the interaction of bodies. Force causes the change
of velocity of a body or the deformation of a body. Force is measured in Newton (N) and has units
of mass m multiplicated by acceleration a: 1 N = 1 kg * m/s²
2) F = m * a,
Moment of force (M) is the product of force (F) and lever arm (l) given as
kg *m/s² *m [Nm].
3) M = F * l
The system is in equilibrium when the sum of all forces and moments is zero.
4) ∑ Fi x,y,z = 0, with F acting in all directions of the
coordinate system.
5) ∑ Mi
x,y,z
= 0, with M acting about all
coordinates.
In an elastic body external forces produce internal stresses and strains. Stress (σ) is the
measure of local intensity of a force (F) per unit area (A) and is measured in N/mm² [MPa]. As
tensile stress (σ+) leads to an increase of the initial length, it is positive by sign. Compressive
stress (σ-) results in a negative elongation, therefore, it has a negative sign.
6)
26
σ=F/A
Strain (ε) can be regarded as local deformations depending on the material properties of
the object and describes the relative elongation (Δ l) of the body.
7)
ε = (l0 – l1)/l0
Young`s modulus (E) defines the ratio of the stress
σ to the strain  attained by the
deformation. Young`s modulus therefore is an indicator of the stiffness of the material and is
given as N/mm² [MPa].
8) E =
σ/
Figure 7 Scheme showing two dinosaurs of different body weight standing on a scale in equilibrium, demonstrating the
relationship between (weight) forces and moments in a static system. F1 and F2 are the acting forces, whereas Fr1 and
Fr2 indicate reaction force of the same value, acting into opposite directions. L is the lever arm (distance from the
insertion point of the force to the pivot of the beam). M is the resulting moment that rotates the beam.
3.4
Finite-element method (FEM)
The finite element method (FEM) in this thesis is used in two different approaches. In a
first step the mechanical stresses, which result from defined loading conditions after calculation,
are analysed and applied to investigate mechanical stresses in solid tetrapod bodies and a 2-D FE
model of the crocodilian shoulder girdle. The second approach, the finite element structure
synthesis (FESS), is applied to investigate the mechanics of the crocodilian and sauropod shoulder
girdle system. Both approaches share a number of methodological procedures and element
descriptions, except for the synthesis in FESS. Therefore, the reduction of the bauraum is the most
distinguishing part between these two approaches, and only refers to the FESS.
27
3.4.1 Generation of the model
All finite element (FE) modelling and calculations in this work are carried out by the FE
program ANSYS Inc. This FE program allows the attribution of different material properties to
each volume of the FE model e.g. Young´s modulus, Poisson`s ratio and material density.
Modelling procedure starts with creating the geometrical model with corresponding material
properties, which is then subdivided into the finite elements building the FE mesh. After that the
so-called boundary conditions are defined for the meshed FE model. The boundary conditions
comprise the loading through the external forces and of the constraints of the free body motion
in space. The external forces, which are applied in this approach indicate gravitation, body weight
or are related to acting muscles. In this investigation two different approaches are used to
indicate acting forces in a FE model, which are link elements and single, directed vectorial forces
(Fig. 8).
3.4.2 Link elements
Muscle action can be simulated by link elements, which represent the required muscles in
the original anatomy of the subject. The origin and insertion sites of the considered muscles are
therefore obtained from the literature. These link elements work like active springs holding each
element in its relative position thus preventing large relative motion of the elements during
calculation (Fig. 8). Pre-tension of each link element is defined by the link crossectional area A
(mm²) and the initial strain , which is the relative elongation of the link element (Δ l/l), which
depends on the Young´s modulus E (N/mm²) of the material chosen for the link elements. The
forces produced by the link elements in the pre-tension in the initial state as well as in the
deformed state can be calculated as follows:
9) F = σ * A, whereby σ =
* E
After the calculation, resulting stresses within the working links can be detected. As the
applied forces are closely related to link stresses and as stress values in the links change during
calculation due to the deformation of the bodies, the resulting forces differ from the initial
desired values as well. To ensure the agreement of link forces with the desired external forces
they have to be calculated precisely following a number of equations. As the presented approach
is not suiTable for a determination of absolute muscle forces, the occurring stress values after
calculation are only described qualitatively. Because of their stabilizing properties during FE
calculations, which are thought to be required to reach equilibrium in the system, link elements
28
are applied in a primary investigation step. They are used to evaluate, whether the FESS could
actually be applied to a multi-body system.
3.4.3 Vectorial forces
Instead of link elements single forces can be used in FE calculations. Hereby defined
forces are applied to the nodes in the meshed FE model. Forces are always indicated as arrows
(Fig. 8). To modify the direction of the force, the node can be rotated in all directions of the
coordinate system. In order to simulate muscle action, one pair of forces is required. Both forces
are of the same value, but with different algebraic signs and are orthogonally directed to each
other. Therefore, one acting muscle can be indicated by one single force or by a bunch of forces,
which in sum refer to the total force value.
3.4.4 Bearings
A FE model, which is positioned in a 2-D, or 3-D space must be restricted in all three or six
degrees of freedom, respectively. Otherwise the model would move according to internal or
external loading. This is accomplished by application of bearings to nodes in the model, which are
indicated as triangles (Fig. 8). Bearings restrict movement in all degrees of freedom or in required
directions, which refers to the actual situation. For example, if only vertically translational
movements should be restricted, only y-directed bearings are required. After calculation reaction
forces can be read out from some of the restricted nodes. These bearings are then termed
indicator bearings, as the relative movement of the model is indicated by the direction of the
resulting reaction forces at this location. This is a useful instrument to gain information about the
statical situation of the model during adjustment of equilibrium in the FE model and to determine
the influence of the applied link elements or forces.
3.4.5 Stress distribution
After the calculation the distribution of the compressive stresses in the FE model is shown
by a contour plot. The spectrum of colours in the contour legend is according to the selected
numeric spread. The regions with high compressive stresses appear blue; the regions with low
compressive stresses are red, tensile stresses are coded in opposite direction. Grey colours
indicate regions, in which stresses are beyond the selected spectrum (Fig.8). In order to decrease
areas, which are beyond the spectrum, the spread can be adapted to the maximum compressive
stress value. This in turn decreases the amount of indicated lower stresses and has to be applied
according to the respective situation, whereas in some cases even the lower stressed areas
contain necessary information.
29
3.4.6 Contact elements
To enable relative movements and transmission of mechanical stresses between the
adjacent parts in the model, contact-elements were applied to the contact and target surface
between the trunk and the scapulocoracoid element, as well as between the humerus and glenoid
surface and within the elbow joint. Contact element properties depend on the friction coefficient,
which determines the resulting friction between the adjacent areas.
3.4.7 Reduction of the bauraum
The term bauraum is here used exclusively for the scapulocoracoid volume, as the
synthesis refers only to this element. The other elements (trunk, humerus, antebrachium) act as
surrogates for the original components. They are required to allocate the muscle forces in the 3-D
space and serve as mechanical counterparts to simulate the situation in an active shoulder girdle
system.
In a number of iterative steps, the primary defined volume is reduced by selecting the
stress bearing areas. The bauraum is modeled only in an approximation to the dimension of the
original structure, and therefore leaves ample space for the spreading stresses. Consequently, the
resulting stresses after calculation and before the first reduction step are below the threshold of
the predicted physiological value (-2 MPa). The numeric spread of the contour legend is selected
to reflect the approximate shape of the desired structure, besides the absolute stress value. As
compressive stress value is defined by compression per unit area, a decrease of the area will
correspondingly result in an increase of the resulting stresses, since the bauraum is reduced. As
soon as the stresses are above the critical threshold of -2 MPa the regions of the geometrical
model with stresses below the threshold of -2 MPa are removed simulating the atrophy of the
unloaded bone. In early reduction steps, the threshold can be adapted to lower values, in order to
preserve the structural integrity of the model.
The reduction of the bauraum is completed for the selected cross-sections of the
scapulocoracoid volume. After the reduction of the selected cross-sections, the new reduced
bauraum is formed. In the next FESS iteration, the FE mesh is generated for this reduced bauraum
and the stress distribution is obtained under the initial boundary conditions. The reduction and FE
calculation steps are applied iteratively until a physiological level of compressive stresses is
reached in the reduced model. In the present investigation, a physiological compressive stress
value for bone between -2 and 20 MPa is predicted. Values betwee -2 MPa to -9 MPa corresponds
to cancellous bone; for compact bone a range of -9 MPa to -20 Mpa is assumed (Witzel &
Preuschoft, 2005).
30
Figure 8 2-D FE model of a simple beam (100mm x 50 mm), indicating the symbols for the boundary conditions, which
are described in the text. Youngs modulus is 17 GPa, Poisson´s ratio is 0.3. A: Meshed 2-D FE model before calculation,
restricted by bearings (orange triangles) at the right side; one single force (10N) is placed at the left side (red arrow).
After calculation the beam is subject to bending, which compresses the ventral and stretches the dorsal side of the
beam (B,C,E,F). B: 2-D FE model after calculation with the highest compressive stresses (blue-green areas) appear at the
ventral side. Low compressive stresses (red) can be detected at the dorsal side. Grey areas are beyond the visible
spectrum in selected spread. Numeric spread is -2.7 MPa to 0 MPa, indicated at the contour legend. Deformation is
indicated by black outlines showing the situation before calculation. C: FE model after calculation with the highest
tensional stresses at the dorsal side (red-orange). Numeric spread is 0 MPa- 2.7 MPa, indicated at the contour legend.
Again grey areas are beyond the visible spectrum in the selected spread. D: Meshed 2-D FE model with a pre-tensioned
link element (green line) instead of a directed force. Cross section of the link element is 1; Pretension is 0.1. E: Resulting
compressive stress distribution after calculation. Numeric spread of the contour legend is -9 MPa to 0 MPa. F: Resulting
tensional stress distribution after calculation. Numeric spread is 0 MPa- 9 MPa. Description of stress distribution is
equal to B and C.
3.5
3-D FE models of two generalized tetrapods
For investigation of basic stress distributions in tetrapod bodies during terrestrial
locomotion and for evaluation of the predicted differences between a sprawling and extended
limb positions of the forelimbs, two rather basic 3-D FE models are created.
31
Overall anatomy and relative mass distribution of the first FE model corresponds to that
of Tiktaalik roseae an early Devonian tetrapod representing the sprawling-type (Fig. 9,10,11).
Tiktaalik roseae was chosen by the author because of its primary morphological features and its
palaeontological relevance as a basic tetrapod. The model shows only the forelimbs, whereas the
hindlimb section is rested on the ground (Fig. 11). This is due to two reasons. First, the study is
focused on the forelimb section and the pectoral girdle. Second, the hindlimbs of Tiktaalik roseae
were not available at date, so that no reliable data could be obtained from the fossil (Shubin,
personal communication). The measurements are based on own measurements of the original
fossil (postcranial skeleton) completed by the author (Fig. 9); and the scaled figure reported in the
original description (skull) (Daeschler et al. 2006) (Fig. 10).
Figure 9 Shoulder girdle of Tiktaalik roseae (NUFV 108) in lateral (A), frontal (B) and dorsal view (C).
32
.
Figure 10 Dimensions of the Devonian fossil Tiktaalik roseae (redrawn after Daeschler 2006), in lateral view (upper
sketch) and in dorsal view (lower sketch). Dimensions are given in mm.
The models are subjected to gravitational forces, whereas the distribution of both
compressive and tensile stresses is described after calculation. First, the static condition is
characterized by a symmetrical limb support at the shoulder region with both forelimbs
supporting the body weight. In the second case, the asymmetrical support during consecutive
phases of a movement cycle is represented by a one-limb support at the shoulder region; a
situation similar to that in walking. All parts of the models possess equal material properties to
ensure that the stress flow is undisturbed and not bundled in the areas with higher stiffness.
Young´s modulus is 10 MPa; Poisson`s ratio is 0.3, both refer to the average values estimated for
vertebrate bone (Carter & Beauprè, 2001; Kummer, 2006). Limbs with ground contact show
bearings, which restrict movements in the y-axis. Forces (red arrows) applied to the models
indicate the limbs held off the ground like in walking (Fig. 11, 12). The weight is defined by the
model’s volume, density (equal to that of water) and the gravitational acceleration (9.82 m/s²).
Figure 11 FE-model of the “sprawling” type in oblique (A) and dorsal view (B). Left forelimb is held off the ground (acting
forces; red arrows), right forelimb has ground contact (bearings; yellow triangles). Hindlimb section is resting on the
ground (bearings; yellow triangle). Coordinate system is placed at the right corner
33
The second model reflects the situation in extant cursorial mammals with the position of
the limbs under the body (extended-type). The model is characterized by a rounded trunk, a long
neck and massive head and columnar legs held under the body (Fig. 12). Overall mass
distributions refer to a medium-sized equine tetrapod.
Fig. 12 FE-model of the “extended” type in oblique view. Left forelimb and right hindlimb have ground contact
(bearings: yellow triangles), right forelimb and left hindlimb are held off the ground (acting forces: red arrows).
Coordinate system is placed at the right corner.
3.6
2-D FE model of a crocodilian shoulder girdle
In order to investigate the static conditions in the crocodilian shoulder girdle an FE model
is created, where the single shoulder girdle elements are placed in the observed anatomical
position. Pictures in frontal and lateral view were taken in order to define the dimensions of each
element. The pictures were transferred to millimetre paper in order to generate the geometrical
coordinates for the FE modelling as input for the FE program, building two 2-D models in frontal
and lateral view (Fig. 13). By this procedure the bauraum of the present model is defined by the
dimensions of each element of the shoulder girdle system, the trunk, the scapulocoracoid, the
humerus, and the antebrachium. Therefore the elements of the FE model can be viewed as
substitutes for the original structure. In contrast to the finite element structure synthesis (FESS)
(Witzel & Preuschoft, 2005), the compressive stress distribution in this case is only analysed, and
therefore leaves out the final synthesis of the structure. Nevertheless, the basic movements and
conditions in a FE multi-body system can be revealed and used for the following 3-D FESS.
34
.
Figure 13 Pictures of a Caiman skeleton, in frontal (A) and lateral (C) view. Dimensions of the original, indicated as black
outlines, are transferred into coordinates, in order to build the bauraum of the 2-D FE model in frontal (B) and lateral
(D) view.
Two 2-D FE models, one in frontal (Fig. 14), and one in lateral view of the object (Fig. 15),
are calculated. The model consists of four single areas, the trunk, scapulocoracoid, humerus, and
antebrachium (Fig. 14,15). As one limb stance is in the focus of the current investigation, only
one-half of the body is modelled in order to minimize the calculation time. Material properties are
Young`s modulus (17 GPa), which is according to cortical bone and Poisson`s ratio (0.3) (Witzel &
Preuschoft, 2005). Body weight force (100 N) is applied to the trunk on one single node, which is
calculated out of an assumed body weight (10 kg) supported by the forelimb in one limb stance
multiplied with gravitational acceleration (9,81 m/s²). At the adjacent surfaces between the trunk
and scapulocoracoid, as well as within the joints, contact-elements are applied. Considering the
possible movements of each part of the model, link elements representing the relevant shoulder
muscles to maintain balance are determined. Each link element exhibits a Young`s modulus, equal
to vertebrate muscles, a pre-tension of 0.01 and a cross section of 0.1 mm². Relative motions of
single elements are constrained by bearings. The trunk is allowed to move vertically, whereas the
antebrachium is free to move sidewards (Fig.14,15).
35
Figure 14 Simplified 2-D FE model of the crocodilian shoulder girdle in frontal view. Trunk (light blue), scapulocoracoid
(orange), humerus (grey), antebrachium (deep blue). Link elements representing the shoulder girdle muscles: m.
pectoralis (1a,b), m. serratus (2), m. rhomboideus (3), m. subscapularis (4), m. deltoideus (5), m. coracobrachialis (6), m.
triceps caput coracoideum (7), m. brachialis (8). Bearings are applied to the trunk and the antebrachium in order to
allow vertical and horizontal movements, respectively.
Figure 15 Simplified 2-D FE model of the crocodilian shoulder girdle in lateral view. Trunk (light blue), scapulocoracoid
(orange), humerus (grey), antebrachium (deep blue). Link elements represent: m. pectoralis (1a,b), m. serratus
superficialis (2a-d), m. serratus profundus (5), m. deltoideus (6), m. triceps caput humerale (7), m. triceps caput
coracoideum (8), m. triceps caput scapulare (9), m. trapezius (10), m. levator scapulae (11), m. latissimus dorsi (12), m.
scapulohumeralis (13), m. costo-coracoideus (14), m. brachialis (15). ). Bearings are present at the trunk and
antebrachium and allow vertical and horizontal movements, respectively.
36
In this part of the approach, link elements were applied instead of the directed forces,
which relate to the crocodilian shoulder girdle muscles. The origin and insertion sites of the
considered muscles are obtained from the literature (Brinkmann, 200; Meers, 2003) (Table 3). In
order to focus on the mechanics in the shoulder girdle system, the muscles need to be considered
according to their mechanical properties. The muscle functions, which maintain balance in frontal
and lateral planes, differ from each other. In frontal view, muscles are required, which act in the
x,y-plane, in order to stabilize the shoulder girdle latero-medially and dorso-ventrally. In lateral
view, cranio-caudal stabilization is maintained by muscular structures, which act in the y,z- plane.
The calculation is finished when the equilibrium in the shoulder girdle system is established by
determination of the required links, which refer to the shoulder girdle muscles.
3.7
3-D FESS of a crocodilian shoulder girdle
In the synthesis part of the study the basic conditions of the frontal and lateral 2-D model
are transferred into a 3-D model (Fig.16). This combination includes the coordinates (x, y, z) in
space, the arrangement of the link elements, and the restriction of movements of the model in
the 3-dimensional space. Again only one-half of the shoulder girdle is modelled in order to
decrease calculation time.
The model consists of four independent volumes, the trunk, scapulocoracoid, humerus,
and antebrachium (Fig.16). Similar to the 2-D model pre-tensioned link elements are used instead
of the direct force application. Body weight force refers to the 2-D FE model. Again material
properties are defined as Young`s modulus (17 GPa), according to cortical bone and Poisson`s
ratio (0.3) (Witzel & Preuschoft, 2005).
The relative motion of each single element in the FE model corresponds to the situation in
the living object. The hindlimbs prevent caudally directed movements, whereas medially directed
movements are prevented by the other body half. These restrictions are realized by bearings in
caudal and medial direction, respectively. The trunk should be able to move vertically and in
horizontal directions, while the antebrachium is restricted to vertical movements, by bearings
directed in the y-axis (Fig.16). Link elements are applied iteratively in order to determine their
function in maintaining the static equilibrium in the shoulder girdle. Each link element
corresponds to a shoulder girdle muscle based on the original crocodilian anatomy. Since
equilibrium in the shoulder girdle is accomplished, the scapulocoracoid and the humerus are able
to move free from constraints and only held in position by the acting links (Fig.16). The calculation
is finished when the equilibrium in the shoulder girdle system is established by determination of
the required links, which refer to the shoulder girdle muscles.
37
Figure 16 3-D FE model of the shoulder girdle region of Caiman crocodilus in oblique view. Trunk element is grey, with
applied weight forces (arrows) and bearings in x-and z-direction (red triangles) up-and downward movements are
possible). Scapulocoracoid element and humerus are light blue. No bearings are applied, indicating free range of
movement. The element representing the antebrachium is deep blue showing bearings in y-direction (red triangles;
fore-and sideward movements are possible).
3.8
3-D FESS Diplodocus longus
The sauropod scapulocoracoid, which is going to be synthesized using FESS method in the
present thesis, is part of the mounted skeleton of Diplodocus longus (SMF) (Fig.17a,b). The
bilateral sternal plates and claviculae are missing in this mount. For completion of the sauropod
shoulder girdle, the appearance and dimensions of the sternal plates and the claviculae from the
Diplodocus specimen HQ1 (SMA) are included (Fig.17,c,d). Similar to the FESS of the crocodilian
shoulder girdle, the investigation of the scapulocoracoid in Diplodocus longus starts by building
the bauraum based on the original dimensions of the bony elements. By contrast, the FESS in this
fossil requires additional considerations. As the arrangement of the shoulder girdle elements in
the skeleton are in the focus of the investigation, the mechanical preconditions for this
arrangement had to be expressed. Not until then, the original dimensions of the bony elements
can be transferred to a 3-D FE model of the shoulder girdle system. Subsequently, the body
weight force of the frontal part of the trunk is estimated, and the determination of the applied
muscle forces and calculation of static equilibrium in the model are described.
38
Figure 17 Diplodocus longus (SMF) in lateral (A) and frontal view (B). Note the distance between both coracoidea in
frontal view. The mount lacks the sternal elements and clavicles. Diplodocus HQ1 (SMA) in frontal (C) and lateral view
(D). Sternal elements and clavicles are present. A und D in lateral view: The different positions of the scapula between
the two mounts are visible.
3.8.1 Mechanical considerations for building the bauraum
The construction of the model including the bauraum for the FE structure synthesis is one
of the most crucial parts of the modelling, because the arrangement of the single elements has a
profound impact on the equilibrium and stress distribution in the system. The arrangement of the
bony elements and the musculature in FESS of the crocodilian shoulder girdle could be
established based on the well-known morphological and empirical data obtained from extant
species. In contrast, the shoulder girdle architecture in sauropods is largely unknown. The only
elements on which one can base the investigation of the body structure is presented by single
39
bony elements. Concerning the anatomical position of the bony elements, musculature, tendons,
and ligaments, or even cartilaginous parts only few information is available to date.
Consequently, the shoulder girdle elements need be reconstructed based on, first, basic
morphological conditions, which are valid in all extant tetrapods; and, second, based on
mechanical considerations in regard to the evolution of the tetrapod shoulder girdle and its
function in statical loading conditions. The morphological conditions can be obtained from the
morphological comparison in extant reptiles, lacertilians, and mammals investigated in this study.
The mechanical conditions refer to the results of the analysis of solid tetrapod bodies and the
FESS of the crocodilian shoulder girdle.
It is predicted by the basic principle in this approach that the form always follows the
function. According to this, in the course of the investigation, those mechanical conditions of the
shoulder girdle system are determined, which lead to the observed shape of the scapulocoracoid.
As the mechanical conditions are ascertained with the aid of FESS, the anatomical position of the
shoulder girdle elements as well as the required musculature is reconstructed. Furthermore, the
magnitudes of forces are determined and lead to the estimation of the muscle volumes required
in one limb stance. To confirm the obtained results, the reverse argument of the approach can be
used. This would imply that invalid mechanical considerations would prevent static equilibrium in
the shoulder girdle muscles and therefore a successful synthesis of the scapulocoracoid of
Diplodocus longus. To prevent circular reasoning, the model is built using only assumptions known
to fulfil the mechanical necessities required to maintain the function of the shoulder girdle during
one limb stance. Function hereby means transmission of body weight from the trunk to the
forelimbs, whereby each element of the multi-body system is kept in equilibrium by the acting
muscles.
Mechanical considerations for the arrangement affect the inclination of the scapula, the
relative position of the scapulocoracoid to the ribcage, position of the sternal elements and finally
the posture of the humerus at the glenoid joint. Those preconditions, which can be confirmed by
the successful synthesis of the scapulocoracoid, will provide ground for reconstruction of the
shoulder girdle elements. The reconstruction of the shoulder girdle in Diplodocus longus is
presented in the results.
The bauraum representing the scapulocoracoid is positioned more cranially in comparison
to the original and lies parallel to the lateral side of the trunk. The caudal edge of the bauraum
exhibits an angle of about 60° to the horizontal. The coracoid part of the bauraum contacts the
most ventral part of the trunk medially. Two positions of the humerus are considered. In the first
40
limb position the humerus in frontal view is directed orthogonally to the glenoid joint, while in
lateral view an angle of about 20° to the perpendicular is assumed (Fig.18a,b). In the second
model, the humerus is slightly abducted in frontal view, with an angle of about 15° to the
perpendicular, while lateral position is similar to the first case (Fig.18a,c). Both models reflect the
static situation occurring in one limb stance. Total body weight of the frontal part of the trunk is
carried by the forelimbs, while the rest is assumed to be supported by the hindlimb section.
3.8.2 Generation of the model
In this part of the study, pictures were taken from the mounted skeleton in frontal and lateral
view (Fig.17). Skeletal elements of the Diplodocus specimen taken into account comprise the
mounted ribcage, the left scapulocoracoid, left humerus and the left lower forelimb. Dimensions
were first determined by drawing the pictures of different views to millimetre paper afterwards
coordinates are transferred to the FE program to generate the geometric 3D-model. Similarly, to
FESS of the shoulder girdle in crocodiles only one half of the body is modelled in order to reduce
calculation time (Fig. 18).
The geometry of the bauraum is according to the dimensions of the original
scapulocoracoid, but leaves ample space for the occurring stresses. The humerus is reduced to a
rod and two spheres, which represents the humerus shaft and its proximal and distal joint
surfaces, respectively and according to the original dimension of the fossil (Fig. 18). The element
representing the lower forelimb is modelled as a cube with an impression for the distal humerus
joint surface to serve as the elbow joint. The trunk includes the frontal part of the body stem up
to the seventh cervical rib and the eighth thoracic rib. The most cranio-ventral part is extracted to
provide a contact surface to the medial side of the scapulocoracoid. Young`s modulus in the
model is 17 GPa, Poisson`s ratio is 0.3 (Witzel et al. 2010). Contact elements are applied to the
contact areas of the shoulder and trunk, the coracoid and sternum as well as in the glenoid and
elbow joint to provide force transmission. Further contact elements are applied along the
adjacent areas of the scapula and the trunk..
41
Figure 18 3-D model of the shoulder region of Diplodocus longus (A-D). In lateral view, with a flexion of the humerus
relative to the glenoid joint of about 20° (A). Forelimb position one with the humerus directed vertically to the glenoid
joint (C). Forelimb position two, with the humerus slightly abducted (20°) (D). Outline of the model in lateral view (E).
Body weight forces (arrows) are distributed along the trunk. Bearings at the trunk are positioned in medial and anterioposterior direction to prevent relative movements; up and downward movements are enabled. Fixed bearings at the
antebrachium prevent all relative movements. No bearings are applied at the scapulocoracoid and the humerus and will
be held in position only by muscle forces, which are determined during the investigation in further steps.
3.8.3 Estimation and distribution of body weight
For investigation of body weight transmission in the shoulder girdle, total body weight in
Diplodocus has to be determined. Body weight represents the initial forces, countered by muscle
forces in order to reach equilibrium in the shoulder girdle system. Determination of body weight
in Diplodocus was calculated from body volume, specific density and earth acceleration. A body
volume of about 13421 m3 for Diplodocus is taken from the literature (Henderson 1999). Under
consideration of the assumed pneumatised postcranial bones, specific density in the living animal
was supposed to be 0.8 (Wedel 2003a, 2003b, 2005; Gunga 2007). Thus, total body mass can be
stated as 10,736 t.
42
In this investigation one limb stance is approximated to the situation in elephants. In
elephants body weight is commonly supported by at least three limbs during locomotion; or by
one forelimb support during every movement cycle (Hutchinson et al., 2007; Ren et al., 2008).
According to Henderson (2006) the centre of mass in Diplodocus is located anterior to the pelvis.
This is calculated for the present study as 11% of 2.5m, which is the total distance between the
shoulder and pelvic joint. Therefore, average load on the forelimbs is 18% of the total body
weight. In the present model this accounts to 966,24 kg acting on each forelimb in four limb
support. If one forelimb is held off the ground this mass is doubled to 1932,48 kg. Multiplication
with earth acceleration (9,81 kg*m s-²) results into a load of about 18957 N during one limb
stance and 9478 N if both forelimbs contribute to body support, respectively. Total body load of
about 18958 N is distributed evenly along the body axis of the trunk volume.
3.8.4 Muscle reconstruction
Currently no complete and reliable description of the shoulder girdle musculature in
sauropod dinosaurs exists. The comparison with the shoulder girdle musculature of extant
crocodiles only provides basic information about the equipment and configuration of the shoulder
girdle musculature applied to the FE model (Tab. 2 ). Therefore, the origins and insertions sites of
the muscles, and their lines of action have to be determined during the process, according to the
mechanical requirements in each calculation step. Similar to the FESS of the crocodilian shoulder
girdle, the muscle forces are applied to the system in iterative steps, in order to reach equilibrium
between the trunk, scapulocoracoid and humerus, whereby the situation in the elbow joint is not
considered in detail. Since static equilibrium in the shoulder girdle is accomplished, muscle forces
exert mechanical stresses onto the bauraum of the scapulocoracoid. As mentioned in the former
paragraph, the distribution of compressive stresses should be in accordance with the fossil.
Consequently, the former precluded muscle forces and the arrangement of bony elements are
confirmed.
3.8.5 Calculation of static equilibrium
In the following, static equilibrium of the 3-D multi-body FE model is calculated, first, with
regard to the principle stress distribution within the bauraum of the scapulocoracoid. Since static
equilibrium is maintained, the stresses are refined by modification of the acting forces. In the
shoulder girdle system, equilibrium results from the weight force (Fg) and the ground reaction
force (Fgr). Equilibrium is maintained, when the sum of forces, acting in the shoulder girdle
system, is near zero. In the present case, this can be accomplished by reduction of the rotational
moments at the joints and translation of the elements by means of muscle forces. The initial
43
situation, where the only applied forces relate to the body weight, is the basis for each calculation
step. The relative motion of the single elements in the shoulder girdle FE model corresponds to
the situation in one forelimb support. As it was described in case of the crocodilian FE model, the
hindlimbs in the living animal usually prevent caudally directed movements. The medially directed
movements are prevented by the contralateral body-half. The restrictions in the FE model are
indicated by bearings in caudal and medial direction, respectively. The trunk element is able to
move vertically, while the antebrachium is restricted to downward movements, by bearings in the
y-direction. The scapulocoracoid is only restricted by indicator bearing, until equilibrium is
accomplished. Then the scapulocoracoid and the humerus are free from restrictions and
positioned only by the acting forces, which are related to the shoulder girdle musculature.
44
Chapter 4
4.1
Results
Morphology of the shoulder girdle
4.1.1 Comparison of extant reptiles and mammals
The following figures illustrate the morphology of the shoulder girdle in extant reptiles
(Fig. 19,20,21) as well as in extant mammals (Fig.22). The examined reptiles belong to five
different orders, Iguania, Sphenodontidae, Cordylidae, Scincidae and Crocodylia. The described
mammals comprise the Artiodactyla (Equus cabalus, Rupicapra rupicapra), Felidae (Panthera leo)
and one Xenarthra (Bradypus spec.). Structural relationships such as the relative position of the
shoulder girdle to the ribcage and forelimbs in reptiles and mammals, as well as the morphology
of the coracosternal joint in reptiles are displayed.
In lateral view, the shoulder girdle in extant reptiles and mammals is always positioned
cranially covering the first ribs, or even in front of the ribcage. In top view, the scapula is always
positioned nearly parallel to the vertebral column. In lateral view the distal margin of the osseous
scapula or the cartilaginous suprascapula extend nearly parallel to the dorsal margin of the
vertebral column. The lateral inclination of the scapula, determined as the angle between the
caudal margin of the scapula and the horizontal plane, ranges from 50° in crocodilians to 60° in
mammals and chameleons.
In all examined reptiles a ventrally positioned tongue-and-groove coracosternal joint is
present. In lacertilians this joint is bony (Fig.19b,c,f; Fig.20b,c; Fig.21e,f), the sternal part is acting
as the groove and the coracoid part as the tongue, together forming a sliding rim. Basiliscus
plumifrons hereby makes an exception. The most cranial part of the medial side of the coracoid
consists of cartilage, whereas the caudal part is osseous. The sternal element is a compact bony
structure. In crocodilians those parts of the sternal and coracoid elements that form the tongue
and groove system consist of cartilage, which fulfils the same function (Fig.21a,b). For a more
detailed observation of the condition in crocodiles see the histological cross sections of the
coracosternal joint in one caiman (Fig.22,23).
45
Figure 19 A: Cyclura nubila, shoulder girdle position in top view. B:Cyclura nubila, dorsal view of coracosternal joint,
arrow points to tongue and rim structure. C: Basiliscus plumifrons, oblique view of the shoulder girdle in relation to the
coracosternal joint (arrow) and the ribcage. D Basiliscus plumifrons, top view of the parallel position of the dorsal
margin of the scapula. E: Iguana iguana, top view of the parallel position of the dorsal margin of the scapula. F: Iguana
iguana, frontal view at the coracosternal joint (arrow). Note the most cranial part of the coracoid is cartilaginous
46
Figure 20 A:Cordylus giganteus, lateral view of the relative position of the shoulder girdle. B: Cordylus giganteus, frontal
view to the coracosternal joint. C: Tiliqua rugosa, top view of the parallel position of the dorsal margin of the scapula. D:
Tiliqua rugosa, dorsal view of the coracosternal joint (arrow). E: Sphenodon punctatus, lateral view showing the cranial
position of the scapulocoracoid overlying the first ribs, together with a slight inclination of the scapula and parallel
position of the dorsal margin of the suprascapula.
47
Figure 21 A: Crocodilus niloticus, lateral view of the overall position of the shoulder girdle, the inclination of the scapula
and its orientation parallel to the ribcage. B: Crocodilus niloticus, frontal view of the cartilaginous parts of the
sternocoracoid joint forming a rim and tongue structure. C: Chamaeleo jacksonii, lateral view. The shoulder girdle is
positioned in parallel to the ribcage, the glenoid lies anterior to the ribcage and the scapula overlies the first ribs. D:
Chamaeleo melleri, lateral view. E: Chamaeleo melleri, ventral view of the coracosternal joint. F: Chamaeleo lateralis,
ventral view of the coracosternal joint.
48
Figure 22 A:Equus caballus, lateral view, The scapula covers the first ribs, while the glenoid region is positioned in front
of the ribcage. B: Equus caballus, view through the ribcage from posterior showing the narrowed trunk at the level of
the shoulder girdle and its parallel position relative to the ribcage. C:Bradypus spec., lateral view showing the relative
position of the shoulder girdle and dorsal margin of the scapular. D: Leo panthera, lateral view of the shoulder. The
scapula covers the first dorsal ribs, the glenoid is positioned shortly before the rib cage. E: Rupicapra rupicapra, lateral
view displays the shoulder girdle overlying the first dorsal ribs, while the glenoid is positioned anterior to the first rib. F:
Rupicapra rupicapra, dorsal view of the parallel position of the scapula relative to the ribcage.
4.1.2 Histology of the coracosternal joint in Caiman spec.
The coracosternal joint in one caiman was dissected, described and a number of cross
sections were taken for further preparation. The transverse sections through the coracoid and
49
sternum were stained and prepared for histological description. The pictures show the
coracosternal joint consisting of cartilage. The cartilaginous part of the sternum forms a groove,
while the coracoid acts as the tongue (Fig. 24). The joint is embedded in connective tissue and is
surrounded by a joint capsule. Thus, the coracosternal joint exhibits all features of a real joint,
compared to other joints as the glenoid joint or the knee joint.
Figure 23 Shoulder girdle of Caiman spec. in ventral view. The scapula (Sc) and the coracoid (C) are drawn laterodorsally to open the coracosternal joint. The cartilaginous tongue (coracoid) and groove (sternum) are visible.
For investigation of the internal structure of the coracosternal joint, transverse sections
were taken in anterio-posterior direction through the articulated joint (Fig. 23). The cross sections
were stained to distinguish the two types of tissue, bone and cartilage. In the coracoid tongue a
proximal bony part and a distal cartilaginous part are present, whereas the distal part surrounds
the bony structure of the coracoid tongue (Fig.24 a-c). In the sternal part only cartilaginous tissue
can be recognized (Fig. 24 a-c). In the most anterior section the sternal rim is compact cartilage at
the dorsal side of the rim and consist of connective tissue fibers at the ventral side. The
connective tissue fibers surround the ventral side of the sternal rim and of the coracoid (Fig. 24 ab). In the most posterior section the entire rim consists of cartilage, which surrounds the distal
coracoid tongue (Fig. 24c).
50
Figure 24 Frontal cross sections through the sternocoracoid joint in Caiman crocodilus. Cranial section (A); middle
section (B) and caudal section (C). Bony tissue is stained red; cartilaginous tissue is stained light blue. The tongue
(coracoid) and groove (sternum) system of the scapulocoracoid joint is clearly visible.
4.2
3-D FE models of solid tetrapod bodies
The first model represents a sprawling limb position in an early tetrapod. The second
refers to an extended position of the limbs as in extant cursorial mammals. In symmetrical limb
support, both limbs are restricted, marked by bearings (triangles). The forces (arrows) indicate the
swinging limb in asymmetrical limb support. The FE models are loaded by gravitational force,
which is indicated as a single arrow at the centre of the model or at the coordinate system.
51
For each loading case, in both locomotor types, the calculations are presented in lateral,
dorsal, and ventral view, showing compressive and tensile stress distributions. In asymmetrical
stance, the stress distributions differ between the right and left body side, and are shown
separately. To show the internal compressive stress distribution, frontal cross sections through
the hindlimb region, the middle of the trunk and at the level of the forelimbs, are given. High
compressive stresses are blue, low stresses are red. Colour code for tensile stresses is indicated by
opposite colours. Absolute stress magnitudes can be obtained from the heading legend
(DMX=total stresses, SMN=maximum tensile stresses, SMX=maximum compressive stresses).
The stress distribution is related to morphological structures, which are required to
sustain the mechanical demands. Compressive stresses are sustained by bony elements, whereas
muscular and tendinous structures intercept the tensile stresses. Bony and muscular elements are
identified and related to structural analogues, according to the amphibian, reptilian and mammal
bone and muscle terminology (Starck, 1979; Romer & Parsons, 1977; Nickel et al. 1996). Although,
the presence of a pectoral muscle, similar to m. pectoralis, was predicted for early tetrapods,
there is little information on the shoulder girdle musculature (Diogo & Abdala 2007; Diogo et al.,
2009). Therefore, with regard to the situation in early tetrapods, the description focuses on the
bony elements.
4.2.1 Early tetrapod
4.2.1.1 Symmetrical stance compressive stresses
During symmetrical stance the model indicates rather low maximum stress values of
about 0.9 MPa. Highest compressive stresses occur at the insertion sites of the limbs, spreading
dorsally over the midline of the trunk to the dorsal part of the constrained pelvic region (25a,b).
The dorsal stresses indicate the presence of a vertebral column (Fig.25b), while the stresses on
the lateral side refer to the presence of a scapula in extant reptiles (Starck, 1979; Romer &
Parsons, 1977) or the cleithrum in early tetrapods. By contrast, the dorsal part of the neck region
exhibits lower compressive stresses (Fig.25b). In ventral view compressive stresses are most
visible between the supporting limbs spreading cranially to the neck region indicated by the least
diameter, corresponding to the coracoid and interclavicles in extant reptiles (Starck, 1979; Romer
& Parsons, 1977) or clavicle in early tetrapods, respectively (Shubin et al. 2006; Coates, 1996).
52
Figure 25 3-D FE-model based on the fossil of Tiktaalik showing the distribution of compressive stresses in lateral (A),
dorsal (B) and ventral view (C). Contour legend (below) shows range of compressive stresses (colour coded); DMX=total
stresses, SMN=maximum tensile stresses, SMX=maximum compressive stresses. Both limbs are constrained
representing a symmetrical stance.
In the frontal cross sections, the compressive stresses at the level of the supporting
hindlimb run symmetrically from the fixed limbs up to the dorsal side (Fig. 26 A). At the middle of
the trunk, the distribution is stratified with the highest values dorsally, to stress free regions
ventrally (Fig. 26 B). Note the high stress values at the insertion site of the supporting forelimbs,
spreading ventrally. Lower values are present dorsally, while the centre is free of stresses (Fig.26
C). The cross sections support the observations shown in Figure 24. Further, the stress free region
in the middle of the trunk is interpreted as the body cavity, which provides space for the
intestines in all extant and extinct tetrapods.
53
Figure 26 Cross sections through the FE-model in symmetrical stance showing compressive stress distribution, starting
at the hindlimb section. A: At the level of the supporting hindlimb section. B: At the middle of the trunk. C: At the level
of the supporting forelimbs. For convention see Fig. 24.
4.2.1.2 Symmetrical stance tensile stresses
Tensile stress values reach a maximum of about 0.1 MPa and are therefore significantly
lower than the occurring compressive stresses (Fig. 27). Highest values of tensile stresses are
present at the ventral side of the trunk near the pelvic region representing abdominal muscles as
m. rectus abdominis in extant reptiles and at the dorsal side of the neck region representing neck
and shoulder muscles, as e.g. m. sternohyoideus, m. sternomastoideus, m. sternocleidomastoideus
and m. spinalis capitis in extant reptiles (Starck, 1979; Romer & Parsons, 1977). No information is
available about muscular structures in early tetrapods. Moderate stresses occur ventrally
between the supporting limbs, which again can be sustained by pectoral muscles, such as m.
pectoralis in extant reptiles (Starck, 1979; Romer & Parsons, 1977).
54
Figure 27 3-D FE-model of an early tetrapod showing tensile stresses in lateral (A), ventral (B) and dorsal (C) view. Both
limbs are constrained in symmetrical stance.
4.2.1.3 Asymmetrical stance compressive stresses
Maximum compressive stress values in asymmetrical support are with up to -43 MPa
significantly higher than in symmetrical stance. In detail, regions showing significant compressive
stresses reach values up to the 2nd power higher than in standing. Stresses are most pronounced
at the insertion sites of the weight bearing limb in contrast to the swinging limb (28 A,B). At the
ventral side between the forelimbs stress values increase and are most pronounced at the
supporting limb spreading medially (Fig. 28 D). Stresses run from the supporting limb obliquely
over the trunk to the level of the resting hindlimb, or in this case the ground-contacting belly (Fig.
28 C). The compressive part of the torsional stresses can be counteracted by compression
resistant ribs, located in the body wall (Starck, 1979; Romer & Parsons, 1977).
55
Figure 28 3-D FE model of an early tetrapod showing compressive stresses in asymmetrical stance. A: right side. B: left
side. C: dorsal view. D: ventral view.
Again, frontal cross sections of the FE model are shown. At the level of the supporting
hindlimb stresses run asymmetrically from the bearing dorsally to the anterior part of the body
(Fig. 29 A). At the middle of the trunk stresses concentrate at the periphery of the trunk leaving
the centre unstressed (Fig. 29 B). Note the highest stress values occur at the insertion site of the
supporting limb and ventrally between the forelimbs. Stresses clearly connect both sides
ventrally. There is a connection on the dorsal side as well, but to a much lower degree. The
alternation of symmetrical and asymmetrical support during locomotion in reptiles therefore
requires a continuous compressive resistant structure medially between the limbs, thus
confirming a ventral coracoid and interclavicula for extant reptiles (Starck, 1979; Romer &
Parsons, 1977) or the frontally positioned clavicles, as in early tetrapods (Shubin et al. 2006;
Coates, 1996).
56
Figure 29 Frontal sections through the FE-model in asymmetrical stance showing the compressive stress distribution.
Supporting limb is right, swinging limb is left. The coordinate system is visible in the middle of the trunk. A: At the level
of the supporting hindlimb. B: At the middle of the trunk. C: At the level of the supporting forelimbs.
4.2.2 Cursorial mammal
4.2.2.1 Symmetrical stance compressive stresses
In symmetrical stance the maximal compressive stresses in the model with an extended
limb position reaches 14 MPa. High compressive stresses occur at the insertion sites of the limbs,
spreading dorsally, with a slightly caudal inclination over the midline of the trunk to the hindlimbs.
Compressive stresses at this location can be counteracted by the scapula in extant mammals with
an extended limb position (Fig. 30 A,B). In ventral view compressive stress values between the
supporting forelimbs are low and not detecTable at the hindlimb section. In extant mammals, a
rod-like sternum is present at this location (Nickel et al. 1996) (Fig. 30 C). Further ventral stresses
spread cranially along the neck region. These stresses can be sustained the ventral position of the
cervical vertebrae, present in mammals with long necks (Fig. 30 C). The ventral part of the trunk is
57
subject to very low compressive stresses, which spread mainly from the extended forelimb to the
side of the trunk (Fig. 30 A,C).
Figure 30 3-D FE-model of a cursorial mammal showing compressive stresses in lateral (A), dorsal (B) and ventral view
(C). All four limbs are constrained in symmetrical stance. Bearing are shown as triangles and indicate the limbs with
ground contact. Coordinate system is located at the most caudal end of the model.
Cross sections confirm the former results of the symmetrical stance (Fig. 31 A-C). At the
middle of the trunk stresses prevail dorsally leaving the ventral side unstressed (Fig. 31 A). In both
fore- and hindlimbs stresses run symmetrically from the limbs up to the dorsal side (Fig. 31 B,C).
At the hindlimbs (Fig. 31 C), no ventral connection is visible, whereas at the forelimbs the cranially
spreading stresses, referred to the cervical vertebrae, are visible (Fig. 31 B).
58
Figure 31 Frontal sections through the FE-model of a cursorial mammal in symmetrical stance showing compressive
stress distribution. A:At the level of the supporting hindlimbs. B: In the middle of the trunk. C: At the level of the
supporting forelimbs.
4.2.2.2 Symmetrical stance tensile stresses
All limbs are constrained in symmetrical stance (Fig. 32 A-C). Highest values of tensile
stresses occur ventrally at the belly and the dorsal side of the neck region (Fig. 32 A,B). The
observed stresses can be assigned to muscular structures e.g. m. sternomastoideus, m.
cleidomastoideus and m. trapezius (Nickel et al. 1996). Moderate stresses can be observed on the
ventral side between the supporting limbs (Fig. 32 C). In extant mammals, the pectoral muscles
can intercept the stresses at this location ((Nickel et al. 1996).
59
Figure 32 3-D FE-model of a cursorial mammal showing tensile stresses in lateral (A) ventral (B) and dorsal (C) view.
4.2.2.3 Asymmetrical stance compressive stresses
In asymmetrical support the stress magnitude increases up to 35 MPa, the distribution is
altered. Stresses are most pronounced at the insertion sites of the weight bearing limbs in
contrast to the swinging limbs (Fig. 33 A,B). In lateral view (Fig. 33 A) the stresses at the forelimb
somewhat expand cranially, but spread mainly dorsally at a slightly caudad inclination. Stresses
spread dorsally from the supporting limbs to the contralateral hindlimb (Fig. 33 C). In ventral view
little compressive stress connecting both sides occur between both fore-and hindlimbs (Fig. 32 D).
Stress distribution in asymmetrical support corresponds to a laterally positioned scapula, in order
to sustain the mechanical stresses, as it is shown before in symmetrical stance. Again, no ventral
compression resistant structure is required in an extended limb position.
60
Figure 33 3-D FE model of a cursorial mammal showing compressive stresses in asymmetrical stance. The left fore- and
the right hindlimbs are within ground contact (triangles). The corresponding limbs are held of the ground (arrows) (see
D for description). A: right side. B: left side. C: dorsal view. D: ventral view.
The cross sections support former results. At the level of the supporting hindlimbs
stresses spread asymmetrically from the constrained limb dorsally to the frontal part of the body
(Fig. 34 A). Similar to the sprawling limb position stresses in the middle of the trunk are
concentrated at the periphery of the trunk leaving the centre unstressed (Fig. 34 B). This
observation is consistent with the situation in the sprawling model. Stress free regions can be
observed in both solid models, thus indicating a body cavity in tetrapods, which provides room for
the intestines (Fig. 34 B). Note the highest stress values at the insertion site of the supporting
limbs, where stresses spread mainly up to the lateral side of the trunk (Fig. 34 A,C). The ventral
aspect shows less compression in both fore- and hindlimbs, compared to the sprawling model.
61
Figure 34 Frontal sections through the 3-D FE-model showing compressive stress distribution in asymmetrical stance. A:
At the level of the supporting hindlimbs. B: At the middle of the trunk. C: At the level of the supporting forelimbs.
4.2.3 2-D FE model of a crocodilian shoulder girdle
In a second approach, the mechanics and stress distribution in a crocodilian shoulder
girdle are investigated. This first attempt of a 2-D multi-body FE model is used to reveal basic
information about the method, the compressive stress distribution and relative movements. The
calculated results are presented as one frontal (Fig. 35) and one lateral 2-D model of the pectoral
girdle in a crocodilian (Fig. 36). Black outlines indicate the initial situation before calculation. In
frontal view and in lateral view the trunk is restricted laterally by x-directed bearings (light blue
triangles), the antebrachium is restricted in the y-direction, only capable of sideward movements.
In lateral view, these constraints are likewise. Contact elements are applied to the adjacent areas
of the trunk and the scapulocoracoid as well as in the glenoid and elbow joint. The required
muscles, which are shown as link elements, are numbered, and indicated as black (initial
situation) and as red lines (after calculation). A coordinate system is placed right at the bottom.
62
Compressive stress values can be obtained from the colour coded contour legend or from the
heading legend, in which maximum and minimum stresses are recorded. High compressive
stresses are blue, low stresses are red. Areas with stresses beyond the spectrum are grey.
4.2.3.1 Static equilibrium
Although the relative motion of the elements is visible, the deformation remains in
tolerable limits, which is a precondition for static equilibrium in the system. In frontal view, the
trunk and scapulocoracoid sink downwards, while the humerus is adducted. The humerus shaft is
subject to bending. The antebrachium tends to move medially. Compressive stresses within the
scapulocoracoid flow from the glenoid joint medially and dorsally towards the trunk (Fig. 35).
Each of the link elements corresponds to muscles in the crocodilian shoulder girdle and produces
forces resulting in mechanical stresses within the predefined structure (Fig. 35). These muscles
contribute to transmission of weight force from the trunk to the pectoral girdle and the
supporting limb or are engaged in maintaining balance in the glenoid joint. Link elements
represented are m. pectoralis (1a, b), m. serratus (2), m. rhomboideus (3), m. subscapularis (4), m.
deltoideus scapularis (5), m. coracobrachialis (6), m. triceps caput coracoideum (7), m. brachialis
(8).
Figure 35 Frontal 2-D FE model. Distribution of compressive stresses after calculation. Black lines indicate the working
link elements representing the shoulder musculature. Black outlines indicate the situation before calculation. Link
elements represented are m. pectoralis (1a, b), m. serratus (2), m. rhomboideus (3), m. subscapularis (4), m. deltoideus
scapularis (5), m. coracobrachialis (6), m. triceps caput coracoideum (7), m. brachialis (8).
In lateral view, the relative motion of the elements is visible. Again, the deformation stays
in tolerable limits. The scapulocoracoid sinks downwards, while the humerus moves caudally.
63
There is little bending of the humerus in lateral view, the shaft is mainly subject to compression.
The antebrachium tends to move caudally. High stress values can be detected at the insertion site
of the limb, which spread from the glenoid joint cranially and dorsally, the inclination of the stress
pattern within the scapulocoracoid is clearly visible (Fig. 36). The results give a first approximation
of the position of the original morphological structures in the crocodilian scapulocoraoid (Fig.
35,36). Link elements represent m. pectoralis (1 a,b), m. serratus superficialis (2 a-d), m. serratus
profundus (5), m. deltoideus scapulare (6), m. triceps caput humerale (7), m. triceps caput
coracoideum (8), m. triceps caput scapulare (9), m. trapezius (10), m. levator scapulae (11), m.
latissimus dorsi (12), m. scapulohumeralis (13), m. costo-coracoideus (14), m. brachialis (15).
Figure 36 2-D FE model in lateral view, only areas representing the scapulocoracoid, humerus and antebrachium. Black
lines indicate the working link elements. Black outlines indicate the situation before calculation. Link elements
represent m. pectoralis (1 a,b), m. serratus superficialis (2 a-d), m. serratus profundus (5), m. deltoideus scapulare (6),
m. triceps caput humerale (7), m. triceps caput coracoideum (8), m. triceps caput scapulare (9), m. trapezius (10), m.
levator scapulae (11), m. latissimus dorsi (12), m. scapulohumeralis (13), m. costo-coracoideus (14), m. brachialis (15
4.3
3-D FESS of a crocodilian shoulder girdle
The present approach is the first attempt of a 3-D multi-body FESS of a crocodilian
shoulder girdle. Body weight force is applied at five nodes, which are distributed over the trunk.
Again, only the left side of the trunk is generated in order to minimize calculation time. Legends
and symbols are similar to the 2-D FE model. Pretension of links and reaction forces within the
indicator bearings for each calculation step can be obtained from the supplement.
64
In the initial situation, the trunk is restricted by bearings in x- and z-direction (triangles) to
prevent medially and anteriorly directed movements, which in the living object are carried by the
right supporting forelimb and the hindlimbs, respectively (Fig. 37 A,B). The antebrachium is
restricted to vertical and caudad movements. To determine the function of the required muscles,
they are applied to the model in iterative steps, considering their impact to the system, which can
be detected as reaction forces in the indicator bearings applied to the scapulocoracoid and
humerus. These indicator bearings provide stability in non-balanced calculations. Since the
shoulder girdle system is balanced, the reaction forces within the indicator bearings are zero and
are removed iteratively during the process. In the second setting, the trunk and the antebrachium
are allowed to move in the formerly defined directions and are still supported by bearings. The
scapulocoracoid and the humerus are free from restrictions and supported only by link-elements,
which relate to the shoulder girdle muscles. In the following, only the final calculation step, with
the shoulder girdle system in equilibrium, is presented. The entire process of equilibrium
adjustment can be obtained from the supplement (Fig. 37,38).
4.3.1 Static equilibrium
In the final calculation step, no bearings are required to keep the scapulocoracoid and the
humerus in position. The equilibrium of the shoulder girdle was accomplished by link elements,
which are related to the shoulder girdle muscles. The trunk and the antebrachium move in the
predicted directions. The scapulocoracoid rotates cranially around the pivot of the glenoid joint
and ventrally. The humerus rotates caudally around the pivot of the glenoid joint and cranially
within the elbow joint, while it shifts the antebrachium in caudal direction. The relative motion of
the elements stays in tolerable limits, which is a precondition for calculation. The relative motion
of the elements in the system and the observed force transmission between the trunk, the
scapulocoracoid and the humerus indicate an effective operation of the contact elements (Fig.
37,38).
65
Figure 37 Final calculation step in frontal view. Stress distribution (contour legend) and deformation (black outlines) in
the crocodilian shoulder girdle are presented. Equilibrium in the shoulder girdle system is maintained by links, which
relate to shoulder girdle muscles in crocodilians. Maximum and minimum stresses can be obtained from the heading
legend. Force transmission between the trunk and scapulocoracoid and within the joints can be recognized.
Figure 38 Final calculation step in frontal view. Stress distribution (contour legend) and deformation (black outlines) in
the crocodilian shoulder girdle are presented. Equilibrium in the shoulder girdle system is maintained by links, which
relate to shoulder girdle muscles in crocodilians. Maximum and minimum stresses can be obtained from the heading
legend. Force transmission between the trunk and scapulocoracoid and within the joints can be recognized.
66
The distribution of compressive stresses, which appear in the bauraum after reaching
equilibrium, gives a first approximation of the scapulocoracoid shape according to the original
bony elements in the mounted skeleton (Fig. 39 A,B). In lateral view, the stresses spread from the
glenoid joint caudad along the posterior border of the bauraum and posteriorily along the
supposed area of the coracoid towards the trunk (Fig. 39 B). Medial transmission of stresses from
the coracoid to the trunk again can be observed in frontal view (Fig. 39 B). Maximal compressive
stress is -20 MPa, whereas the indicated threshold is between 0 and -0.09 MPa. There is a large
amount of grey areas, i.e. stresses beyond the visible spectrum. The numeric spread was used to
point out the areas with the highest amount of compressive stresses, because absolute values of
compressive stresses are low at this point of the approach. To accumulate the stresses, the
volume of the bauraum is reduced according to the stress bearing areas, while the stress free
areas are removed in iterative steps. This reduction will lead to a refinement of the structure and
an increase of compressive stress value.
Figure 39 Comparison of the stress pattern in the final calculation step with the skeletal elements of the crocodilian
shoulder girdle (black outlines).
67
4.3.2 Muscle function
All links in the model relate to the shoulder girdle muscles in crocodilians (Fig.40) and are
described according to their mechanical function in static equilibrium. The majority of forces
exerted by the links act mainly in caudal (z-axis) and dorso-ventral (y-axis) direction, and therefore
are most visible in lateral view (Fig. 40).
The link elements, which connect the scapulocoracoid and the trunk, are related to m.
serratus superficialis, m. trapezius, m. levator scapulae, m. costo-coracoideus, m. triceps caput
scapulare and m. triceps caput coracoideum (Fig. 40). Their lines of action point mainly craniocaudally (z-axis), and in addition exert a lower amount of forces in the medio-lateral direction (xaxis). M. serratus superficialis is represented by four single links according to its distribution along
the insertion site at the scapula and the thoracic ribs. The muscle transmits the body weight from
the trunk to the shoulder girdle and acts as caudal rotator of the scapulocoracoid. M. trapezius,
m. levator scapulae and m. costo-coracoideus contribute to static equilibrium in the shoulder
girdle, as they rotate the scapulocoracoid cranially. They act as antagonists to the trunksuspending m. serratus superficialis, which rotates the scapulocoracoid caudad. M. costocoracoideus is found to contribute to static equilibrium, as it pulls the scapulocoracoid backwards.
As this muscle has a smaller lever arm to the pivot of the shoulder joint compared to m. trapezius
or m. levator scapulae, the main function is rather caudad translation than craniad rotation of the
scapulocoracoid. To increase the amount of required forces in m. costo-coracoideus and m.
trapezius, the diameter and pretension in the link elements are enlarge (Fig. 41).
M. biceps brachii and m. supracoracoideus run between the scapulocoracoid and the
humerus anterior to the glenoid joint and below its pivot (Fig. 40). Their line of action is directed
in the x,y-and z-axis. According to the line of action these links presumably support the former
muscles in cranial rotation of the scapula and the humerus. Their function cannot be determined
beyond doubt, because no visible impact was detecTable during adjustment of static equilibrium.
Although the function of the former muscles cannot be finally determined, they do account for a
decrease of maximum stress value in the final calculation of up to -20 MPa.
M. triceps caput scapulare and m. triceps caput coracoideum run between the
antebrachium and the scapulocoracoid (Fig. 40). The former inserts posterior to the glenoid joint
at the scapula, the latter, anterior to the glenoid at the coracoid. Functionally they rotate the
scapulocoracoid caudad and craniad, respectively. Their impact on this assumed function cannot
be determined during the process. In fact, they contribute to prevent a cranial translation of the
antebrachium, thus supporting equilibrium in the elbow joint, while pulling the antebrachium
68
forward. Reaction forces within indicator bearings at the antebrachium still indicate a strong
tendency for caudal movements, thus mm. triceps cannot account for reduction exclusively. To
prevent caudad movement of the antebrachium m. brachialis, m. spiralis and m. triceps caput
humerale were applied, as their line of action runs posteriorily from the antebrachium to the
humerus (Fig. 40). Again, no modifications could be detected. In consequence, horizontally
directed link elements were applied to the anterior border of the antebrachium to hold its
position and to keep the elbow joint in equilibrium. It is assumed at this point of investigation,
that further musculature, running between the antebrachium and the forefoot, is required.
In frontal view, medially directed forces at the model are mainly exerted by m. pectoralis,
m. rhomboideus and m. coracobrachialis. In the final calculation, the humerus moves medially
towards the trunk, as indicated by deformation. This is due m. pectoralis, which acts as an
adductor of the humerus (Fig. 40).
Further applied links, representing m. subscapularis, m. deltoideus scapularis, m.
subscapularis, m. scapulo-humeralis, m. teres major and m. latissimus dorsi show no detecTable
impact for maintaining equilibrium. In fact m. teres major and m. latissimus dorsi prevent static
equilibrium in any calculation step during the process, and are therefore not included in the final
calculation.
69
A
A
B
A
Figure 40 Shoulder girdle muscles investigated by the former calculation in frontal (A) and in lateral view (B). Boundary
conditions refer to the final calculation step. Weight forces applied to the trunk are distributed along the trunk (arrows).
According to their demands under static loading conditions, the link elements exhibit a
certain amount of mechanical stresses, which can be detected after calculation. In Table 8 the
resulting stresses are compared to the forces initially present in the links, defined by pretension
and diameter. The maximum stress values can be detected in the horizontally directed link
70
elements applied to the antebrachium. These elements have been applied to prevent caudad
movements of the antebrachium, but do not influence to equilibrium in the shoulder girdle
joint.The highest stresses are visible in links related to m. serratus superficialis, where the most
posterior links show the highest values (250-300 N/mm²) (Fig. 41). For links representing m.
triceps humerale, m. brachialis, m. deltoideus clavicularis, m. supracoracoideus and m.
costocoracoideus values of >150 N/mm² can be detected (Fig. 41). Due to the requirements of
static equilibrium, the forces exerted by m. costocoracoideus increased during the adjustment
process. Therefore, the cross section of the link representing this muscle is 10 times the initial
cross section of about 0.1mm². Further, pretension was doubled, thus leading to higher absolute
stress values. To enhance cranial rotation of the scapulocoracoid the cross section of m. trapezius
was increased as well, which results into significantly higher absolute stresses, than can be
detected in the link after calculation (50-100 N/mm²) (Fig. 41). Therefore, m. costo-coracoideus
and m. trapezius are structures in the system with the highest load. Lower stress values (100-150
N/mm²) can be viewed in m. biceps brachii, m. pectoralis major et minor, m. triceps caput
scapulare, m. triceps caput coracoideum, and m. subscapularis (Fig. 41). Finally, in the links
representing m. levator scapulae, m. spiralis and m. serratus profundus no stresses are detectable
after calculation (Fig. 41).
Figure 41 Diagram showing absolute value of link stresses after calculation in N/mm² for each link representing the
involved musculature. Note the highest stress values in m. serratus superficialis 1-4, m. supracoracoideus and m. triceps
caput humerale. M. levator scapulae, m. spiralis and m. serratus profundus exhibit no stresses.
71
4.3.3 Synthesis of the scapulocoracoid
The resulting compressive stresses of the final calculation step are accumulated in
iterative steps, in order to reduce the bauraum of the 3-D FE model. The designated shape then
corresponds to the original crocodilian scapulocoracoid. First, the 3-D volume of the
scapulocoraoid bauraum is dissected in antero-posterior direction (Fig. 42). Each section shows
areas with more or less compressed regions. The internal stress distributions reveal the
transmission of forces between the scapulocoracoid and the trunk as well as between the
humerus at the level of the glenoid joint (Fig. 42). Areas with a certain amount of compressive
stresses are selected. Subsequently the selected sections are spliced to form the new bauraum.
After remeshing of the reduced bauraum, the model is recalculated under the predefined
boundary conditions according to the final calculation step (Fig. 43). The compressive stresses are
accumulated in the reduced volume, so that compressive stresses will increase until they reach
physiological values according to compact bone (-2 to -20 MPa).
Figure 42 Selected cross sections of the model after the final calculation step in frontal view. Virtual dissection of the
model runs from anterior to posterior (1-6). Note the stress distribution in section 5 and 6, where stresses spread from
the glenoid joint medially to the trunk (coracoid) and dorsally along the bauraum (scapula).
72
Figure 43 Different steps for reduction of the bauraum are presented. A: Scapulocoracoid bauraum after calculation
showing the stress distribution in oblique view. White lines indicate selected areas, which show a certain amount of
compressive stresses. B: Model in lateral view. Position of selected sections is indicated as vertical lines. C: Spliced areas
of the selected sections, before remeshing of the new volume in lateral view. D: Remeshed new bauraum in its original
position together with the trunk, humerus and antebrachium (dashed lines) in oblique view.
After the 2
nd
recalculation in the 3
rd
reduction step an overall decrease of material in
comparison to the initial bauraum can be observed (Fig. 44). The virtual synthesis of the
scapulocoracoid approximately resembles the original skeletal part, as it exhibits a caudally
inclined slender form, similar to the crocodilian scapulocoracoid. The coracoid part of the
synthesis connects the trunk at the predicted position (Fig. 44). Further reduction steps would
probably lead to a more defined structure, but the amount of additional knowledge is assumed to
be insignificant. However, the investigation so far reveals necessary basic information about the
application of the FESS method to a multi-body 3-D shoulder girdle system. The observed relative
movements of the shoulder girdle elements, and basically defined mechanical muscle functions in
this part of the study provide ground for further steps. However, no quantitative muscle values
could be estimated. Consequently, in the FESS of the scapulocoraoid of Diplodocus longus the link
elements are replaced by directed forces in order to gain the required information.
73
rd
Figure 44 The 3 and last reduction step of the crocodilian scapulocoracoid model. A: Reduced scapulocoracoid
bauraum placed within the whole model, in lateral view (A), frontal view (C), oblique view (E). Single reduced
scapulocoracoid in frontal (B), lateral (D) oblique view (F).
4.4
3-D FESS of the scapulocoracoid of Diplodocus longus
3-D FESS of the scapulocoracoid of Diplodocus longus is performed via calculation of two
different forelimb positions. The first FE model shows the humerus in a vertical position relative
to the glenoid joint (forelimb position I.); in the second FE model (forelimb position II.) the
humerus articulates at the glenoid joint with an angle of about 15° to the perpendicular. Laterally,
both models exhibit an inclination angle at the humerus of about 20° to the vertical.
The elements of the shoulder girdle system are held in position first, by bearings
(indicated as triangles) and second, by forces preventing the movement. Therefore, two settings
with different boundary conditions can be determined. The first is the initial situation, where the
trunk is restricted in x- and z-direction to prevent medial and forward movements. In living
animals, this function is adopted by the right supporting forelimb and hindlimb, respectively. The
antebrachium is restricted by a fixed bearing. The scapulocoracoid and the humerus require
74
indicator bearings from which reaction forces can be read out after calculation. The indicator
bearings will be removed iteratively, leading to the second setting in which the trunk and
antebrachium are allowed to move in physiological directions. In equilibrium of the shoulder
girdle system, the scapulocoracoid and the humerus are free from restrictions and held in position
only by forces, which represent the shoulder girdle muscles. To determine the function of the
required muscles, the forces are applied to the model in iterative steps, considering their impact
on the system, which again can be detected at the indicator bearings. Changes concerning the
boundary conditions (applied forces and bearings) will be described in the figure legends.
Static equilibrium can not be accomplished in the first FE model. Nevertheless, this first
attempt reveals basic information with regard to the method and function dependent
implications of the diplodocoid shoulder girdle. To explain the consecutive nature of the method,
where each result provides the basis for the following step, the most relevant information drawn
from the adjustment of equilibrium in both FE models (forelimb position I and II.) is given. The
entire adjustment processes can be obtained from the supplement.
The results of the second FE model are presented as follows. The results comprise one
frontal and lateral overview of the model, to demonstrate the applied forces representing the
shoulder girdle muscles. Two contour plots, in frontal and lateral view, show the distribution and
magnitude of compressive stresses after calculation. Black outlines indicate relative motion of the
elements according to deformation after calculation. Contact elements are present at the
adjacent areas of the trunk and the scapula, in the region of the assumed coracosternal joint and
at the glenoid and elbow joints. Values of muscle forces applied to the model and reaction forces
within the bearings can be obtained from the supplement. Forces are shown as arrows. The
coordinate system is placed at the lower right corner. Compressive stress values can be obtained
from the colour coded contour legend or from the heading legend (maximum and minimum).
High compressive stresses are blue, low stresses are red. Areas with stresses beyond the
spectrum are grey.
After the static equilibrium in the FE model was accomplished, the scapulocoracoid of
Diplodocus longus is successfully synthesized with regard to the compressive stress distribution of
the second FE model. The synthesis resembles the characteristic features of the original
diplodocoid scapulocoraoid. According to the successful synthesis of the scapulocoracoid, the
precluded mechanical considerations concerning the shoulder girdle of Diplodocus longus are
confirmed. Based on the results the shoulder girdle is reconstructed, which includes the
75
arrangement of the bony elements, and the position, function and forces of the shoulder girdle
musculature.
4.4.1 Forelimb position I.
In a number of iterative steps (see supplement for entire calculation steps) equilibrium
was reached after completion of muscle arrangement running between trunk, scapulocoracoid,
humerus, and antebrachium, except for m. latissimus and m. teres major. No arrangement could
be established in which the latter muscles contributed to stability in the system. No bearings are
required to keep scapula and humerus in balance, though rotation and translation of the
scapulocoracoid and humerus are prevented by the applied forces. Maximum compressive stress
value is -25 MPa therefore closely corresponding to the formerly predicted physiological stress
values, ranging from -2 to -20MPa. Although equilibrium was maintained only by means of forces
the model did not fully correspond to the original. Especially the stress distribution within the
acromial region needed to be enhanced. In a first attempt, m. serratus superficialis was
repositioned and forces in m. trapezius and m. triceps scapularis were enlarged. None of these
approaches modified the distribution of stresses. Furthermore, the scapulocoracoid shows no
rotation. The lack of rotation of the humerus within the glenoid joint can obviously not be
affected by any rearrangement of muscle forces and is definitely no physiological condition.
Missing rotation probably results from an inhibited sliding between the humerus and glenoid
surface area. To avoid canting between the scapulocoracoid and humerus both elements were
remeshed using smaller finite-elements (Fig. 45 A). Further, the friction coefficient for the contact
elements was lowered from 1 to 0.1 (Fig. 45 B). Even these modifications were insufficient (Fig. 45
A,B). Nevertheless, refinement of the elements has a visible effect on the stress distribution. In
consequence the refined ground model will be applied to all following calculations (see
supplement).
Finally the convergence criteria had been lowered from 0.1 to 0.001. To preclude any
further influences the applied forces are reduced to the initial situation. Lowering convergence
then results into a significant decrease of deformation and an increasing rotation and sliding
between the contact areas of the scapulocoracoid and the humerus (Fig. 45 C,D; Fig. 46). Even
with an enhanced sliding between the elements of the shoulder girdle, no stress distribution,
resembling the shape of the original scapulocoracoid could be achieved (Fig. 45 C,D). In order to
reach a more elaborate distribution of stresses, the model will be recalculated, this time with the
humerus slightly abducted. As the ground reaction force supposedly flows in more medial
direction, this procedure should enhance compressive stresses in the coracoid element and the
acromial part of the scapula.
76
Thus, 3-D FESS was not accomplished, even though static equilibrium could be established
by forces. First, convergence criteria could be revealed as the most crucial factor for a decrease of
friction between areas provided with contact elements. The great numbers of contact elements
applied to the joint areas are sensitive to changes in convergence criteria. Decreasing
convergence criteria lead to a higher accuracy of the solution, but simultaneously increases
solving time and sensitivity to model discrepancies during calculation of static equilibrium. These
discrepancies are obviously responsible for the enhanced contact between the scapulocoracoid
and humerus, leading to non-physiological results. As an increased convergence criteria-setting
affects the distribution of stresses as well as deformation and state of equilibrium, all following
calculations will be calculated using a convergence criteria of about 0.001.
Figure 45 Calculations of the refined model showing a reduced amount of forces. M. serratus superficialis, m. serratus
profundus, m. levator scapulae, m. trapezius and m. costo-coracoideus are reduced to a single vector respectively.
Equilibrium between the trunk and scapulocoracoid is maintained, occurring stresses are plotted. The refined model
was calculated with a convergence criteria of 0.1 in lateral and frontal view of the object (A, B), and, for comparison,
with the convergence criteria set to 0.001. (C, D).
77
Figure 46 Magnified detail of the glenoid joint region showing the free rotation of the humerus within the glenoid fossa
indicated as black outlines showing the initial situation before calculation.
4.4.2 Forelimb position II.
Friction coefficient for contact elements is 0.01. Convergence criteria for calculation are
0.001. The refined model is calculated using the same boundary conditions as in forelimb position
I. Again, adjustment of static equilibrium starts with forces running between the trunk and the
scapulocoracoid. The scapulocoracoid is restricted to translational movements by one single
indicator bearing at the distal corner. The humerus is restricted by two fixed bearings at both
distal and proximal ends of the shaft. Based on this initial situation equilibrium has been reached
in iterative steps (for details see supplement).
Compared to the first FE model, the forces in the present steps were subdivided and
placed along the predicted insertion sites at the scapulocoracoid to obtain a more distributed
compressive stress pattern and to minimize stress peaks, which occur especially at insertions of
high forces, while the absolute force values are kept constant. Calculation of the initial situation
resulted in an increase of reaction force at the z-directed bearing at the distal edge of the
scapulocoracoid, thus indicating strong forward translation, which results into imbalance of the
trunk and the scapulocoracoid. Although forces and rotational moments were calculated
precisely, the trunk was entraining the scapulocoraocid along its direction. Friction, even with a
lowered friction coefficient of about 0.1 is beyond tolerable limits, because of the large number of
contact elements between the trunk and scapulocoracoid (for details see Fig. 75, Table 14 in the
supplement).
To avoid misleading results from uncertain reactions between the trunk and
scapulocoracoid, the trunk was removed, and the scapulocoracoid along with the humerus and
78
the antebrachium were calculated separately. The focus in this investigation is on stress
distributions within the scapulocoracoid element. Areas of the trunk adjacent to the
scapulocoracoid preventing medially directed movement were replaced by adequate bearings to
fulfil the function. Assumed regions of main force transmission are defined at the predicted
coracosternal joint and the acromial region, where the claviculae presumably contact the scapula
and where forces can be transmitted. Further applied bearing should serve as indicator bearing
and to provide stability until equilibrium was reached. Forces directed to the trunk maintain their
line of action positioned towards the trunk. After calculation a comparison between both models
reveals a reduction of translational movements at the scapulocoracoid and deformation within
the model calculated without the trunk (for details see supplement, Fig. 76). Therefore,
equilibrium was calculated under the described conditions. The final results include medial and
caudal views in addition to the lateral and medial view to show the additional information. All
steps leading to the static equilibrium can be obtained from the supplement (Table 15).
4.4.2.1 Static equilibrium
Static equilibrium in the shoulder girdle system has been reached after a number of
iterative steps (Fig. 47). The bearings, which were required in the initial situation, are removed
and the shoulder girdle system is held in position only by forces, which refer to shoulder girdle
muscles. First, equilibrium was maintained between the trunk and the scapulocoracoid by
muscles running between the two elements. Indicator bearings at the scapulocoracoid were
removed. The humerus then was still restricted by fixed bearings. In the following steps forces
were applied by mechanical requirements to reach equilibrium not only between the trunk and
the scapulocoracoid, but also within the glenoid joint. The calculations of the single iterative
steps, the applied force values and reaction forces in the bearings, can be obtained from the
supplement.
In the final calculation all indicator bearing were removed, leaving those bearings, which
provide stability in accordance to the anatomy in the sauropod shoulder girdle (Fig. 47). Therefore
medially directed bearings (x-axial) at the scapulocoracoid are still present at the assumed
sternocoracoid joint and the acromion, indicating the presence of the clavicle. The sum of all
reaction forces, which appear at the sternocoracoid joint are high, approximately 4492 N, thus
indicating strong medially directed force transmission. The reaction force, which is detected at the
acromion, is less pronounced (1171 N), but still indicates medially directed force transmission at
this location (for a complete list of reaction forces see supplement). The humerus is free from
restrictions and therefore able to rotate within the glenoid and the elbow joint. As the resultant
79
of all forces, which act on the shoulder girdle system, runs through the pivot of the glenoid, the
shaft of the humerus and the pivot of the elbow joint, rotational moments have been decreased,
thus leading to a balanced system. The antebrachium is still restricted by fixed bearings. The
reaction forces within the bearings indicate strong backwards and sidewards translational
movements. These reaction forces can only be countered by the application of the complete limb
musculature and are not considered in this study. The sum of all y-dircted forces (18387 N) is
approximately the value of body weight (18956 N), which has been initially applied to the system.
The difference between these two values is not significant and is referred to minor errors during
calculation.
The overall stress distribution of compressive stresses in the bauraum is according to the
original shape of the scapulocoracoid (Fig. 47). Only the slight curvature at the caudal border of
the scapula cannot be synthesised. The scapulocoracoid and the humerus are mainly subjected to
compression stresses. Concurrent bending within the structures was reduced. The deformation
after calculation shows the relative movements of the shoulder girdle elements, but still is in
physiological limits (Fig. 47 ). The maximum value of compressive stresses is -24.7 MPa. The
contour legend, which indicates the compressive stress values as different colours, was set
throughout the calculation process to show a range of 0 -0.9 MPa. This was required in order to
detect even low compressive stresses during the calculation. In the final calculation the numeric
spread of the contour legend ranges from 0 to-1.8 MPa. The visible spectrum (orange-blue) now
reflects only those areas, which provide basis for the following reduction steps. In those areas,
which resemble the shape of the original scapulocoracoid, the value of compressive stresses is still
lower than the precluded physiological stress value in compact bone (-2 to -20 MPa). By the
reduction of the bauraum these values are then increased to a physiological value (Fig. 47).
80
Figure 47 Final calculation of the 3-D FE-model of Diplodocus longus- Outlines of the FE-model with all applied forces in
lateral (A) and frontal (B) view. Distribution of compressive stresses in: lateral view (C), frontal view (D), medial view (E)
and top view (F). The system is kept in equilibrium only by means of forces, which can be related to shoulder girdle
muscles. All indicator bearings at the scapulocoracoid and the humerus are removed. Numeric spread of compressive
stress values in the contour legend range from
0 MPa to -1.8 MPa. Stresses beyond the spectrum (grey plotted
areas) are reduced. The desired stress pattern (orange-blue areas) is visible.
The complete FE-model of Diplodocus longus, which includes the trunk element, is
recalculated under the same boundary conditions as presented in the final calculation step (Fig.
81
48). Again, equilibrium in the shoulder girdle system was accomplished by means of forces, which
refer to the shoulder girdle musculature (Fig. 49). The completion of the model shows, that the
calculated muscles of the shoulder girdle not only maintain stability in the scapulocoracoid and
the humerus, but also in the trunk. No vertically directed bearings are required to keep the trunk
in equilibrium. Therefore the former assumed forces are confirmed. Total value of compressive
stress
is slightly increased. A reduction of compressive stresses and decrease of force
transmission can be observed at the acromial region, which is assumed to be the contact area
between the clavicles and the scapulocoracoid (Fig. 48). Nevertheless, in comparison to the
former calculation, compressive stress distribution shows only little modifications. Therefore, the
boundary conditions, which led to static equilibrium and distribution of compressive stresses, can
be confirmed. The applied forces in the FE model are related to muscles of the shoulder girdle
system and will be described according to their function.
Figure 48 Stress distribution in the complete FE model of Diplodocus longus including the trunk element. Calculation is
accomplished under the same boundary conditions as in the final calculation (Fig. 47)
82
Figure 49 FE-model with all muscles of the shoulder girdle, keeping the shouldergirdle -system in static equilibrium. Lateral view (A), frontal view (B). Muscles connecting the scapulocoracoid
and trunk are red; muscles running from the trunk to the humerus are dark blue and muscle groups originating from the scapulocoracoid and inserting at the limb are indicated as green lines.
83
4.4.2.2 Synthesis of the scapulocoracoid in Diplodocus longus
For reduction of the bauraum to the designated shape according to the diplocoid
scapulocoracoid, the resulting compressive stresses are accumulated in iterative steps.
Compressive stress values at this point are non-physiological, but their distribution corresponds to
the original diplodocoid scapulocoracoid. Compressive stresses are concentrated within the
region of the coracoid, showing transmission of stresses to the sternal region. Highest stresses can
be viewed at the insertion site of the humerus at the glenoid joint. The internal compressive
stress flow is undisturbed through the whole element, showing a reduction of material along the
scapula. The reduction is based on the results of the single calculated model, excluding the trunk
element to prevent any friction related deformations and modifications. First, the three
dimensional volume of the bauraum is dissected in antero-posterior direction at eight locations
(Fig. 50). Each section shows an area with more or less stressed regions. The areas with a certain
amount of compressive stresses are selected (Fig. 50), retained and then spliced to form a new,
reduced, bauraum. Subsequently, after remesh of the reduced bauraum, the model is
recalculated under the predefined boundary conditions according to the final calculation step. By
this procedure, the compressive stresses are accumulated in the reduced volume and increase
until they reach a physiological stress value.
Figure 50 Frontal cross sections at eight different positions through the bauraum after calculation (1-8; A). Selected
areas showing a certain amount of compressive stresses, which are retained to be spliced to form the new reduced
three dimensional volume (B).
84
The reduction of the original bauraum of the scapulocoracoid was accomplished in three
consecutive steps. The position and orientation of the required forces are equal to the condition
in the final calculation step. Insertion and origin of the forces at the bauraum are maintained
throughout the reduction. In each of the steps a decrease of the structure and an increase of
compressive stresses according to the shape of the scapulocoracoid can be obsereved (Fig. 51).
The acromial region of the scapula became visible after the first reduction step. The numeric
spread of the legends is increased stepwise according to the increased compressive stress values.
Compressive stresses in the aimed regions before the first reduction range from -0.4 to -1.8 MPa.
After the first and second reduction the stresses range from -0.9 to -2.7 MPa and -1.5 to -4.5
respectively. In the recalculation of the third and last reduction step the structure shows
compressive stresses within a range of -2 and -9 MPa, which is in the predicted range of
physiological compressive stress value of bone. In order to account for an assumed impact of
acceleration during locomotion the reduced FE model was recalculated with forces, which are
scaled up according to two times earth accelaration. The stresses are then within a range of -3 to 20 MPa, which is in accordance to the predicted compressive stress loading (Fig. 51). The
maximum value of compressive stresses in the last reduction step is -141 MPa and is located at a
small area at the most posterior border of the glenoid facet. All three reduction steps, which
include frontal cross sections of the first and second reduction, can be obtained from the
supplement.
Figure 51 Reduction of the orignal bauraum in three consecutive steps. The range of compressive stress values is shown
in the contour legends. The numeric spread of the legends is increased stepwise according to the increased compressive
stress values. The reduction of the structure is visible, even in the first reduction, in which the acromial region of the
scapula became visible.
After reduction of the bauraum the final synthesis was smoothed by a virtual surface
overlying the original mesh. By this procedure the great number of edges and nodes are reduced,
emphasizing the main features of the synthesis structure (Fig. 52). Comparing the synthesized
85
finite-element model to the original fossil the main characteristics of the scapulocoracoid in
Diplodocus longus can be described (Fig.53). The fossil consists of a long slender scapula with a
slight impression at the anterior half of the distal end. The scapula thickens disto-proximally, and
is divided in its length by a bony ridge running from the first to the last third (Fig. 53a). The
described features can again be viewed in the synthesized model (Fig. 52). This holds true
especially for the overall shape; the elongated scapula, the acromial region and the relative
thickness of the elements (Fig. 52). Further the bony ridge and the impressions on the lateral side
can be recognized (Fig. 52 A,E). The scapular body is characterized by the presence of a prominent
acromial region, which forms a deep incisure between the acromion and the scapula (Fig. 52,53).
The acromial region exhibits a deep impression along two thirds of its structure (Fig. 52 A). In the
fossil the glenoid joint is formed by parts of the scapula and the coracoid, and is the most
compact part of the scapulocoracoid (Fig.53 C). The shape of the acromion and the incisura, as
well as the distribution of bony material is similar in the synthesis (Fig. 53). The most compact
region again can be viewed above the glenoid joint (Fig. 52 A,C,D). The lateral impression at the
acromion is not as pronounced in the FE-model, nevertheless it is marked by a slight lateral
groove at this location (Fig. 52 F). The ossified epiphysis between the scapula and the coracoid is
still visible in the fossil. The caudal border of the scapula is slightly convex along the whole length
(Fig. 53 A). These two features cannot be synthesized in the FE-model (Fig. 52 A).
The scapulocoracoid in the original is medially concave in frontal view. Both ends of the
scapulocoracoid, the distal end of the scapula and the medial part of the coracoid bend towards
the trunk (Fig. 53 B). The concavity is present in the FE-model as well (Fig. 52 B). The acromion
forms a thin plate, which is thickened caudal-ward towards the scapula. A bony ridge running
along the acromion to the glenoid joint marks the borderline between both parts of the scapula
(Fig. 53 A,B). The relative thickness of the acromion and the scapula can be viewed in the FEmodel, whereby the formation of the acromial ridge is less pronounced (Fig. 52 B). The coracoid
plate of the fossil is positioned fronto-ventrally to the aperture of the ribcage, its medial border
shows a cranio-caudad inclination (Fig. 53 B). Considering the coracoid as a rectangular element,
its frontal border has the greatest thickness, with a strong rugose structured rim. The rugosities
are present along the whole frontal length, but are most visible at the first half (Fig. 53 B,C). All of
these features were successfully synthesized in the FE-model. Similar to the original, the frontal
border of the coracoid is most thickened, while the overall coracoid part of the model has a
smooth and rounded appearance (Fig.52 B,C,F).
86
The medial view of the scapulocoracoid of Diplodocus exhibits a more or less smooth
surface, which is laterally convex corresponding to the concave bending of the scapulocoracoid.
The most prominent impressions of the structure are visible along the midline of the coracoid
plate and in the middle part of the scapular body (Fig. 53 C). These shallow impressions, on the
coracoid, the scapular body and the distal end of the scapula again are visible in the synthesis (Fig.
52 F).
Figure 52 Smoothed scapulocoracoid bauraum of the final reduction step in different views. A: Lateral view. Arrows
point at the long slender scapula, the prominent acromial region and the deep incisura in between them. B: In frontal
view the scapulocoracoid is flexed medially concave. Arrows point at the rounded coracoid, a thin plate of the aromion,
and the thickened anterior side of the scapula. C: Medial view: The most massive bony regions are located where the
scapula and coracoid are forming the glenoid joint. Another arrow points at the most thickened part of the coracoid. D:
Back view: Medial concavity of the scapulocoracoid is visible. Arrows point at the thin distal end of the scapula thickens
along its length to the glenoid joint. E: Lateral oblique view: Arrows point to a ridge running along the last and second
third of the scapula. Further arrow indicates a slight impression at the most anterior, distal end of the scapula and
another shallow groove at the acromial part of the scapula. F: Medial oblique view: Arrows indicate impressions at the
distal end of the scapula, superior to the glenoid joint and along the coracoid..
87
Figure 53 .Charateristic features of the scapulocoracoid of Diplodocus longus (SMF) in lateral (A), frontal (B) and medial
view (C). Lateral view (A): Long and slender scapula (I.), with a slight impression at the distal end (II.) and a bony ridge
along the whole length of the scapula (III:). A prominent acromial region (IV) and a deep fossa on the lateral side of the
extended part of the scapula (V) and the acromial ridge (VI). thickened bony structure above the glenoid joint (VII).
Frontal view (B): The scapulocoraoid exhibits a medially convex shape. Coracoid is thickened and show a rugose
structure at its medial side pointing towards the trunk (I.); Acromial ridge in frontal view (II.); The scapula is very thin,
which is most pronounced at its cranial border (III.). The coracoid is positioned nearly in the same plane as the scapula
(IV.) Medial side (C) Rugosities at the medial border of the coracoid (I.) and impressions at the coracoid blade (II.). Deep
fossa at the medial side of the scapula (IV.). The most compact bony regions can be observed at the scapular part of the
glenoid facet (V.).
88
4.4.3 Reconstruction of the shoulder girdle
The reconstruction of the shoulder girdle of Diplodocus longus is based on mechanical
pre-conditions for transmission of body weight from the trunk to the forelimb in one-limb stance,
which are confirmed by means of the successful FESS of the scapulocoracoid. The reconstruction
affects the overall position of the scapulocoracoid and the sternal plates as well as the position of
the humerus (Fig. 54). The arrangement of the shoulder girdle elements allow for a closed circle
of forces, kept in equilibrium by the required shoulder girdle muscles.
4.4.3.1 Skeletal elements of the shoulder girdle
The scapulocoracoid is positioned parallel to the ribcage covering the first thoracic ribs in
lateral view. The lateral inclination of the scapula now exhibits an angle of about 65° between the
posterior side of the scapula and the horizontal plane, compared to 50° in the original mount (Fig.
54 A). The distal expansion of the scapula is slightly inclined to the dorsal margin of the vertebrate
spines, leaving room for a cartilaginous suprascapula (Fig. 54 A). The coracoid plates now lie
medio-frontally in the aperture of the ribcage. In contrast to the original mount they move
towards each other, which results into a narrowed distance between the two halfs of the
shoulder girdle (Fig. 53 B). Ample room is left for the adjacent sternal plates, which are therefore
placed medio-posteriorily to the coracoids on the ventral side of the trunk (Fig. 54 B).
The long axes of the paired sternal plates lie medially within the longitudinal axis of the
body, firmly united by cartilage, whereas the whole plates are supposedly embedded into a
cartilaginous plate, which expands at least over one half of the ventral side of the trunk (Fig. 55).
The coracoid contacts the sternal plate at its short anterior border. Both elements together form
the coracosternal joint (Fig. 55). This joint is assumed to exhibit a cartilaginous rim and tongue
structure, where the coracoid rim moves within the sternal tongue, thus enabling relative
movements between the shoulder girdle and sternum (Fig. 55). The lateral aspect of the
triangular sternal element and the caudally expanding cartilaginous plate then provide room for
connection to the distal ends of the thoracic ribs (Fig. 55 ). Following the reconstruction of the
shoulder girdle the glenoid fossa faces slightly posteriorily (Fig. 53A). The humerus therefore
articulates to the glenoid joint at a moderate angle of about 20° in lateral view, whereas the angle
in frontal view is assumed to be not less than 15°. The maximal expansion of the proximal
articulation facet of the humerus here is supposed to lie in the transverse plane of the glenoid
facet (Fig. 53 A).
89
The position of the clavicles remains somewhat ambiguous. In the final calculation step
without the trunk, one bearing was placed at the acromial region related to the contact area of
the clavicle to the scapula. The reaction forces in this bearing was low compared to the
sternocoracoid joint, transmitting the major amount of forces. The final calculation including the
trunk shows a reduction of forces at this point, while showing constant conditions in the other
regions. It is assumed, that the clavicles were positioned at the acromial regions, while their true
function remains unknown. Thus, they were not included in the reconstruction.
Figure 54 Shoulder girdle reconstruction of Diplodocus longus, in (A) lateral and (B) frontal view. Both situations before
(black outlines) and after (grey) are shown. In lateral view the shoulder girdle is positioned more cranially covering the
first rib. Inclination of the scapula is 60° after reconstruction, compared to 50° before. The humerus is rotated 45°
around its y-axis; the maximal expansion of the humeral facet lies in the transverse plane of the glenoid joint. Position
of the humerus is inclined 20° to the perpendicular (A). In frontal view the shoulder girdle is placed parallel to the
ribcage; both coracoids are shifted medially, and contact the sternal plates.
90
Figure 55 Reconstructed shoulder girdle in ventral view (A). Coracoid plates contact the sternal plates ventro-medially.
Sternal plates are embedded in a cartilaginous plate, where the distal ends of the ribs insert posterio-laterally. The
coracosternal joint is built by cartilaginous parts of the coracoid and the sternal plates (A). The coracoid part is the
tongue, while the sternal part builds the sliding rim or groove providing contact for the coracoid tongue (B).
Circle of forces in one sauropod shoulder girdle is shown in frontal view (Fig. 56).
Compressive forces in the shoulder girdle system require bony elements, while tensional stresses
can be sustained by muscles or tendons. In a static system the weight forces of a body are
transmitted from the trunk through the shoulder girdle elements to the supporting limb and back,
while the ground reaction force acts in the opposite direction. For sauropods the circle of forces is
reconstructed as follows (Fig. 56). Weight force of the body is transmitted by means of m.
serratus superficialis from the trunk to the scapulocoracoid and via the glenoid joint to the
supporting limb. The presumably abducted position of the limb, relative to the glenoid joint, leads
to medially directed components of forces. They run from the scapulocoracoid via the sternal
plates and over the ventral ribs back to the trunk, thus closing the circle of forces. In an abducted
position of the humerus m. pectoralis will sustain the tensional forces, while m. deltoideus
(clavicularis and scapularis) serve as the abductors of the forelimb. Further, in one limb stance
tensional forces occur at the distal end of the scapula, which are taken by means of m.
rhomboideus (Fig. 56).
91
Figure 56 Circle of forces in a sauropod shoulder girdle in frontal view. Compressive forces (white arrows) are taken by
bony elements (black); tension (black arrows) is sustained by muscles (white). Weight force of the body (Fg) is
transmitted by m. serratus from the trunk to the scapula, via the glenoid joint to the humerus. A medial component
runs from the scapula to the coracoid via the sternal plates back to the trunk, closing the circle of forces. Ground
reaction force (Fgr) therefore acts from the opposite direction. In an abducted position of the humerus (dashed
humerus) m. pectoralis sustains the occurring tensile forces, while m. deltoideus serves as the abductor of the limb.
4.4.3.2 Muscles of the shoulder girdle
The shoulder girdle muscles are reconstructed according to their mechanical function to
keep the shoulder girdle in equilibrium during one limb stance (Fig. 57). Muscles are described as
vectorial forces contributing to static equilibrium with regard to their line of action, whether they
are single or subdivided, and their position in the model. Each force consists of components in the
x-y- and z-axis, which are described separately to emphasize their contribution to movement.
From this information anatomical position, origin and insertion site, as well as the function of the
corresponding muscle is reconstructed.
Reconstruction and determination of function was possible for all muscles running from
the scapulocoracoid to the trunk, humerus, and antebrachium. Furthermore, m. pectoralis major,
running between the trunk and humerus, contributes to static equilibrium. M. latissimus dorsi and
m. teres major are the only muscles associated with the shoulder girdle system, which are
excluded from calculation, as they supposedly play a major role during locomotion, but do not
contribute to static equilibrium during one limb stance (Fig.57). Insertion sites at the model
92
compared to reconstruction can differ in some muscles. To simulate a laminar insertion at the
bone, as it can be expected for m. serratus superior, m. levator scapulae or m. trapezius, the
vectorial forces are shifted following their line of action. For example in m. serratus superficialis
the force vectors are placed more cranially than it is described for reconstruction. This results into
a homogenous distribution of stresses, the mechanical function is not affected.
M. serratus superficialis- The total force of m. serratus superficialis is subdivided into a
number of single vectors distributed evenly along the cranial border of the dorsal end of the
scapula. In lateral view the forces are inclined posteriorily, inserting at the ventro-lateral side of
the trunk. They rotate the bauraum in caudal direction. In frontal view forces are placed in the
middle of the bauraum, pointing sidewards, inclination angles increase from the most cranial
vectors to the vectors placed caudally. The forces draw the distal part of the scapula away from
the trunk. M. serratus superficialis produces high y- and z-directional forces. Forces in x-direction
are correspondingly lower. The muscle therefore originates from the fifth-eighth thoracic rib,
running to the caudal border of the distal part of the scapula. M. serratus superficialis transmits
body weight from the anterior part of the trunk to the scapula. As the main portion of the body
weight is transmitted by this muscle, the required forces had to be correspondingly high. The
determined insertion areas at both, the scapula and the ribs, provide ample space for an effective
muscle diameter.
M. serratus profundus- The total force is divided, with each single vector positioned at the
distal end of the scapulocoracoid. The most cranial vector points to the frontal part of the trunk,
the second is more or less perpendicular. The line of action in the third vector runs through the
glenoid joint, while the most caudal vector is nearly parallel to the fibers of m. serratus
superficialis. Altogether this muscle has a fan-shaped appearance. In frontal view the vectors
point medially from the scapulocoracoid to the trunk. In sum all vectors produce zero rotational
moments at the glenoid joint, while the main direction of forces lies in the y-axis. M. serratus
profundus is positioned medially between the distal end of the scapula and the most anterior part
of the trunk. It originates from the last cervical to the third or fourth thoracic rib. Its insertion is
reconstructed at the medial side of the distal end at the scapula, ventrally to m. rhomboideus. M.
serratus profundus supports the transmission of body weight and does not contribute significantly
to rotation of the scapulocoracoid. As its fibers are slightly directed medially it prevents the
scapula from sliding aside. The supportive nature of this muscle is most visible in its function to
absorb the negative y-directional forces produced by muscles acting in cranial rotation of the
scapula (i.e. m. trapezius and m. levator scapulae), while counteracting the caudal rotation of m.
93
serratus superficialis. Following this m. serratus profundus can be termed as a set-screw in this
part of the shoulder-girdle system, keeping the sum of forces in equilibrium.
M. trapezius- The total amount of force in this muscle is subdivided into a number of
single vectors, which are positioned at the last third of the most caudal border of the
scapulocoracoid. Corresponding vectors are placed at the dorsal midline of the trunk, anterior to
the scapulocoracoid. All vectors are parallel to each other, their line of action is widely separated
from the pivot of the glenoid joint. An increased lever arm, compared to m. serratus superficialis,
allows for a decrease of forces, while countering the rotational moments at the scapulocoracoid,
which are induced by the main weight bearing muscle. At the same time vectors of m. trapezius
have high cranially directed z-axial components, while y-and x-axial components are low. M.
trapezius originates from the dorsal midline of the trunk at the level of the last cervical and first
thoracic vertebrae. Its cranial extension could be even more cranially, but this would not affect
the results as forces. Its insertion is placed at the last third of the most cranial border of the
scapula, dorsally to m. levator scapulae. M. trapezius is a cranial rotator of the scapula and
together with m. levator scapulae serves to counteract the caudal rotation induced by m. serratus
superficialis.
M. levator scapulae- This muscle consists of a superior (m. levator scapulae superior) and
an inferior part (m. levator scapulae), each of them subdivided into single vectors. The superior
part is placed in the middle of the last third of the scapula; the inferior part is positioned below at
the middle of the second third. In both parts corresponding vectors are placed at the most cranial
end of the trunk. The vectors are parallel to each other, while the superior part of m. levator
scapulae has a greater lever arm to the pivot of the joint than the inferior part. Muscle forces in
the superior part therefore can be lowered to produce the same cranially directed rotational
moments at the scapulocoracoid. As in m. trapezius high z-axial components of forces in both
parts of m. levator scapulae translate the scapula cranially. The y-and x-axial components are
comparably low. M. levator scapulae originates from the cranial border of the second third of the
scapula to the acromial region. Its insertion is at the ventral side of the neck, presumably at the
last four or five cervical vertebrae and/or adjacent cervical ribs. M. levator scapulae serves as a
cranial rotator of the scapulocoracoid, supporting m. trapezius in counteracting caudal rotation.
The wide distribution of this muscle along the cranial border of the scapula is reasonable as the
elongated scapula acts as a long lever arm and provides ample space for m. levator scapulae in
order to exert the required moments to the scapula.
94
M. rhomboideus- Vectors run parallel between the most distal end of the scapulocoracoid
and the dorsal margin of the trunk. The vectors are mainly directed in the x,y-plane and consist of
high x-axial forces. They prevent the most distal part of the scapula from sliding laterally. M.
rhomboideus originates from the dorsal margin of the vertebral column. Its parallelly oriented
fibers insert at the distal margin of the scapula and an assumed cartilaginous suprascapula. The
cartilaginous part cannot be reconstructed beyond doubt. M. rhomboideus draws the distal end of
the scapula towards the trunk.
M. costo-coracoideus-The total force of this muscle is subdivided and runs parallel from
the caudal border of the coracoid to the middle of the trunk, passing the glenoid joint ventrally.
The forces exert mainly z-axial components, while x,y-components are low. It is rather close to the
glenoid joint, and therefore has a short lever arm. Both cranial rotators of the scapula (m.
trapezius, m. levator scapulae) produce high cranially directed z-axial forces. To establish static
equilibrium not only rotational moments have to be balanced, but also the sum of forces has to
be equalized. This is established by the cranially directed z-axial components of m. costocoracoideus drawing the bauraum caudally. It should be noted that without these forces no static
equilibrium was reached in any calculation. M. costo-coracoideus originates from the ventral part
of the first thoracic ribs, and inserts dorsally at the caudal border of the coracoid plate. It
somewhat contributes to cranial rotation of the scapulocoracoid, but its major function is to draw
the scapulocoracoid caudally relative to the sternum.
M. costo-coracoideus pars profundus- The components of m. costo-coracoideus profundus
are located in the middle of the coracoid and the ventral midline of the trunk. They are mainly
directed medially, with a certain amount of z-axial forces. Components in the y-axis are low. The
x-directed components draw the coracoid part of the scapulocoracoid towards the sternal part of
the trunk. M. costo-coracoideus pars profundus originates from the sternum and in part from the
first thoracic ribs and inserts at the dorsal side of the coracoid plate, anterior to m. costocoracoideus. This muscle serves as the main stabilizer of the coracosternal joint, as it prevents
luxation of the joint by pressing the coracoid tongue towards the sternal rim.
M. supracoracoideus- The total force of this muscle is subdivided and runs between the
ventral side of the coracoid part of the scapulocoracoid and the anterior side of the humerus.
Forces are directed mainly in the z-and x-axis, drawing the scapulocoracoid posteriorily and the
humerus medially. Forces rotate the scapulocoracoid cranially, and at the same time exert
posteriorily z-directed components, which are necessary to compensate for the large anteriorly
directed z-components of m. levator scapulae and m.trapezius. The presence of this muscle allows
95
for reduction of forces in m. levator scapulae and m. trapezius as it serves as a support in cranial
rotation. M. supracoracoideus originates from the anterior part of the ventral side of the coracoid
plate and part of the most proximal border of the scapula, below the insertion of m. levator
scapulae. Its insertion is on the lateral aspect of the humerus, near the glenoid joint. In one limb
stance it serves as a stabilizer of the glenoid joint, playing a major role in establishing equilibrium
between the scapulocoracoid and the humerus.
M. coraco-brachialis brevis- The forces of m. coraco-brachialis brevis are placed at the
coracoid part of the bauraum in front of the pivot of the glenoid joint and at the anterior side of
the humerus. As in m. supracoracoideus the components are mainly directed in the z,x-axis,
drawing the scapulocoracoid posteriorily and preventing the humerus from sliding laterally.
Functionally these forces are similar, although its lever arm is longer, thus exerting higher
rotational moments to the scapulocoracoid. M. coraco-brachialis brevis is reconstructed to
originate from the ventral side of the coracoid plate above the insertion site of m. biceps brachii
and posterior to m. supracoracoideus. Its insertion site is on the lateral side of the humerus below
m. supracoracoideus. As the function is comparable to m. supracoracoideus, both muscles can be
viewed as agonists during joint stabilization during one limb stance.
M. biceps brachii- This muscle is positioned as one single force at the most medial part of
the coracoid bauraum and the anterior part of the antebrachium. In lateral view this force runs at
a far distance from the glenoid joint, but close to the pivot of the elbow joint. This results into a
large rotational moment exerted on the scapulocoracoid, while expansion of forces can be kept
low. At the same time the elbow joint is little affected. In frontal view distance to glenoid joint
enables high rotational moments just as much as in lateral view, resulting into lateral movement
of the scapulocoracoid and concurrent medial movements of the antebrachium. The laterally
directed forces on the bauraum are balanced by m. costo-coracoideus pars profundus.
Components are mainly y-, and x-axial, with a low portion directed posteriorily (z-axial). M. biceps
brachii is reconstructed to originate from the ventro-lateral surface of the coracoid plate near the
coracosternal joint. It presumably inserts on the anterior part of the antebrachium, whether on
the ulna or radius cannot be determined beyond doubt. Along with m. coraco-brachialis and m.
supracoracoideus it is an important stabilizer of the glenoid joint and contributes to minimize
anterior movements of the scapulocoracoid in one limb stance.
M. triceps caput coracoideum- It is positioned as one single force at the coracoid just in
front of the pivot of the glenoid joint and at the middle of the antebrachium near the pivot of the
elbow joint. In frontal view its lever arm to the glenoid joint is somewhat longer than at the elbow
96
joint, which results into moderate cranial rotation of the bauraum, while the antebrachium is not
affected. Force components in the y-axis dominate, x-and z-directed components are low. M.
triceps caput coracoideum originates from the medial side of the coracoid in front of the glenoid
joint. It inserts on the posterior side of the ulnar element of the antebrachium. In an extend
position of the humerus, as it is assumed in this reconstruction, m. triceps caput coracoideum
contributes to moderate cranial rotation and backward drawing of the scapulocoracoid.
M. triceps caput scapulare- This muscle originates from the lateral margin of the most
proximal aspect of the scapula posterior to the glenoid joint and inserts along with m. triceps
caput coracoideum behind the elbow joint in the middle of the antebrachium. In lateral view the
distance to the glenoid joint is farer than to the elbow joint. The resulting force rotates the
bauraum in caudal direction and contributes to posterior movements of the scapulocoracoid. In
frontal view line of action runs close to the pivots of both joints, thus exerting only minor
rotational moments in this plane. M. triceps caput scapulare takes its origin from the caudal
border of the scapula proximal to the glenoid joint. The muscle inserts along with m. triceps caput
coracoideum posteriorily at the ulna. M. triceps caput scapulare rotates the scapulocoracoid in
caudal direction, while drawing it slightly backwards. Both parts of the triceps complex originate
from either the posterior or anterior side of the glenoid joint. In an active muscle system they can
act as mechanical antagonists, while balancing either forward or backward rotation, as needed.
M. pectoralis- Total force of m. pectoralis is represented as one single vector. It is
positioned at the posterior side of the humerus and near the midline at the last third of the trunk.
Line of action lies mainly in the y,z-plane and exerts high z- and x-components of force, which
results in posterio-medial movements of the humerus. Y-axial components of force are low.
Forces applied to the anterior side of the humerus (m. supracoracoideus and m. coracobrachialis
brevis) pull the humerus in cranial direction, whereas m. pectoralis counteracts these movements.
M. pectoralis originates from the whole length of the ventral side of the sternal plates, at least to
the level of the fifth thoracic rib. Its overall extension along the cartilaginous part of the sternum
cannot be finally determined, due to lack of reliable information but it presumably does not to
extend beyond the level of the fifth or sixth thoracic rib.
M. pectoralis pars profundus- The establishment of static equilibrium requires a certain
amount of caudally directed force components in the z-axis acting on the bauraum of the
scapulocoracoid. This is due to the large amount of caudally directed z-components exhibited by
m. trapezius and m. levator scapulae, which counteract the caudal rotation of the scapula. A
certain portion but not all of these caudally directed z-components can be taken by m. costo97
coracoideus. M. pectoralis pars profundus was applied to produce the required z-directed
components. Its line of action runs from the middle part of the trunk to the bauraum close to the
glenoid joint. Main components are in z-and y-direction, whereas the x-directed component of
force is low. As the force is applied close to the pivot of the glenoid joint, rotational moments
exerted on the bauraum are low. M. pectoralis pars profundus originates from the medial side of
the scapula, most profoundly at the glenoid joint and inserts at the second and third thoracic rib.
Its main function in one limb stance is to pull the scapulocoracoid caudally. Further, it supports
force transmission from the trunk to the scapulocoracoid, but only to a minor degree.
M. deltoideus clavicularis- In frontal view the line of action in m. deltoideus clavicularis
runs from the lateral side of the bauraum to the lateral side of the humerus. The total force is
subdivided into a great number of single vectors applied to the scapulocoracoid, directed laterally,
as they do at the humerus. The resulting angles between these pairs of forces are required to
simulate a muscle running over the glenoid joint. By this arrangement the x-directed components
of the muscle are higher, as if the pairs of forces were directed perpendicularly to each other. The
sum of forces shows a high amount of medially directed forces acting on the bauraum. In order to
reach equilibrium, these forces have to be countered by means of m. deltoideus clavicularis in the
described modality. In lateral view line of action runs close to pivot of the glenoid joint, thus
preventing lateral rotation of the scapulocoracoid bauraum. M. deltoideus clavicularis originates
from the cranial edge of the scapula, in the region of the acromion and nearly the whole lateral
aspect of the scapulocoracoid above the glenoid joint. In the fossil a deep impression is present at
this location, providing ample space for the muscle. It inserts at the deltopectoral crest laterally at
the humerus, along with m. deltoideus scapularis and m. pectoralis. Its serves as a stabilizer of the
glenoid joint, as it prevents medial translation of the scapula. Further, following its line of action
at the humerus, it serves as a lateral abductor of the humerus.
M. deltoideus scapularis- The force of m. deltoideus scapularis is subdivided into a number
of single vectors, which run from the middle of the scapula to the lateral aspect of the humerus.
The arrangement is similar to m. deltoideus coracoideus. The forces are located superior to m.
deltoideus coracoideus, which in comparison results into a longer lever arm and laterally directed
rotational moments in frontal view. In lateral view rotational moments are lower. The forces are
mainly directed in the y-axis. A certain amount of x- directed components translate the
scapulocoracoid laterally. Little movement is present in the z-plane. M. deltoideus scapularis
originates from the middle of the scapula. Cranially, it is positioned superior to m. deltoideus
clavicularis. It runs close to the glenoid joint and inserts at the deltopectoral crest along with m.
98
deltoideus clavicularis and m. pectoralis. It stabilizes the glenoid joint, rotates the scapula laterally
and, similar to m. deltoideus clavicularis, serves as a lateral adductor of the humerus. Its thickness
is assumed to be lower than that of m. deltoideus clavicularis. This is because the muscle is able to
exert the same amount of rotational moment with a lower amount of forces, because of the
increased lever arm acting on the pivot of the glenoid joint in frontal view.
M. subscapularis- The forces of m. subscapularis point from the medial side of the
scapulocoracoid to the medial aspect of the humerus at the level of the glenoid joint. The
arrangement of single vectors can be compared to that of m. deltoideus clavicularis and
scapularis, as it is applied to balance the occurring rotational moments at the glenoid joint, but in
opposite direction. The main components of this force lie in the y-and x-axis. The medially
directed x-components at the scapulocoracoid provide stability as they move the scapula to the
trunk, but to a lower degree than the counteracting forces exerted by the muscles of the deltoid
group. M. subscapularis originates from the medial side of the scapula and covers a large area,
which is provided by a bony impression at this location. It inserts deep at the medial side of the
glenoid joint. Rotational moments can only be viewed in frontal view, whereas muscle actions do
not affect lateral stability. This muscle is a stabilizer of the glenoid joint in one limb stance and is
required to counter forces exerted by all laterally acting muscles.
M. scapulo-humeralis- This muscle is applied as one single vector, pointing from the
lateral side of the humerus to the second third of the scapula. The force is comparably low as
laterally directed x-axial components are only required to a minor degree at this location. In
frontal view it serves to pull the scapula laterally, thus minimizing bending in the scapula at this
position. At the same time the forces exert moderate caudally directed rotational moments at the
scapulocoracoid. M. scapulo-humeralis originates from the second third of the scapula, lying
anterior to the insertion of m. serratus superficialis, below m. teres major. Its insertion is clearly
separated from m. deltoideus scapularis by a prominent bony ridge running from the most distal
end of the scapula to the proximal basis of the scapula near the glenoid joint. It contributes less to
the establishment of static equilibrium. Considering its line of action, running from the scapula to
the humerus posterior to the glenoid joint, its major function is that of a retractor of the humerus
during locomotion cycles.
M. teres major- This muscle is represented by one single force. Its line of action runs from
the posterior, middle part of the scapulocoracoid to the posterior side of the humerus. The
distance to the pivot of the glenoid joint is smaller than for m. scapulo-humeralis, therefore the
resulting rotational moments are lower. The major component of force is directed to the y-axis; x99
and z-axial components are low. M. teres major was determined not to contribute to static
equilibrium in one limb stance, therefore it is excluded from the muscular setting in the final
calculation.M. teres major presumably originates either from the caudal border of the last third of
the scapula, lying superior to m. scapulo-humeralis, or posterior to the bony ridge of the scapula,
superior to m. deltoideus scapularis. As no calculation could be completed with this muscle
included in the system this cannot be determined beyond doubt. M. teres major has no stabilizing
function in one-limb stance, but it could in addition to m. latissimus dorsi serve in retraction of the
humerus during locomotion.
M. latissimus dorsi- It consists of one single force running from the dorsal margin of the
trunk to the posterior side of the humerus. The line of action runs at a wide distance to the pivot
of the glenoid joint, which results in high rotational moments at the humerus. The main
components are in the y-and z-axis, but a certain amount of forces drags the humerus towards
the trunk. The muscle is not present in the final calculation. M. latissimus dorsi presumably
originates from the dorsal fascia at the thoracic vertebrae and inserts at the second third of the
posterior side of the humerus. According to the long lever arm of m. latissimus dorsi at the
glenoid joint, it serves as a strong retractor during locomotion, but does not contribute to static
equilibrium in one limb stance.
Figure 57 Reconstruction of the insertion sites of the muscles to the scapulocoracoid in lateral (left) and medial view
(right). M. serratus superficialis(1), M. serratus profundus (2), M. trapezius (4), M. levator scapulae superior (5), inferior
(6), M. rhomboideus (3), M. costo-coracoideus (10), M. costo-coracoideus profundus (11), M. supracoracoideus (16), M.
coraco-brachialis brevis (19), M. biceps brachii (17), M. triceps caput coracoideum (18), M. triceps caput scapulare (8),
M. pectoralis pars profundus (9), M. deltoideus clavicularis (15), M. deltoideus scapularis (12), M. subscapularis (7), M.
scapulo-humeralis (14), M. teres major (13).
100
The cross sectional areas (CSA) of the muscles determined above is estimated by dividing
muscle force by the maximal isometric stress of lacertilian skeletal musculature [0.2 N/mm² (see
methods, Medler, 2002)] (Table 2). The largest CSA is present in m. serratus superficialis, formerly
determined as the most weight bearing muscle in one limb stance. Large CSAs are also present in
m. serratus profundus, m. trapezius, m. costo-coracoideus, m. pectoralis profundus, m.
supracoracoideus, m. coracobrachialis brevis ventralis, m. deltoideus scapularis, m. deltoideus
clavicularis and in both parts of m. levator scapulae (scapularis et clavicularis). CSAs estimated for
m. rhomboideus, m. costo-coracoideus pars profundus, m. pectoralis, m. triceps caput
coracoideum, m. biceps brachium and m. subscapularis are comparably lower (Table 2). The
smallest CSAs are found in m. rhomboideus, m. triceps caput scapulare and m. scapulo-humeralis.
CSAs are presented in cm² to give an impression of their overall dimensions and to be compared
to the skeletal elements to which they insert (Table 2). To evaluate, whether the muscles found
ample space in the skeleton or were underestimated, the dimensions of the skeletal insertion
sites for some muscles are compared to the estimated CSAs. Muscular architecture differs
between the respective muscles. Some of them presumably insert directly to the bone or at
distinct areas via a tendon. A comparison of the insertion sites with the estimated CSAs, although
only approximately, is only possible for muscles in which length and breadth of the assumed
insertion sites can be measured. If the muscle inserts via a tendon to the bone, it cannot be
correlated to the assumed insertion site, as it does not reflect muscle body dimension. Muscles
that count for the first condition are m. serratus superficialis, m. trapezius and m. levator scapulae
(superior). The insertion site of m. serratus superficialis at the scapula is assumed to be 60 cm in
length and maximally 3 cm in breath, accounting to 180 cm². The area of origin at the thoracic ribs
is assumed to extend over 100 cm in length and maximally 5 cm in breadth, accounting to 500
cm². A maximal CSA of about 880 cm² was estimated for this muscle. Assuming a mean value for
insertion length of about 80 cm², the maximal thickness of about 11 cm² in this muscle is required.
M. trapezius is assumed to insert along the scapula over a distance of 40 cm. The maximal width
at the cranial border is 3 cm. Insertion area therefore would be 120 cm². The origin of this muscle
extends over an assumed distance of approximately 80cm in length, its width cannot be
determined beyond doubt. If the latter is assumed to be maximally 5 cm², according to the
breadth of the vertebral spines, the estimated value accounts to 400 cm². The estimated CSA of
m. trapezius is 215 cm², thus the formerly determined forces would be underestimated. The
insertion sites of both parts of m. levator scapulae (superior) are assumed to be 30 cm and 40 cm
in length, and a maximal breadth of 2 cm, respectively. M. levator scapulae therefore exhibits 60
cm² at its superior part and 80 cm² at the inferior part of the muscle. The origin extends
101
presumably 120 cm along the ventral side of the neck to insert at the cervical ribs and has a
maximal breadth of about 5 cm. CSA in this case would be 600 cm². The CSAs for both muscles
together has been formerly estimated as 208 cm², which is significantly lower than the CSAs
provided at the muscle insertion sites.
Table 2 Cross sectional areas of the shoulder girdle musculature in Diplodocus longus, calculated out of maximum
muscle forces obtained from FESS.
No
Muscle
Force (N)
A (mm²)
cm²
1
2
M. serratus superficialis
17603
88015
880,15
M. serratus profundus 1
301
1505
15,05
3
M. serratus profundus 2
1895
9475
94,75
4
M. trapezius
4300
21500
215
5
M. levator scapulae
3000
15000
150
6
M. levator scapulae superior
1175
5875
58,75
7
M. rhomboideus
1000
5000
50
8
M. costo-coracoideus
4800
24000
240
9
M. costo-coracoideus pars prof.
3000
15000
150
10
M. pectoralis profundus
4500
22500
225
11
M. pectoralis
3500
17500
175
12
M. triceps caput coracoideum
1900
9500
95
13
M. triceps caput scapulare
1000
5000
50
14
M. supracoracoideus
4500
22500
225
15
M. biceps brachii
2000
10000
100
16
M. coracobrachialis brevis ventralis
4000
20000
200
17
M. deltoideus scapularis
5000
25000
250
18
M. deltoideus clavicularis
6000
30000
300
19
M. subscapularis
2000
10000
100
20
M. scapulo-humeralis
500
2500
25
a
with a specific tension like that of Iguana of about 0.2 N/mm² (Medler, 2002)
102
Chapter 5
5.1
Discussion
Form and function of the tetrapod body
The stress analysis of the 3-D FE models of solid tetrapod bodies reveals the distribution
of mechanical stresses in a tetrapod-like body by the time the body is released to gravity. Two
different positions of the forelimbs were investigated, the sprawling type representing the
primary condition observed in early tetrapods as well as in many extinct and extant reptiles; the
extended type as it is exhibited in cursorial mammals.
The 3-D FE models reveal stress patterns in regions representing the limb girdles and
trunk, which clearly depend on the mass distribution in the head, neck and trunk as well as the
position and posture of the supporting limbs. In both the reptilian and mammalian type the
compressive and tensile stresses at local regions are significantly higher during asymmetrical
stance, which emphasizes the relevance of locomotion for the demands on the supporting
musculoskeletal system.
In the FE-models the trunk is subject to bending moments which compress the dorsal and
stretch the ventral side during symmetrical support. Compression stresses spread between both
fore- and hindlimbs and the dorsal side of the trunk, thus indicating a connection to the vertebral
column at both regions. This has to be taken as the initial condition. In tetrapods, dorsal stresses
are sustained by compression resistant structures like the vertebral column. However, no
connection exists in the shoulder region of more derived tetrapods. By contrast, the mobility
between the pelvic girdle and the vertebral column is restricted. During locomotion the
movement of the body is accomplished by transmission of ground reaction forces resulting in
forward propulsion. In tetrapods acceleration is mainly accomplished by the hindlimbs, whereas
the forelimbs act in deceleration and direction changes (Hildebrand & Goslow, 2001; Alexander,
2003). In early tetrapods the shoulder region is more engaged in lifting the body off the ground
than in true locomotion. A connection to the vertebral column therefore is reasonable, as the
forelimbs are the first limbs to develop, carrying the entire body weight. In Tiktaalik, who
supposedly represents a semi-aquatic locomotor type living in shallow water, the shoulder region
shows no direct connection to the head, due to a loss of the operculum and extrascapular
103
elements, thus enabling striking head movements (Shubin et al. 2006; Downs et al. 2008).
Nevertheless, it has a medio-dorsal connection via the anocleithrum and supracleithrum (Shubin
et al. 2006; Downs et al. 2008). This can be regarded as an intermediate stage, where a certain
stiffness is maintained via a connection to the vertebral column, but an enhanced mobility of the
shoulder girdle can be established based on these preadaptations. As soon as the tetrapod
hindlimbs have fully developed, the main propulsive action is transferred caudally. An enhanced
mobility at the forelimbs and shoulder region still requires a connection to the vertebral column
which is realized by more flexible structures as muscles and ligaments (Shubin, 1995; Coates,
1996; Shubin et al. 2006; Downs et al. 2008; Boisvert, 2008).
The highest values of compressive stresses are located at the insertion sites of the limbs
at the trunk, i.e. the glenoid joints. In the sprawling type two portions of maximum compressive
stress patterns spread from the inserting limb, first, dorso-laterally along the trunk and, second,
ventrally to the midline of the trunk. In extant reptiles the scapula (lateral aspect) and the
coracoid plate along with an interclavicula connected to the sternal elements (ventral aspect) are
located at this position. In early tetrapods like Tiktaalik or Acanthostega the scapula is not as
developed as in later forms, additionally lacking the sternal elements, whose their function can be
adopted by the cleithrum (lateral aspect) and the well developed clavicle (ventral aspect) (Shubin
et al. 2006; Coates, 1996). As soon as the animal lifts one forelimb off the ground (asymmetrical
support) as in walking, the stress distribution and values are modified. At the position of the
scapula of the standing limb tension occurs dorsally, while compression increases ventrally. Thus,
the alternation of symmetrical and asymmetrical support during sprawling locomotion confirms
the relevance of a continuous compressive resistant structure medially between the limbs. This
increased stresse is significantly lower in extended than in sprawling limb models. In extended
limb position only minor compression is visible between the forelimbs both in symmetrical and
asymmetrical support. Thus, no compression resistant structures are required. Mammals possess
a sternal element, but it is reduced to a rod-like element and has no bony connection to the
shoulder girdle.
In the mammalian type compressive stresses concentrate dorso-laterally spreading from
the inserting limb to the trunk. This inclination is more pronounced in an extended limb stance
than in the sprawling model, which in fact corresponds to observations in extant cursorial
mammals and lacertilians. Compression is also visible at the ventral side of the neck region, which
is in accordance to the more ventral location of the cervical vertebrae, especially in mammals
exhibiting a long neck. In contrast, crocodiles exhibit a more inclined position of the scapula,
104
intermediate between the condition of mammals and lacertilians. Although they are mostly
engaged in aquatic locomotion, crocodiles can hold their forelimbs under the body during high
speeds, termed the high walk and are able to perform gaits similar to the gallop (Zug, 1974; Webb
& Gans, 1982; Renous, 2002). Crocodiles descend from quadrupedal terrestrial ancestors and
therefore have a derived forelimb condition, which can not be equalized to the situation in
modern lacertilians (Hutchinson & Gatesey, 2000; Benton, 2005).
Additionally, in one limb support, torsional stresses occur at the surface of the trunk,
which are higher and spread over larger areas than bending stresses in symmetrical stance.
Stresses run obliquely from the scapula of the weight-bearing limb to the contralateral support
(hindlimb). These stresses have tensile as well as compressive components. The first can be
sustained by the oblique muscles of the ventro-lateral body wall; the latter necessitate segmental
bony elements, namely the ribs (Preuschoft et al. 2007). The compressive stresses are
concentrated in the periphery, therefore leaving the centre of the trunk without stress. Thus, the
body cavity is less important to absorb mechanical stresses resulting from locomotion and
therefore can provide space for the intestines. Torsional stresses increase in sprawled and
extended limb positions, but are always concentrated in the periphery of the trunk. The tensile
stresses have been related to appropriate muscular structures, as the m. rectus abdominis, which
sustains tensile stresses along the ventral contour in animals of both sprawling and extended
locomotor type. Correspondingly, on the dorsal side of the anterior trunk and posterior neck
tensile stresses occur, which again can be sustained by neck and shoulder muscles connecting the
head to the trunk and pectoral girdle. The corresponding muscles in extant tetrapods can be
determined, although homologies differ between taxa. This will not be discussed in detail, as the
aim of the study is to relate the structures due to their mechanical requirements, not their
phylogenetic relationships. Shoulder girdle musculature in early tetrapods had not been described
in detail by now. In Tiktaalik the adductor of the pectoral fin presumably corresponds to m.
pectoralis in more derived tetrapods, as it originates at the coracoid or a more ventral element
and inserts at a prominent crest at the medial side of the humerus. The only muscles considered
in comparative studies on the evolution of pectoral girdle musculature in basal tetrapods are
adductors and abductors of the pectoral girdle, although further muscles are supposed to be
required (Shubin, 2006, Diogo & Abdala, 2007).
According to the presented stress distributions at the shoulder girdle, the transmission of
body weight between the trunk and the forelimbs was realized in two different ways in mammals
and reptiles. In an extended limb position the weight forces of the anterior part of the body are
105
carried by the forelimbs, the scapula, the ribs, and sternum. The reaction forces expand as
follows: 1. ground- forelimb- scapula via mm. serrati to the ribs and the vertebral column, 2.
forelimb and scapula via the m. pectoralis profundus to the sternum (Preuschoft et al., 2007).
Therefore, the first ribs are straight and strong, and the thorax is narrow. Cursorial mammals
usually do not display great abduction movements of the forelimbs to prevent horizontal forces,
which cannot be countered by the existing structures. The narrowness of the trunk also helps to
minimize the rotational moments in one limb stance, which is essential for saving muscle energy
(Preuschoft et al. 2007).
A different situation is presented by living reptiles, amphibians and early tetrapods, which
usually hold their limbs in a sprawling position. The weight forces are carried by the humerus, the
scapulocoracoid and the ribs. The sprawling posture necessitates horizontal tensional forces
which can be produced by the m. pectoralis. The resultant of ground reaction force and pull of the
m. pectoralis passes through the shoulder joint. These joint forces can be separated in two
components, a horizontal part and a vertical part. The ground reaction forces run through the
forelimb to the shoulder joint. The horizontal part of the forces flows through the coracoid plate
to the sternal apparatus, whereas the vertical part runs via the scapula, transmitted by the
posterior part of m. serratus over the middle ribs to the vertebral column (Preuschoft & Gudo
2005, Preuschoft et al. 2003).
Figure 58 Scheme of two basic shoulder girdle constructions. Compressive force transmission is indicated as grey
vectors, tensile stresses as black arrows. A: sprawling position as in early tetrapods and recent reptiles. Compressive
force transmission from the shoulder girdle to the trunk is accomplished from the coracoid via the sternal element to
the ribcage. Tensile stresses are carried by the m. rhomboideus, m. serratus and m. pectoralis. B: extended position as
in cursorial mammals. No bony connection is present between the shoulder girdle and trunk. Transmission of body
weight force Fg and ground reaction force Fgr is solely accomplished via muscular structures.
106
The observed compressive stresses in both the sprawling and the cursorial model are in
remarkable accordance with the major skeletal elements described in extant species. According to
the results of the morphological comparison the position of the shoulder girdle elements is similar
in all tetrapods. The scapula is always positioned parallel to the rib cage covering the first ribs
latero-cranially. In addition the dorsal margin of the osseous scapula, or cartilaginous
suprascapula, lies nearly parallel to the vertebral column. Lateral inclination of the scapula ranges
from 50° to 60°. M. serratus has been determined to play the major role in transmission of body
weight to the forelimbs (English, 1978, Jenkins & Weijs, 1979; Jenkins & Goslow, 1983; Larson &
Stern, 2007). Enhanced function in body weight transmission leads to an increase of vertical
compression within the ribs. Scapular position in extant quadrupeds correlates to the position of
the most weight bearing ribs.
Thus, all tetrapod bodies underlie the same mechanical conditions since the body is
exposed to earth gravity. In case of the shoulder girdle this means that appropriate bony,
muscular and tendinous structures are required for a successful transmission of body weight over
the limbs to the ground to enable locomotion. These results will be the basis for the further
investigation of the shoulder girdle in extant crocodiles and sauropod dinosaurs. Although
sauropod dinosaurs exhibit the "typical" components of a reptile shoulder girdle, morphological
changes of the shoulder girdle elements as well as in the musculature are visible and necessary.
As a consequence of the presented results, further investigations should combine a
functional approach with an analysis of rugose attachment sites in the fossil. The requirements of
the musculo-skeletal system engaged in terrestrial locomotion could then be correlated to
muscular and tendinous structures. With this approach the musculature can be reconstructed,
taking into account phylogenetic as well as functional considerations for extinct animals.
During evolution the number of shoulder girdle elements in both extinct and extant
tetrapods underwent some alterations. The most profound changes appear at the transition from
water to land, when it comes to a striking increase of element number and structural expansion,
and during mammalian evolution, where a reduction of elements takes place (Romer & Parsons,
1977). With regard to the results in this study this is strongly related with fundamental changes in
locomotor behaviour. Besides a variety of selective pressures like climate changes and other
exogene influences (Clack, 2007; Carroll, 2005) biomechanical conditions have to be considered as
selective pressures during the evolution of the tetrapod bauplan as well. This was recently
pointed out in a study about body mechanics in the early tetrapod Tiktaalik roseae (Hohn et al.,
submitted). It was shown that Tiktaalik was able to organize its movements so as to keep the
107
necessary muscle force at the lowest possible level, because energy expenditure is proportional to
the muscle forces required for movements (Alexander, 2002).
Movement dependent mechanical stimulation has a significant impact on cartilage and
bone remodelling during embryogenesis and postembryonic life (Carter 1987; Lightfoot & German
1998, Jones et al. 2007; Gomez et al. 2007). Influenced by epigenetic factors the structure of bone
seems actually to depend on its mechanical function, its shape directly results from mechanical
stress acting within the locomotor system (Müller & Streicher 1989; Newman & Müller, 2005).
Thus the overall mechanical impact on bone formation is beyond discussion, but as it has been
stated in the first paragraph, the major role of compressive stresses still is neglected by some
authors (Biewener et al., 1991; Carter & Beauprè, 2001; Currey 2002; Liebermann et al., 2003).
Mechanical stresses inducing bone formation in this approach are always compressive stresses, as
predicted by Wolff (1892) Pauwels (1965) and Kummer (1962). This has been recently supported
by both cell physiological studies (Kaspar et al., 2000; Rath et al., 2008, Takahashi et al. 1998, Sato
et al., 1999; Ikegame et al., 2001) and biomechanical approaches using the finite-element method
(Witzel, 1985, 1996, 2000, 2003, 2005a, 2005b; Sverdlova & Witzel, 2010).
The origin and development of the tetrapod bauplan are widely discussed in anatomy,
developmental biology, and palaeontology. The importance of epigenetic factors in ontogenetic
and evolutionary processes was pointed out by Newman & Müller (2005) stating that “the
increasingly autonomized skeleton continues to be reshaped and embellished by novelties that
arise from the continued responsibility of chondrogenetic tissue to the biomechanical
environment.” A question still unacknowledged is how the evoked stresses lead to the
development of novel elements and their position in the locomotor system. One of the most
dramatic changes in pectoral girdle evolution is the dissociation from the head skeleton enabling
the forelimbs to move independently from the head. Studies on chicken cranial neural crest cells
indicate their influence on the organisation of the skeletomuscular connectivity between the head
and the trunk (Huang et al., 1999) and help to explain how skeletomuscular connectivity between
the pectoral girdle and the head is maintained during tetrapod evolution, even though the
pectoral girdle and the skull became dissociated during this process (Mc Gonnel et al. 2001). The
repositioning of muscular attachment sites due to migrating cells during the evolution of the
vertebrate skeleton could provide ground for a better understanding of the formation of novel
construction within the tetrapod bauplan (McGonnel, 2001). If muscles change their position
within in the musculoskeletal system, they evoke different stresses acting on the involved skeletal
elements or developing tissue. This was recently pointed out by a study on hip joint
108
endoprothesis investigating physiological loading conditions in the femur in different loading
cases during locomotion using finite-elements (Sverdlova & Witzel, 2010). In an evolutionary
context this may result into remodelling of actually present bony tissue during ontogenesis or
induce the formation of novel structures.
Morphological changes occur gradually rather than suddenly. During early tetrapod
evolution a striking increase of the enchondral elements (scapula, coracoid) of the shoulder girdle
can be observed (Shubin et al. 2006). At the same time the dermal elements (cleithral series) of
the shoulder girdle, which are functionally engaged into aquatic locomotion and feeding are
reduced (Romer & Parsons, 1977; Shubin 2006; Downs et al. 2008). Enhanced terrestrial
locomotion behaviour, as indicated by the fossil record, led to an increase of body weight, as a
result of gravitational acceleration. This results in an increase of the counteracting muscle forces,
which are required during locomotion. The gradual changes in the shoulder girdle construction,
which can be observed in basal stem tetrapods, therefore indicate a close relationship between
form and function in the vertebrate skeleton. However, Romer & Parsons (1977) attributed the
enlargement of the ventral elements to the downwards acting body weight which otherwise
would push both coracoids apart. By contrast, the results gained from the analysis of the 3-D FE
models of tetrapod bodies indicate that body weight is transmitted from the trunk via muscular
structures like the mm. serrati to the scapula and to the supporting forelimbs. Thus, it is
suggested that the ventral elements increased as a result of the mechanical stresses occurring
during terrestrial locomotion. In consequence the ventral elements provide additional attachment
sites for the necessary musculature. The formerly described traits of the tetrapod bauplan can be
viewed as biomechanical necessities of standing and walking on limbs. Some of these features like
an enhanced development of the shoulder girdle elements especially on the ventral side and the
appearance of short ribs can first be observed in Tiktaalik (Shubin et al. 2006).
In marsupials, the appearance and reduction of certain elements can be observed during
the individual development. In marsupials the development of the shoulder girdle consists of two
parts. The embryonic and neonate stages show a strong similarity to the shoulder girdle
architecture in monotremes, where the scapula, metacoracoid, procoracoid, first rib, paired
sternal elements, unpaired sternal elements and clavicle together form the shoulder girdle. There
is a strong functional- adaptive correlation in this temporary formation, as the neonates have to
reach the mother`s pouch after birth on their own, crawling with the aid of well developed
forelimbs (Klima, 1987). The stresses occurring in the marsupial pectoral girdle require structures
that are able to transmit large medially directed forces, a condition which is only present in a
109
sprawling limb posture. When the neonate has reached the pouch certain elements are reduced
and the tight connection between both sides is lost. In particular, the metacoracoid is reduced to
the coracoid process, and the procoracoid becomes the praeclavium. The unpaired sternal
element fuses with the paired elements generating the manubrium sterni (Klima, 1987). The
maintenance of function in a system at all times is a precondition for morphological
transformation. Even if the fossil record is often lacking gradual transitional forms along the
evolutionary line between different clades, there is no doubt that function was maintained in
every stage. Furthermore, a structural element can only be transformed within the constraints
established by the surrounding structural elements and the organism as a whole (Gudo &
Homberger, 2002).
The present data are consistent with these conclusions. Indeed, our results reveal that
location and shape of bony structures and associated muscles depend on the occurrence and
distribution of compressive and tensile stresses, respectively.
5.2
FESS of the crocodilian shoulder girdle
In the present approach, equilibrium within the 3-D finite-element model of a crocodile
has been reached by application of functionally required muscular elements in iterative steps.
Function here is related to the relative movements of each shoulder girdle element in the 3dimensional space. To maintain equilibrium element movement is countered by the necessary
muscles, which allow determining their function in maintaining balance. In the following, muscles
are discussed according to their function revealed in the study, compared to former descriptions
and their impact on the system, which relates to link stresses.
M. serratus superficialis transmits weight forces from the trunk to the scapulocoracoid. It
supports the trunk, while simultaneously forcing the scapulocoracoid to rotate caudally. The large
expansion of m. serratus superficialis from the lateral side of the trunk to the caudal border of the
scapula is advantageous for transmission of large forces. Muscles are distributed over a wide area,
thus producing sufficient muscle force. At the same time, these forces produce significant
rotational moments at the glenoid joint, therefore, the scapula acts as a lever. The weight
supporting function of m. serratus superficialis as described in the literature (Brinkmann, 2000;
Meers, 2003), is consistent with the recent results. Although m. serratus profundus is supposed to
support transmission of weight forces, its function cannot be determined during the process.
However, it can be stated that it does not account for any rotation of the scapulocoracoid and it is
less loaded after calculation compared to the resulting link stresses.
110
The rotational moments induced by m. serratus superficialis need to be compensated by
appropriate muscles, which were identified as m. trapezius and m. levator scapulae. Their
function as cranial rotators of the scapula has formerly been described in the literature
(Brinkmann, 2000; Meers, 2003). By contrast, their function in counteracting these movements
during one limb stance had yet to be determined.
It became obvious during calculation that m. trapezius and m. levator are not sufficient to
prevent the rotation of the scapulocoracoid. M. costo-coracoideus is required to maintain
equilibrium in the shoulder joint, as it contributes to cranial rotation of the scapulocoracoid and
translation of the coracoid relative to the sternal region of the trunk. Both m. trapezius and m.
costo-coracoideus have been enhanced in the final calculation step by an increase of link`s cross
section and pretension leading to higher forces. As no calculation was possible before, this
procedure points out the importance of these structures for cranial rotation of the
scapulocoracoid and caudal translation of the coracoid for static equilibrium. Although m. levator
scapulae is required to maintain equilibrium during calculation, this function is not reflected by
the resulting link stresses, which were near zero.
In crocodiles m. pectoralis major has a broad fan-shaped appearance, originating from the
most cranial aspect of the sternum almost reaching the pelvis (Meers, 2003; Brinkmann, 2000).
M. pectoralis in the present approach is separated into two different functional compartments.
The anterior part prevents lateral movements of the humerus and is confirmed in its role as the
main humeral adductor. According to the dissection, the majority of its musclefibres are directed
anterio-posteriorily, indicating its relevance as a humeral retractor. Considering the aquatic
lifestyle in crocodiles, strong humeral retraction is required to support swimming motions. Based
on the present results, these parts of the muscles do not contribute to static conditions. Link
stresses after calculation reveal their contribution under static loading.
In the course of adjusting equilibrium, m. teres minor, m. latissimus dorsi and m. scapulohumeralis were applied to the system (calculation not shown in the results). None of these
calculation steps could be accomplished by any possible combination. They obviously do not
contribute to static equilibrium in the shoulder girdle system and were excluded from further
calculations. Muscles originate from the posterior part of the vertebral column, insert at the
caudal edge of the humerus (m. latissimus) or run from the medial (m. teres minor) or caudal edge
(m. scapulo-humeralis) of the scapula to the humerus. These muscles are commonly described in
the literature as humeral retractors (Meers, 2003; Brinkmann, 2000), although their line of action
additionally forces the scapula to rotate caudally
111
At this point of the investigation, it can be stated that function could only be determined
for those muscles, which clearly contribute to static equilibrium of the shoulder girdle in one limb
stance. The function of all further muscles associated with the shoulder girdle remain unclear.
Nevertheless, they have a certain impact to the system as they contribute to decrease maximum
compressive stresses, from -105 MPa in early stages of the process to -20 MPa in the final
calculation, which is within the predicted physiological stress range of -2 to -20 MPa typical for
cortical bone (Witzel & Preuschoft, 2005).
As the absolute muscle forces cannot be determined at this point of investigation, they
should be evaluated qualitatively, by comparison of their load after calculation. For this purpose
link stresses after calculation were read out. For some of the muscles the observed stress values
correlate with their functional relevance in static loading during one limb stance (e.g. m. serratus
superficialis, m. costo-coracoideus, m. trapezius and m. pectoralis), while for others no functional
significance was revealed during calculation (m. levator scapulae, m. serratus profundus). As for
volume elements, deformation in the model after calculation is indicated by black outlines. The
direction of the link elements changes their predefined position according to the relative
movements of each element in the model. Consequently, the line of action of some muscles is
altered. In some elements this does not lead to relevant changes, and these elements still retain
their function. In case of m. levator scapulae, the trunk and scapulocoracoid movement lead to an
expansion of the link element by changing its relative line of action. Expansion due to an
alteration in the lines of action of the muscles is therefore the most reasonable explanation for
the observed zero stresses within m. spiralis and m. serratus profundus as well.
The finite-element method has been part of a number of studies investigating vertrebrate
skull (Daniel & McHenry, 2001; Rayfield, 2004, 2005; Dumont et al., 2005; Ross et al., 2005;
McHenry et al., 2006; Wroe et al., 2008) or rib morphology (Fujiwara, 2009), but none of them
used a multi-body approach to reveal locomotion related stresses within the skeletal structure.
The present study is the first attempt of a multi-body FESS applied to the shoulder girdle. Link
elements were used according to former finite-element approaches, because they provide
stability during calculation, especially if a number of elements act in a 3-dimensional space.
Although stability was maintained, the resulting forces were found to be insTable during
calculation. Again, it is emphasized that former studies never included more than two separate
volumes connected by link elements (McHenry et al., 2006; Wroe et al., 2008), thus the presented
difficulties had not been reported until now. It is concluded from the results that link elements
are insTable during the calculation process and do not provide any information about absolute
112
force values, which can be compared to the musculature. As mechanical forces act in the x,y,zaxis, they provide detailed information, thus directed forces instead of link elements are
recommended for further multi-body approaches. The maximal muscle forces, which can be
determined reflect the occurring stresses in one loading case. For a reasonable comparison of
muscle cross sections, as it was initially aimed at by this approach, a number of different load
cases should be calculated and superimposed, as it has been performed in the FESS of a
Diplodocus skull (Witzel & Preuschoft, 2005). Such an approach would lead to an overall
impression of the muscular activity of the shoulder girdle musculature. Unfortunately, there are
no EMG data of the crocodilian shoulder musculature available, neither in statics or during
locomotion, which could provide ground for further comparison.
Nevertheless the present 3-D FESS of a crocodilian shoulder girdle shows compressive
stress distributions, which are near the physiological value of bone and enables the first synthesis
of the main features in the scapulocoracoid of a crocodile. The function of the shoulder girdle
muscles could be determined for some of the elements with special emphasis on their ability to
maintain equilibrium in the glenoid joint. The findings provide the basis for the following
investigation of the scapulocoracoid in sauropods. Considering the present results together with
the information gained from morphological comparison in extant tetrapods, and 3-D FE analysis
of solid tetrapod bodies, the preconditions for a 3-D FESS of Diplodocus longus will be configured.
5.3
3-D FESS of the shoulder girdle of Diplodocus longus
5.3.1 Reconstruction of the shoulder girdle of Diplodocus longus.
5.3.1.1 Skeletal elements
One of the most debated issues in reconstruction refers to the overall position of the
shoulder girdle on the trunk and especially the inclination of the scapula. Several indicators for
scapula reconstruction in sauropod dinosaurs were considered in former publications. In first
attempts the position was deduced from either comparison with extant reptiles without any
functional considerations (Hatcher, 1901; Osborne and Mook, 1921) or refer to the "death pose"
in the excavated fossil (Gilmore, 1932), leading to 60° or 45° inclination of the scapula to the
horizontal plane. The taphonomic processes during the decay of a vertebrate body affect both the
individual bones and the skeleton as a whole (White and Folkens, 2005). The position of the
individual bone in the skeleton after decay is different from the initial situation. Each bone looses
its connection to the skeletal composite at different stages of the decay, depending on the
113
strength of the adjacent ligaments and muscles. Therefore bones with predominantly muscular
connections, as the scapula, are disconnected earlier from the composite than vertebraes do.
Further distortions could arise from other animals feeding on the carcass and physical influences,
like water, heat and geological pressure during the fossilization process (Lyman, 1994). Therefore,
the scapular position deduced from "in-situ" finding is somewhat doubtful.
The orientation of the glenoid joint serves as another factor affecting scapular position.
The glenoid is reconstructed facing ventrally in order to allow the forelimb to articulate in an
upright columnar stance, reflecting a graviportal forelimb configuration (McIntosh et al. 1997,
Wilson & Sereno 1998, Bonnan 2001, Upchurch et al. 2004, Wilhite, 2003). Since the limbs in
quadrupeds are straightened with increasing body size to prevent high rotational moments at the
joints (Biewener, 1989, Reilly et al. 2007), highly abducted limbs in sauropods are unlikely. Our
results, however, show that optimal force transmission is not limited to a strictly vertical
orientation of the humerus but is also present at small angles of up to 20° between the humerus
and the perpendicular in lateral view. A slight abduction of the humerus as presented in this
thesis therefore is reasonable, because without any flexion in the joint, the occurring compressive
stresses acting on the joint’s surface would be too high to resist (Christian et al., 1999).
Furthermore, if the humerus was positioned orthogonally to the glenoid joint, the scapula would
possess an inclination angle of about 40° or less (McIntosh et al. 1997, Wilson & Sereno 1998,
Bonnan 2001, Upchurch et al. 2004, Wilhite 2003). According to the results of the basic finiteelement tetrapod models an inclined position of the scapula is visible in both the sprawling as well
as in the extended limb stance, but it never reaches a degree of inclination below 50°.
Wilhite (2003) furthermore argued that, resulting from best-fit articulation of digitized
skeletal elements, the humerus has to be articulated in a ventrally facing glenoid. He assumed the
thickness of cartilage to be uniform across the joint surfaces. Thus, the distances between
articulating elements should be uniform as well. The distance between joint surfaces therefore
should be the minimum distance required for the highest bone surface rugosities to rotate across
the joint without touching the corresponding joint surface. The condition in Alligator joints is used
to confirm these assumptions (Wilhite, 2003). The study provides no mechanical confirmation. Up
to now, reconstruction of cartilage is not possible for fossil vertebrates because a number of yet
unknown parameters to calculate material properties and structural behaviour is not available.
Thus, the absolute dimensions of cartilage in extant animals remain speculative.
An orthogonal position of the humerus of Diplodocus is not only assumed in lateral but in
frontal view. The orientation of the maximal width of the saddle shaped humerus facet is
114
predicted to lie in the transverse plane of the glenoid joint. It was rotated along the long axis of
the humerus at 45° compared to the original mount of Diplodocus longus (SMF), which is
generally accepted until today (Bonnan, 2001; Wilhite, 2003, Schwarz et al. 2007; Remes et al.
2007). Following the assumption of a columnar forelimb with a posteriorily facing glenoid joint,
little bending in the brachial-antebrachial joint and a restricted movement of the humerus in an
anteromedial- posterolateral direction is assumed (Wilhite, 2003). By contrast, in the present
study abductional movements in the forelimbs are predicted, because the enormous length and
weight of the neck causes moments in the anterior part of the trunk, which can only be countered
with abducted forelimbs (Preuschoft und Gudo 2005). The saddle shaped glenoid facet and the
flattened caput humeri of sauropods indeed enable abduction/adduction as well as
extension/flexion. Therefore, a moderately extended posture of the forelimbs, with slight flexion
in the shoulder as well as in the elbow joint in lateral view, and a moderate ability for abducting
movements in frontal view is assumed.
In the present study the glenoid joint is represented by a simple ball and socket structure,
which does not resemble the situation in the fossil. No differentiation was made between bone
and cartilage at this point. This was done because of the uncertain dimensions and material
properties in sauropod cartilage as stated before. Nevertheless the model was satisfactory for the
present attempt as the glenoid joint is able to transmit forces and allow for rotational movements
with a minimum of friction, a state which can be termed as physiological. The results reveal a
concentration of stresses in the scapulocoracoid at the position of the glenoid joint. These
stresses tend to spread caudally, if the bauraum would provide more space than provided in the
model. An increase of contact area provided by an assumed cartilage cover would have led to
caudal expansion of compressive stresses and therefore of bone material at this location. It is
assumed that the missing caudal inclination of the scapula border could relate to this
circumstance. Finally, this would lead only to marginal modifications. It is recommended for
future investigations to take these considerations into account, thus leading to an elaborated
implementation of the method.
In this investigation the scapulocoracoid is positioned parallel to the rib cage, covering the
first thoracic rib, whereby the glenoid is placed shortly before the ribcage. This is determined to
be the basic morphological conditions as they are present in all extant quadruped tetrapods (see
morphological comparison in extant tetrapods). Beside this apparent relationship a close
correlation between the most weight bearing ribs and scapular position was recently described
(Fujiwara et al., 2009). These ribs serve as attachments for the body weight transmitting muscle,
115
the m. serratus, and show the highest resistance against vertical compression. As these ribs lie
almost beneath the scapula, scapular position was concluded based on this information (Fujiwara
et al., 2009). Although the weight bearing ribs can be determined the relative position of the
scapula in extinct animals remains unclear if it is deduced exclusively based on the ribs. Length of
m. serratus superficialis as well as the inclination of the line of action in the fibers of the muscle is
unknown in fossils. If these variables change, scapula position will change its direction as well. In
the present investigation the ribs, which serve as attachment sites for m. serratus superficialis lie
closely behind the scapula as it is predicted by the former approach. In contrast, their position
was reconstructed based on mechanical necessities, providing a mechanically effective lever arm
for the acting musculature leading to the observed shape of the scapulocoracoid. Although
determination of weight bearing ribs does not allow for a final reconstruction, it contributes
necessary information. At this point further investigations are contemplated including
measurements on a high number of individuals to evaluate the most weight bearing ribs in
sauropods.
In the shoulder girdle system the position of one element has an impact on the other
component. Thus the reconstruction of the coracoid position derives from former assumptions. A
number of questions arise concerning the presence of any contact between the coracoids and the
sternal plates, the morphology at this location, and the relative position of the elements. The
sternals are often misinterpreted in the fossil record or found not to fit to any "reasonable"
reconstruction and were therefore left out of the whole mount (McIntosh, 1988). The original
mount of Diplodocus longus (SMF) presents no sternal plates, with the coracoids placed more
laterally, thus leaving a wide gap between the coracoid plates in front of the ribcage. By contrast,
in the mount of HQ1 (SMA) the sternal plates lie in front of the ribcage and in close contact to the
coracoids. These two extreme positions are reflected by the literature. The mounted skeletons of
a number of sauropods indicate no direct connection, in fact, the sternals were placed posterioventrally far behind the shoulder girdle. The first reconstructions of Diplodocus and
Camarasaurus, although differing in many issues, recommend a possible connection between the
coracoids and the sternals, but leave out any information on the definite position (Hatcher, 1901;
Gilmore, 1932). As soon as the coracoids are seen as integral parts of the sternal apparatus, they
should be connected with the sternal plates (Schwarz, 2007). Thus, several reconstructions
recommend that the coracoids should be oriented vertically in front of the rib cage (BorsukBialynicka, 1977; McIntosh, 1997; Wilhite, 2003).
116
According to the observations in the present study and of other researchers (Schwarz et
al., 2007) such a morphological condition cannot be detected in any extant tetrapod, because it
would result into the loss of the ventral connection of the sternals and the distal ends of the
thoracic ribs. The connection therefore can only be maintained if the coracoids are placed in the
same plane as the sternal plates, thus leading to an inclination angle of the scapula of 60° to the
vertebral column (Schwarz et al., 2007). This can be seen as strong morphological argument,
which is confirmed by the results of the present investigation.
If the scapulocoracoid is placed relative to the ribcage as it is predicted in the present
study, the frontal aperture is narrowed, bringing the coracoids closer to the midline. With the
assumed limb posture a medial component of forces runs through the glenoid and extends to the
midline of the trunk. Therefore, the circle of forces in the horizontal plane can only be completed
if the forces flow from the humerus via the coracoid to the sternals and back to the trunk via the
distal ends of the ribs. This is consistent to the results of the 3-D FE models of tetrapod bodies, in
that locomotion requires a compression resistant structure between both half's of the shoulder
girdle in order to provide effective force transmission within the shoulder girdle system. A similar
condition is present in the shoulder girdle of extant reptiles, where the transmission of forces is
most visible in the FESS of the crocodilian shoulder girdle. Here high compressive stresses occur at
the contact region between the trunk and the coracoid. The coracoids and sternals in sauropods
do account for this function in the present case. The coracoid therefore resists the compressive
stresses which occur between the shoulder joint and the sternum. In the present case the sternals
lie in close medio-caudal contact to the coracoid plate to transfer the occurring forces to the ribs.
Compared to the condition in extant reptiles and mammals the circle of forces in
sauropods reveals an alternative way. In brief, force transmission in the mammalian shoulder
girdle is accomplished only by the shoulder girdle muscles. Forces in the extended limb position
flow mainly vertically, reflected by straight first ribs lying beneath the scapula, giving the thorax a
narrow appearance (Preuschoft et al. 2007, Preuschoft et al. submitted). No medially directed
forces occur, as mammals usually do not display significant abduction in the humerus. The
sprawling limb position in extant reptiles and early tetrapods produces additional horizontal
forces, which require a medially positioned compression resistant structure as stated in the
former paragraph. Compared to mammals, crocodilians possess short, but stout anterior and
rather strong middle ribs, where only the latter are connected to the sternum. According to the
present reconstruction, the anterior trunk in sauropods is narrow like in mammals. They show
particular strength of their middle ribs, while the anterior ribs are longer than in crocodiles and
117
seem to have reached the sternal elements. However, the curvature of the middle ribs is similar
to cursorial mammals. The vertical portion of the occurring stresses is enhanced, while the
horizontal parts are lowered. In consequence the sauropod coracoids are reduced in their
dimensions and modified in shape, but still present. It can be concluded that the shoulder girdle
system in sauropods was modified to the requirements of a graviportal locomotor type using their
reptilian skeletal components.
Like their overall position in the shoulder girdle system, the morphology of the connection
between the coracoids and sternal plates is strongly debated. Two different conditions are
currently assumed. In the first condition the elements are embedded in a ventral cartilaginous
plate, connected to each other by sutures and/or synchondroses (Whilhite, 2003; Schwarz et al.,
2007, Remes, 2008). Alternatively the elements are connected via a joint, similar to the situation
in extant reptiles (Bakker, 1987). Such a condition would lead to an increase of the arc of rotation
in the forelimbs and enhance mobility in the shoulder girdle. Some authors assume that the only
motion within the shoulder girdle occurred in the gleno-humeral joint (Henderson 2006; Bakker
1987, Paul 1987; Christiansen 1997; Wilhite 2003).
During quadrupedal locomotion translational movements of the shoulder girdle occur. In
lacertilians (varanids, chameleon) the coracoid moves translational relative to the sternal
apparatus (Peterson, 1973; Jenkins & Goslow, 1985; Lilje, 2007). The junction between coracoid
and sternum in these reptiles exhibits a” groove and tongue” mechanism, where the bony part of
the sternum acts as the groove in which the coracoid is moving. In crocodilians, these movements
are enabled by rotation of the elongated coracoid relative to the sternum. Coracoid and sternum
exhibit a “rim and tongue” structure as well, but consist of cartilage rather than bone. In both
cases, the junction is surrounded by a joint capsule embedded in connective tissue. The condition
in sauropods differs from modern lacertilians, as no bony rim and tongue structure can be
detected. This holds true for the condition in crocodiles as well as the cartilaginous parts would be
diminished during decay and fossilization. In fact the cartilaginous tongue and rim structure in
crocodiles has been either not reported or their relevance has been underestimated in the
literature. The histological cross sections of a Caiman presented here point out the structure in
the coracosternal joint, which indeed holds all features of a true joint. Based on this observation a
similar condition is assumed to have been realized in sauropods. In fact the mobility in the
shoulder girdle becomes restricted with an increase of body size and weight. The FESS of the
scapulocoracoid of Diplodocus shows a model in which the coracoid part and the sternal region
are connected through contact elements enabling relative movements. The joint is hereby
118
stabilized by muscles e.g. the m. costo-coracoideus. The mobility of this joint although restricted,
is indispensable for the function of the shoulder girdle, because it provides the ability for force
transmission from the coracoid to the sternum and at the same time leaves a certain degree of
mobility in the shoulder region.
5.3.1.2 Muscles of the shoulder girdle
One major difference in the reconstruction of the shoulder girdle muscles in sauropods in
this investigation compared to other attempts is the mode of determination. Muscle
reconstructions were performed based on the presence and position in extant crocodilian
myology, like for Opisthocoelocaudia (Borsuk-Bialinicka, 1977; Schwarz, 2007), Camarasaurus,
Apatosaurus, and Diplodocus (Wilhite, 2003; Schwarz, 2007) or with the aid of the extant
phylogentic bracketing (EPB) method, in which both archosaur outgroups, crocodilians and birds
are considered (Remes, 2008). The latter study investigates the basal sauropods up to
Patagosaurus, which is the most phylogenetically derived sauropodomorph. Except for Remes
(2008), no reconstructions include all muscles acting on the shoulder girdle. In the present study
the musculature was determined by its function to maintain static equilibrium and transmission of
weight force from the trunk to the limbs. Although crocodilian myology serves as a basis for
muscular structures suited for this function, their line of action and forces values were chosen by
mechanical necessity. During the process all muscles assumed to be present in the sauropod
shoulder girdle based on the terminology of Alligator mississipiensis could be determined
according to their possible function in static equilibrium. This includes muscles connecting the
trunk and the pectoral girdle as well as muscles running from the pectoral girdle to the humerus
and the antebrachium. Additionally some muscles were included, which connect the humerus and
the antebrachium.
In Schwarz et al. (2007) m. serratus superficialis in Diplodocus inserts at the medial surface
of the suprascapula and the caudal margin of the scapula. Orientation of the cranial bundles of
muscle fibers is supposed to be ventrally, as the caudal bundles are orientated caudo-ventrally. Its
origin is assumed to be at the fourth or fifth thoracic rib. An insertion at the caudal margin of the
scapula is consistent with the present results. M. serratus superficialis is the main weight bearing
muscle, exerting high forces on the scapula. Therefore an insertion at the cartilaginous part of the
scapula is mechanically inadequate. The calculation further reveals the caudal bundles to be
oriented not only caudo-ventrally but laterally to a certain degree, thus contributing to reduce
bending in the long scapula. Remes (2008) suggested the muscle to insert at the first third of the
119
scapula. A placement at this location would inhibit an effective lever arm of the muscle and is
inconsistent with any comparison with extant reptiles and crocodiles. In contrast, he
reconstructed the m. serratus profundus in two portions inserting on the medial part of the distal
scapula; one originating from the 3-6th thoracic rib, and the other part from the last cervical and
first and second thoracic rib. Functionally the caudal part is equivalent to m. serratus superficialis
in the present approach, while the cranial part is consistent with the results. In the present study
m. serratus profundus originates from the last cervical and first to third thoracic rib, inserting at
the medial part of the distal scapula. Its line of action is mainly vertical, and transmits a lower
amount of weight forces than m. serratus superficialis, while rotating the scapula only to a minor
degree. It contributes to equilibrium of forces in the shoulder girdle system by countering the
negative vertically directed forces exerted from the cranial rotators of the scapula. This functional
implication could only be revealed by a functional approach and was not documented before.
However, no mechanical evidence is present to support the assumption of a cranial expansion of
this muscle as predicted by Schwarz et al. (2007). In the studies by Wilhite (2003) and BorsukBialinicka (1977) this muscle was not taken into account for reconstruction.
The elongation of the scapula is one apparent feature of sauropods. The scapula acts like
a long lever since muscle forces insert there. In the current reconstruction the area below the
scapula provides space for an effective distribution of m. serratus superficialis according to the
high forces it has to exert. As rotational moments refer to the distance to the pivot, the inclination
of muscles fibers and therefore the distance to the glenoid joint had to be low enough to prevent
high rotation, but at the same time wide enough for an effective muscle diameter. The most
reasonable placement was chosen in this calculation. If the scapula would be oriented horizontally
this would not provide enough space for the muscle, meanwhile rotational moments would
increase and result in larger cranial z-directional forces. Both, the rotational moment and the
cranially directed z- forces would have to be countered. This would result in an unphysiological
amount of forces to be produced by the antagonistic muscles. Therefore, elongation of the
scapula is interpreted as an adaptation to high body weight. According to m. trapezius and m.
levator scapulae a shorter lever arm could not provide sufficient area for muscle insertion to exert
the forces required to counteract the high rotational forces induced by the weight bearing m.
serratus superficialis. Counteracting muscles would be thickened beyond a physiological limit to
produce the same forces, in order to reach sufficient counteracting rotational moments.
M. rhomboideus in this study originates from the dorsal margin of the trunk and inserts
medially at the most distal end of the scapula. It is assumed to have inserted in part at a
120
cartilaginous suprascapula, which was not modelled in the present case because of limitations
described above. Schwarz et al. (2007) assumed a cranial expansion of this muscle, while the
insertion and origin are consistent with the current results. Its main function here is to prevent
lateral movements of the distal scapula, as its direction of forces is oriented in the medial plane.
Again no mechanical evidence is provided in the present study to assume a different position as
predicted by Remes (2008). The expansion of the suprascapula declines during sauropod
phylogeny and correlates with a reduction of size and function in m. rhomboideus. This is
supported by the predicted cross sections for this muscle and the provided space for a distally
expanding cartilaginous part in the present model.
Both m. trapezius and m. levator scapulae act as cranial rotators of the scapula, thus
counteracting the caudal rotation of the scapulocoracoid resulting from m. serratus superficialis
forces. The first was reconstructed to run from the last third of the scapula to the dorsal midline
of the trunk at the level of the last cervical and first thoracic vertebrae. The latter expands in two
portions from the cranial border of the second third of the scapula and the acromial region to the
cervical ribs. The reconstruction by Remes (2008) is consistent with our results, except he
assumed the muscular lines of action to be crossed, leading to a ventral insertion at the scapula of
m. trapezius and dorsal insertion of m. levator scapulae, respectively. Functionally this
corresponds to the present results, but anatomically makes less sense, as such a condition is
absent in extant tetrapods. Reconstruction of muscle orientation in m. levator scapulae by
Schwarz et al. (2007) is mainly in accordance to the present study, although the cranial expansion
in the study is enhanced. The presence of m. trapezius was not considered in this investigation
(Schwarz et al. 2007), or by Borsuk-Bialynicka (1977) and Wilhite (2003).
The present study reveals that m. costo-coracoideus in this study has a large impact on
static equilibrium in the shoulder girdle. The coracoids were reconstructed to allow for this
muscle a line of action running caudally beneath the glenoid joint. Because of its small lever arm,
it contributes to cranial rotation to a minor degree, but translates the coracoid caudally, which is
essential in maintenance of static equilibrium. Without m. costo-coracoideus no equilibrium could
be established in any calculation. The coracoid plates were reduced during sauropod evolution,
but are still present, thus provide insertion areas required by this muscle. M. costo-coracoideus
has been reconstructed to run from the lateral side of the coracoid anterior of the glenoid to the
ventral ribs (Remes, 2008) or from the medial side of the coracoid dorsally to the ribs in front of
the glenoid joint (Borsuk-Bialynicka, 1977). Both reconstructions do not contribute to the former
determined function of this muscle, as the predicted line of action in that study would prevent
121
any translational movements of the coracoid relative to the sternum. The muscle has not been
described by Schwarz et al. (2007) and Wilhite (2003).
A differentiation in a superficial and a profund part of m. costo-coracoideus was assumed
only by Meers (2002). M. costocoracoideus pars profundus in crocodiles originates from the
cranial surface of the free ribs, without any fibers inserting at the sternum. In contrast, the
superficial part of m. costo-coracoideus has been found to insert at the sternum (Meers, 2002), a
stabilizing function of the sternum was predicted to be insignificant (Meers, 2002). The presence
of m. costo-coracoideus profundus in sauropods is first described in this study. In the present
investigation it has been reconstructed because of its requirement as a stabilizer of the
sternocoracoid joint, as it prevents luxation of the joint under static loading conditions. If the
muscle fibers of m. costo-coracoideus, which insert to the sternum, belong to a profound or
superficial part of the muscle is immaterial, because the existence of such parts is supported by
functional requirements.
M. supracoracoideus has been decribed to originate from the lateral side of the coracoid
at its distal end (Remes, 2008; Wilhite, 2003) or placed more proximally (Borsuk-Bialynicka, 1977),
while its insertion is positioned at the deltopectoral crest (Remes, 2008; Wilhite, 2003) or extends
to the lateral head of the humerus (Borsuk-Bialynicka, 1977), respectively. The first placement is
consistent with the present study. It is defined as a humeral extensor, but no function under
statical loading conditions has been mentioned. In the present study m. supracoracoideus is
described as a stabilizer of the glenoid joint, as it exerts the required translational forces to keep
the scapulocoracoid in equilibrium relative to the humerus. Again this muscle has not been
considered in the study by Schwarz et al. (2007).
Schwarz et al. (2007) do not illustrate or mention m. coraco-brachialis brevis. In the study
by Borsuk-Bialynicka (1977) it was placed to originate on the distal medial bend of the coracoid
and to insert at the humerus at two points, the anteroproximal surface of the humerus and
anteromedial surface of the medial condyle. Further, an origin from the lateral surface of the
short coracoid has been assumed or at least placed directly above the glenoid, while its insertion
would have been on the cranial surface of the humerus medial to the deltopectoral crest (Wilhite,
2003; Remes, 2008). The latter reconstructions are consistent to the present strudy. M. coracobrachialis is a strong adductor in crocodiles and lacertilians, corresponding to the strong
abduction of the limbs. It has been assumed to be reduced in sauropods because of their upright
posture. A shift of function therefore is reasonable. In crocodiles the coracoid is positioned at an
approximated angle of 90° to the scapula, therefore the origin of the muscle is far from the
122
glenoid joint and the humerus, accounting for an adductional function. In sauropods the coracoid
has been shifted nearly into the same plane as the scapula, which results into a modification of
the line of action of m. coraco-brachialis. In the present study its line of action runs rather
anterior-posteriorily than medio-laterally. This indicates higher rotational and translational
movements of the scapulocoracoid and protraction of the humerus, rather than an adduction of
the humerus. Therefore its function is comparable to m. supracoracoideus, which acts as an
agonist in maintaining equilibrium between the scapulocoracoid and the humerus.
M. biceps brachii has not been mentioned by Schwarz et al. (2007) and Wilhite (2003).
Borsuk-Bialynicka (1977) placed the origin of this muscle dorso-laterally at the coracoid, just
below the assumed origin of m. coracobrachialis brevis. Its insertion is assumed to be at the distal
end of the radius. Remes (2008) assumed a different insertion at the ulna and found its origin to
be placed more ventrally at the lateral side of the coracoid just below the origin of m.
coracobrachialis brevis (Remes, 2008). The origin at the coracoid assumed in the latter study is in
accordance to the current results. The insertion site cannot be determined with certainty as ulna
and radius were not differentiated in the present investigation. Referring to its line of action, m.
biceps brachii serves as a protractor of the antebrachium, while rotating the scapulocoracoid
caudally. Further functions revealed by this study are the prevention of craniad movements of the
scapulocoracoid. Therefore it contributes to equilibrium between the scapulocoraoid and the
humerus.
The function as well as origin and insertion sites for the m. triceps complex have been
described differently in the literature. In crocodiles this muscle complex consists of five heads, in
which the first two originate from the coracoid and scapula anterior and posterior to the glenoid
joint, respectively. The latter three heads are placed at the humerus. As no conditions could be
evaluated by this study in which the humeral heads contribute to static equilibrium, they were
not considered for reconstruction. Therefore, their presence can be neither confirmed nor
excluded. In crocodiles the main function of the muscles is the extension of the antebrachium,
according to the sprawling limb position. The upright posture of the limbs in sauropods, in
contrast, does not correspond to this function. A reduction of the muscles is indicated by the loss
of a prominent ulnar olecranon, which serves as insertion site of all triceps heads at the ulna in
extant reptiles and mammals. The scapular head of the triceps complex in this study has been
reconstructed to originate proximally from the caudal side of the scapula posterior to the glenoid
joint. The coracoid head is placed just anterior to the glenoid joint laterally at the coracoid. In one
limb stance they contribute to static equilibrium by either caudad (scapular head) or craniad
123
rotation (coracoid head) and caudad translation of the scapulocoraoid, as well as craniad
translation of the antebrachium. The latter function probably plays a major role in equilibrium
within the elbow joint. The triceps complex has been excluded in the study by Schwarz et al.
(2007). Borsuk-Bialynicka (1977) reconstructed only two heads, a scapular and a humeral one. The
origin and insertion of the scapular head is consistent with the present study, but the humeral
head has been placed at the humerus just below the glenoid joint. This is not supported by any
functional indication. Wilhite (2003) assumed five different heads for sauropods similar to the
condition in crocodiles. No information was given about their origins at the scapulocoracoid,
although a tendinous insertion at the posterior process of the ulna has been assumed. Remes
(2008) reconstructed the origin of only one head at the middle of the caudal scapular border. The
insertion is placed at the anterior side of the proximal humerus. Again this is in not in accordance
with the function of the muscle defined in the present study.
M. pectoralis in the present study orginates from the bony sternals and an assumed
cartilaginous sternal plate, which extends to the position of the fifth thoracic rib. M. pectoralis
contributes to retraction of the humerus during locomotion, its main function has been
reconstructed according to its lever arm. It indicates an anterior-posterior translation of the
humerus relative to the glenoid joint, and a caudad rotation of the humerus within the glenoid
joint. Although a certain amount of force is directed medially, the function as a humeral adductor
is less important in one limb stance. M. pectoralis is the main adductor in the crocodile forelimb
and has been described mainly according to this function (Borsuk-Bialynicka, 1977; Wilhite, 2003;
Remes, 2008). It was not included in the reconstruction by Schwarz et. al. (2007). Its origin at the
sternal plates and insertion at the deltopectoral crest of the humerus is consistent with the
present results (Borsuk-Bialynicka, 1977; Wilhite, 2003; Remes, 2008). An origin at the sternal
plates would lead to a different function, if the sternals were oriented in front of the ribcage
(Bialynicka, 1977; Wilhite, 2003). In such a position the force exerted by the muscle would run
medio-dorsally from the humeral insertion. Consequently the humerus would not only be
abducted, but also protracted and lifted. According to the present study such a condition would
be disfunctional and cannot be observed in any other extant tetrapod. In addition this supports a
more caudo-ventral position of the sternal plates. According to the limited dimensions of the
sternal plates the size of this muscle seems to be reduced. In an upright limb posture the
abduction of the humerus is reduced and therefore requires lower adductional forces. A
correlation between size of the sternal plates and dimension of the m. pectoralis is reasonable.
Nevertheless, the extension of the cartilaginous sternal plate is unknown, which could have
provided insertion for part of the muscle. In the present study m. pectoralis is reduced in size
124
corresponding to former investigations. It plays an important role in static equilibium as it
translates the humerus caudally.
M. pectoralis pars profundus in the only muscle assumed to be present in sauropods,
which is not described in crocodiles and is not mentioned in any other study. It was included for
calculation, because of the requirements during static loading conditions. The muscle runs from
the first thoracic ribs to the medial side of the scapula, and inserts deep above the glenoid joint.
Therefore it translates the scapulocoracoid caudo-medially, without exerting any rotational
moments. The presence of a profound part of m. pectoralis is currently known only for therian
tetrapods (Maier, 1971). Here it is termed m. pectoralis minor, which runs from the thoracic ribs
to the scapula, where it inserts deep at the glenoid joint. It has been stated before, that the
archosaur musclulature remains conservative along the evolutionary line. Nevertheless, the
apparent modification in the bony sauropod shoulder girdle justifies the former assumption based
on functional requirements. Modifications of the musculature take place whenever the habit of
locomotion has changed. The acquisition of flight in birds, for example led to alterations of the
primary reptilian musculature in order to perform the new behaviour. Investigations on migrating
cells associated with the shoulder girdle indicate that muscular attachement sites are
repositioned during evolution, thus providing new functions (McGonnel, 2001). The homology
and derivation of the assumed profound part of m. pectoralis can be disputed. The mechanical
requirements of the sauropod shouler girdle indicate that such a muscle is required in the system.
M. deltoideus coracoideum originates from the deep fossa and cranial border of the
cranially extended part of the scapula (acromial region) and inserts proximolaterally along with m.
deltoideus scapularis and m. pectoralis at the deltopectoral crest of the humerus. The line of
action indicates lateral abduction of the humerus. In one limb stance this muscle prevents a
medial translation of the scapula. The same term was used by Remes (2008). It was named in the
literature as m. deltoideus scapularis (Wihite, 2003) or m. scapulohumeralis anterior (BorsukBialynicka, 1977). The muscle was not considered for reconstruction by Schwarz et al. (2007).
Although its insertion at the humerus again is largely consistent with the present results, its
scapular origin is different. Remes (2008) assumed the origin to be exclusively at the clavicles,
which were positionend at the acromial region of the scapula. In basal sauropods no cranially
expended part of the acromial region is present. Therefore, a comparison with Diplodocus is
inadequate. A certain amount of fibers probably inserts at the clavicles in Diplodocus, which
would not alter the function described above. An expansion of the muscle over the extended
acromial region is suggested by the present data. The acromial ridge and the deep fossa of the
125
scapula were assumed as the scapular origin by Wilhite (2003), which is consistent to the present
study. The major function here would be to draw the humerus dorsally and posteriorly. A
horizontal position of the scapula was assumed in this study, leading to the described line of
action for the muscle. Even in this reconstruction the muscle overlies the glenoid joint, thus
leaving little lever arm for an effective protraction of the humerus. No information was given
about the influence of the muscle on the scapula.
The insertion of m. deltoideus scapularis was reconstructed at the deltopectoral crest
along with m. deltoideus coracoideum, while the origin is placed at the middle of the scapula
cranial to the scapular ridge. Similar to m. deltoideus clavicularis, the muscle serves as an
abductor of the humerus and reduces bending in the elongated scapula, In contrast to the former
muscles it exerts little caudal rotation to the scapula. In the literaturs it is equivalent to m. dorsalis
scapulae (Wilhite, 2003; Remes, 2008) or m. deltoideus scapularis (Borsuk-Bialynicka, 1977). The
muscle was excluded from the investigation by Schwarz et al. (2007). The condition in basal
sauropods (Remes, 2008) is different from Diplodocus, however, the presence of such a muscle
has been noted. The insertion and origin of m. deltoideus scapularis is consistent with the current
results (Borsuk-Bialynicka, 1977; Wilhite, 2003), but the horizontal position of the scapula, which
has been assumed in the investigations, results into different lines of action and function of the
muscle. In the assumed position the muscle could neither act as an abductor of the humerus nor
serve in humeral retraction, because of its small lever arm and direction of forces.
M. subscapularis in the present study originates from the medial side of the scapula and
inserts deep at the medial side of the glenoid joint. Its presence is required for static equilibrium
in one limb stance, as it exerts medially directed rotational moments to the scapula and therefore
can be termed as an antagonist of the deltoid muscles. M. subscapularis corresponds to m.
subcoracoscapularis described by Borsuk-Bialynicka (1977). The assumed origin and insertion are
consistent with the present study, whereas again the line of action and function in the muscle are
altered due to the horizontal scapula position. The placement of the muscle and shoulder girdle
position assumed by Remes (2008) are comparable to the current results. The muscle was not
described by Wilhite (2003) and was omitted from reconstruction by Schwarz et al. (2007).
In the present investigation m. scapulo-humeralis originates from the second third of the
scapula, anterior to the insertion of m. serratus superficialis and below the m. teres major. It
inserts posteriorly at the humerus below the glenoid joint. It plays a minor role for static
equilibrium, although it could serve as retractor of the humerus during locomotion. In other
studies the muscle was not mentioned (Borsuk-Bialynicka, 1977; Wilhite, 2003) or was excluded
126
from considerations (Schwarz et al., 2007). Remes (2008) assumed two different parts of m.
scapulo-humeralis to be present in basal sauropods; m. scapulo-humeralis caudalis and cranialis.
Both parts originate ventrally from m. deltoideus scapularis and insert distally at the posterior
side of the humerus. While the humeral insertion is consistent with the recent findings, the origin
is placed more ventrally than assumed by the present study. Retraction of the humerus is also
possible in this reconstruction, but to a lower degree, as the distance to the pivot of the glenoid
joint would be decreased.
M. teres major in the present investigation is assumed to originate either from the caudal
border of the last third of the scapula, superior to m. scapulo-humeralis, or posterior to the bony
ridge of the scapula, superior to m. deltoideus scapularis. The muscle was found to have no
stabilizing function in one-limb stance, but it could in addition to m. latissimus dorsi serve in
retraction of the humerus during locomotion. Remes (2008) and Schwarz et al. (2007) did not
mention or even excluded this muscle from reconstruction.
M. latissimus dorsi presumably originates from the dorsal fascia at the thoracic vertebrae
and inserts at the second third of the posterior side of the humerus. According to the long lever
arm of m. latissimus dorsi relative to the glenoid joint, it serves as a strong humeral retractor
during locomotion, but it does not contribute to shoulder girdle stability and static equilibrium in
one limb stance. Borsuk-Bialynicka (1977) and Wilhite (2003) did not figure or reconstruct m.
latissimus dorsi in their studies. Remes (2008) described a similar condition for this muscle
referring to the origin and insertion sites and the predicted function. In comparison to the present
results, a shoulder girdle stabilizing function is assumed for m. latissimus dorsi (m. dorsohumeralis
in their study) by Schwarz et al. (2007). In the investigation the muscle overlies the scapula and
inserts posteriorly to the humerus. Stability of the shoulder girdle would then be achieved by
tension of the body wall musculature, which would result in a hydraulic tension of m. latissimus
dorsi (m. dorsohumeralis) (Schwarz et al. 2007). Following the results obtained in this study no
such stabilizing function is required, as equilibrium or stability in the shoulder girdle system can
overall be maintained by means of the acting muscles determined in the present investigation.
5.3.1.3 Muscle forces and dimensions
Muscle forces and dimensions were estimated for all muscles involved in static
equilibrium in one limb stance. Largest cross sectional areas and consequently the highest forces
can be found for m. serratus superficialis. Cross sectional areas (880 cm²) and forces (17603 N) are
at least 2 or 3 times as high as was estimated for the other muscles. In comparison cross sectional
127
areas and force values for all other muscles ranged between 25 mm²-300 mm² and 500 N- 6000 N
, respectively. The discrepancy between the observed values for m. serratus superficialis and the
other shoulder girdle muscles can be explained.
From the models, cross sectional areas can only be determined approximately as further
informations are required to gain absolute values. The internal architecture of the muscles
(Alexander & Ker, 1990; Payne et al., 2005) and physiological conditions of individual muscle
fibers (Azizi et al, 2008) are assumed to have a large impact on the values of the exerted forces
and the correlation between forces and cross sectional areas.
Because of the extensive calculation process in this first attempt of a 3-D multi-body FESS
only one load case was calculated. By comparison, in the synthesis of the diplodocoid skull by
Witzel & Preuschoft (2005) four load cases, which refer to biting, lateral pulling to either side,
gravity and opening the snout, were calculated and summarized by a load case technique (Witzel
& Preuschoft, 2005). By this physiological superposition of forces, the range of motion and
function of all muscles involved can be considered. The range of motion and function in the
shoulder girdle musculature is usually not restricted to the situation in one limb stance, as shown
before. Most muscles are able to perform a minimum of two different functions. M. serratus
superficialis, therefore, is an exception. It has exclusively been described, throughout the
literature, to be the main weight bearing muscle in quadruped tetrapods. Electromyographic data
for an extant reptile shows activity in this muscle during the whole locomotor cycle (Jenkins &
Goslow, 1983). Thus, a similar restriction of muscle function is assumed for sauropods. M.
serratus superficialis is the only muscle in the shoulder girdle system, which exerts its maximal
force in one limb stance, which therefore leads to a reliable estimation of its cross sectional area.
The estimated dimensions of m. serratus superficialis were found to be placed at both the origin
and insertion sites of the skeleton in a reasonable manner. For determination of cross sectional
areas of the other shoulder girdle muscles it is recommended to calculate at least three more load
cases. They should represent different stages of the locomotor cycle and in addition one single
load step, in which only neck movements are considered, as they are assumed to be performed by
m. trapezius and m. levator scapulae. A physiological superposition of the muscle forces then
would lead to an increase of maximal muscle forces and cross sectional diameters as predicted for
m. trapezius and m. levator scapulae.
Although in former investigations functional aspects or mechanical advantages of a
muscle were considered, the definition of function remains unclear. For instance, some muscles
were supposed to stabilize a structure or not, without giving a detailed description of the
128
respective muscle function or how it can be performed. Muscle action in this study was described
regarding all degrees of freedom to comprise its function as a whole, which leads to a more
detailed understanding of the muscle action. Even retraction and protraction were defined as a
performance in all three dimensions. Stabilizing, in this context, is therefore determined as the
ability of the muscles to keep the skeletal elements in equilibrium at each point of movement by
exerting the necessary amount of force components (x,y,z-axial) in the required direction.
5.3.2 FESS method
The FESS method itself is an innovative technique, and has been applied to a number of
anatomical features in extinct and extant species (Rossmann et al. 2001; Witzel & Preuschoft,
2002, 2004, 2005; Witzel & Gössling, 2006; Witzel, 2006, 2007; Gössling et al. 2008; Moser et al.
2008), but it was formerly not applied to a 3-D FE multi-body system of the free forelimbs.
In the present thesis, the FESS method was applied to a 3-D multi body FE model of the
shoulder girdle system of Diplodocus longus. Two different limb positions in frontal view were
considered, the first with the humerus in a vertical position relative to the glenoid joint, and the
second with the humerus slightly abducted.
Using forces instead of links in the Diplodocus FE model is the major difference between
both attempts of the FESS in this thesis. The link elements are less sensitive for deformation
during calculation and therefore enable stability within the system. It was revealed by FESS of the
crocodilian FE model, that the link elements prevent the determination of absolute muscle force
and function. Link elements are widely used in FEA to simulate muscle forces, but in none of the
studies the use of link elements was discussed with regard to their mechanical properties (Huber
et al. 2005; McHenry et al., 2007; Wroe et al., 2008).
It was shown by the present approach that using forces instead of link elements allows for
a more detailed investigation. The decreased stability of the model in turn reveals a number of
technically related conditions, which were not observed in the former approach of the crocodilian
shoulder girdle. These issues include the size of each finite element, the convergence criteria for
calculation, and the value of friction defined for the contact elements, which obviously have a
great impact on stress distribution and statics in the system. Lowering the friction, refinement of
the FE model, and modification of the friction coefficient for contact elements had a visible
impact on stress distributions, but still did not lead to satisfactory results. Possibly an ongoing
refinement of the volumes in the FE model could lead to more accuracy in calculation. Finite
element number is linked to the number of nodes required in the mesh of the volume, which are
129
concurrently increased. Unfortunately, the number of nodes available in the FE Program is
limited. Finally, an increase of the convergence criteria during calculation led to functional
conditions in the FE model. Despite longer calculation times, the improved technical conditions
were adopted for further calculation.
However, the calculation of the first FE-model, in which the humerus was positioned
vertically to the glenoid joint, could not be finished satisfactorily. No physiological pattern of
compressive stress distribution within the bauraum could be achieved as the results lack
significant compressive stresses i.e. in the coracoid region. Therefore, abduction of the humerus
was proposed to increase the medially directed forces, which are necessary to maintain
compressive stresses within the coracoid. Nevertheless, the performed calculations of the first
attempt revealed necessary information and implications for the method. The basic movements
of the multi-body system resulting from forces and bearings are demonstrated and equal to the
condition in the FESS of the crocodilian shoulder girdle. Downward movements of the trunk,
resulting from weight forces, rotation of the scapulocoracoid under the influence of m. serratus
superficialis, and the impact of the counteracting muscles providing stability in the system could
be recognized.
These methodological implications were incorporated in the second FE-model. As the
method is obviously sensitive to the arrangement of forces, the forces applied to the second FE
model of Diplodocus longus were refined. Muscles like m. serratus superficialis, m. trapezius, and
m. levator scapulae are assumed to insert along the scapula. To accommodate this, total forces
are subdivided, which results into a more distributed stress pattern. This procedure also
minimizes stress peaks, which occur especially at insertions of high single forces.
Even though, calculation of the whole body model revealed high friction between the
trunk and scapulocoracoid, which was still visible and prevented the establishment of static
equilibrium. It was proposed, that the high number of contact elements between the trunk and
the scapulocoracoid increases adhesion between the adjacent areas, even if friction coefficient
was lowered. In consequence, a separate FE model, which consists of only the scapulocoracoid
bauraum, the humerus and the antebrachium, was calculated. The mechanical function of the
trunk was accomplished by bearings applied to the scapulocoracoid bauraum.
After a number of iterative steps, the equilibrium in the shoulder girdle system of the
second FE model was accomplished and results in a satisfying distribution of compressive stresses
in the scapulocoracoid bauraum. The system is kept in equilibrium only by muscle forces, which
130
were applied to the FE model in a mechanically proper arrangement according to their function in
one limb stance. The compressive stresses, which result from transmission of body weight to the
scapulocoracoid serve as the blueprint for the synthesis-part of the FESS method. In three
consecutive steps, the bauraum of the scapulocoracoid was then reduced, until physiological
loading in the structure was achieved. The completed virtual synthesis of the scapulocoracoid of
Diplodocus longus corresponds to the original fossil.
With the aid of FESS the shoulder girdle of Diplodocus longus was investigated according
to its function in one physiological loadcase, which here was defined as the transmission of body
weight forces in one limb stance. The results of the analyis of the solid tetrapod bodies and the
primary 2-D FE model of the crocodilian shoulder girdle indicate the requirement of a closed circle
of forces in the shoulder girdle. Static equilibrium in the shoulder girdle can only be maintained if
the weight forces are continuously transmitted through the elements of the shoulder girdle
system. As compressive forces are sustained by the bony elements a ventrally closed shoulder
girdle was predicted. The results in fact indicate a transmission of medially directed forces from
the glenoid to the coracoid to the sternal plates via a coracosternal joint, which was considered in
the reconstruction. This transmission of forces reveals an alternative way compared to extant
reptiles or mammals. Therefore the shoulder girdle in dinosaurs cannot be investigated only
based on functional comparison with extant tetrapods, as it was assumed in the beginning.
According to Wolff`s law, bony structures are functionally adapted to mechanical loading,
which is determined by mechanical stimuli, thus leading to the shape of bone. This condition is
utilized by the FESS method as it allows building of virtual bony structures according to
mechanical loading in a physiological loadcase (Witzel et al. 2010).
The physiological loadcase is defined as mechanical loading of bone, which results from
muscle and joint forces during everyday use. In a physiological loadcase bone is mainly subjected
to compressive stresses. The physiological compressive stress loading presumably ranges from -2
to -20 MPa, as it can be observed for vertebrate bone (Witzel & Preuschoft, 2005; Witzel, 2007).
After reaching equilibrium in the shoulder girdle of Diplodocus longus in this study, the resulting
compressive stresses in the FE model range from -0.2 to 1.8 MPa. In the following three reduction
steps the compressive stresses were increased to -2 to -9 MPa in the last reduction of the
bauraum. After reduction a maximal value of compressive stresses (-141 MPa) is present at one
discrete point at the most posterior border of the glenoid facet. This condition can be technically
described as "kantenpressung" and is referred to the geometry of the FE model. To account for
acceleration during locomotion the reduced model was then recalculated under loading
131
conditions according to two times earth acceleration. The resulting compressive stresses range
from -3 to -20 MPa, which corresponds to the formerly predicted physiological compressive stress
loading (-2 to -20 MPa) (Witzel & Preuschoft, 2005; Witzel, 2007, 2010). The synthesized
scapulocoracoid resembles the original and shows a significant reduction in the structure
compared to the initially generated bauraum. In comparison, the original scapulocoracoid is still
thinner than the synthesis. Additional reduction steps should lead to a further decrease of the
structure, which at the same time would lead to a concentration of the stressed areas. Because of
the time consuming procedure the value of the additional knowledge seems doubtful.
According to Pauwels (1965), bending stresses induce remodelling of bone, which results
into bone formation in the compressed areas until bending is minimized (Witzel et al. 2010). This
in consequence leads to the formation of a light weight structure, which is assumed to be an
advantage for energy saving strategies (Alexander, 2002; Reilly et al. 2007) especially for animals
with high body weights like sauropod dinosaurs. In fact, with regard to the size of the animal the
dimensions of the scapula are remarkably small. The scapula of Diplodocus exhibits the greatest
dimension in the lateral plane, whereas its maximal thickness ranges from approximately 10 cm at
the caudal part of the glenoid facet to only 1 cm at the cranial border of the scapula. Despite its
appearance the scapulocoracoid bone is able to support approximately 1/5 (2 tons) of the total
body weight (10 tons) in one limb stance.
In fossil remains, information about the musculature and the position of the bony
elements are rare and sometimes misleading as shown in the present thesis. In this context the
applied FESS method allows for the determination of the position and values of muscle forces, as
they correspond to the physiological compressive stress loading in the scapulocoracoid bone. This
condition makes the FESS an appropriate method for the investigation of the musculoskeletal
system, especially in extant vertebrates.
132
Chapter 6
General conclusions and future
perspectives
6.1
General conclusions
The relationship of form and function in vertebrate bone can be sufficiently investigated
by means of mechanical engineering. The form or shape of an individual bone therefore is
adapted to resist the mechanical stresses, which occur under loading conditions. It is further
concluded, that the major determinant of bone formation is compression stress and therefore
consistent with Wolff´s law.
In general, all morphological and evolutionary implications, which were drawn from the
different approaches using finite element method (FEM) in the present investigation, can be
referred to this circumstance. This includes the analysis of the stress distribution in two 3-D FE
models of solid tetrapod bodies and the crocodilian shoulder girdle, the FESS of the crocodilian
shoulder girdle and finally, the FESS of the shoulder girdle in a sauropod dinosaur (Diplodocus
longus).
The initial loading condition in a terrestrial vertebrate body is defined by the gravitational
force. Since the animal is released to gravity, the body weight force acts towards the ground,
which results into ground reaction forces. Both gravitational force and ground reaction force exert
mechanical stress on the body. This primary condition reveals the most basic requirements of the
presence and location of the musculoskeletal elements in a terrestrial vertebrate. Bones and
muscles were found to be positioned in accordance to the observed mechanical stress patterns.
The middle of the trunk is free from stresses even under loading and thus provides ample space
for the intestines. The bony elements are defined to resist the compressive stresses and the
musculature is able to intercept the tensional parts. The defined basic loading conditions describe
two different locomotor types, one sprawling and one extended limb position. They can be
distinguished by the different mechanical stress patterns reflecting the morphological structures
in “lower tetrapods” (amphibians, most saurians) and dinosaurs and mammals. Furthermore, the
results emphasize that the gradual changes of the shoulder girdle elements during early tetrapod
133
evolution are the result of an enhanced terrestrial locomotor behavior related to an increase of
gravitational loading.
FESS is based in principle on the calculation of the equilibrium of forces and moments in a
mechanical system. The determination of the boundary conditions in a FE model is the most
crucial point of the approach. The most basic condition of the shoulder girdle was determined
with the aid of morphological comparison with tetrapod vertebrates and the stress distribution in
the tetrapod body under the influence of gravitational loading.
The primary approach of the 2-D and 3-D multi-body FE models of the crocodilian
shoulder girdle offers the opportunity to evaluate, whether it is possible to transfer biological
structures to a mechanical system. The results clearly indicate that this approach is in fact
appropriate to evaluate relative movements of the skeletal elements and the function of the
muscles under a functional point of view.
In addition to the determination of basic functions of shoulder girdle musculature,
fundamental implications related to the method could be revealed. Although the results provide a
first insight into the shoulder girdle mechanics, the use of link elements for this approach could
not be approved. In consequence, directed forces instead of link elements were applied in the
final approach of the 3-D FESS of the sauropod shoulder girdle.
As the analysis of the 2-D model leads to implications for the relative movements in the
shoulder girdle system, the results of the 3-D FESS reach further. The alteration of each value,
position and direction of the forces lead to visible changes of stress distribution and static
condition, pointing out that with these iterative steps a more detailed investigation of muscle
function can be established. As forces are related to muscle function, investigations therefore
should aim to determine function in all three dimensions. The results obtained by the FESS of the
crocodilian and Diplodocus shoulder girdle lead to the further conclusion that in a functional
shoulder girdle, a closed circle of mechanical forces is required.
The relationship of form and function of bone is the central aspect of the FESS method.
With a mechanical equilibrium in the shoulder girdle system achieved, the distribution of
compressive stresses in the predefined bauraum resembles the shape of the original model. A
physiological stress distribution then could be accomplished by iterative reduction of the bauraum
according to the stress bearing areas. Although the results of the FESS of the Diplodocus longus
shoulder girdle are more developed than in case of the crocodilian shoulder girdle, both
134
approaches reveal the compressive stresses to be the major determinants of bone formation and
therefore support the primary results obtained from the solid tetrapod FE models.
Because the system is in equilibrium, the mechanical stress patterns can again be related
to the bony structures and in turn allow for the deduction of the mechanical function of the
shoulder girdle muscles. The predefined basic mechanical conditions in the tetrapod shoulder
girdle in general and the shoulder girdle of Diplodocus longus in particular.
The deductive nature of the FESS method accounts for its capacity to investigate the
mechanics of the musculoskeletal system in extant animals. Although there is a lack of
information of soft tissue in fossil remains, muscle structures and the relative position of the
elements can be reconstructed by their function.
A number of initially defined morphological and functional considerations, such as a
closed circle of forces, and the most cranially position of the shoulder girdle, in which the scapula
is lying parallel to the ribcage, were found to be present in all extant tetrapods. Other features,
such as the number of shoulder girdle elements and their relative position in the forelimb girdle
system are different. The function, which was determined as the transmission of body weight
from the trunk to the shoulder girdle, is consistent for all taxa included in this investigation.
According to the transmission of body weight, the shoulder girdle of Diplodocus longus was found
to fulfill the mechanical requirements of an extended limb position, similar to a mammalian-like
locomotor type but with the components of a modified reptile-like shoulder girdle. These results
account for a careful evaluation of comparisons with extant homologues and functional
analogues; they cannot be applied to extinct ancestors regardless the individual functional
adaptations. Nevertheless morphological comparison is necessary, and was also included in this
approach, but always with regard to the locomotor behavior between the object of interest and
the extant analogue. In general the results emphasize the highly adaptive nature of bone in order
to maintain function during the evolution of sauropod dinosaurs. In both FESS approaches the
structure of the primary generated bauraum was significantly reduced in iterative steps. The
resulting physiological loading condition of bone was established most successfully for the
scapulocoracoid of Diplodocus longus.
Under physiological conditions, bending in the scapulocoracoid synthesis was significantly
reduced. A minimization of bending in bone leads to an overall reduction of the material, which
can be observed in the synthesis and in the original fossil. It can be concluded that the maximal
function of the bone is maintained by the investment of the minimum of material. A light weight
135
construction in sauropod bones was also assumed for the vertebrae and is here shown to be
realized in the shoulder girdle skeleton as well. A minimization of energy expenditure can be
achieved by an extended position of the forelimbs and a light weight construction of the shoulder
girdle skeleton, which are therefore determined to be preconditions for an increase of body size
in sauropod dinosaurs. Light weight construction in the remaining skeletal parts has to be tested
but seems reasonable. This would be a strong argument for their success and is probably a
precondition for gigantism.
FESS is an innovative approach using the FE method, and an appropriate tool not only to
study the maximal strength of bone, but to explain the nature of bone according to its mechanical
function and to shed light on the mechanisms of evolution of the great diversity in vertebrate
bones.
6.2
Future perspectives
The present study provides ground for further investigations on the evolution of the
shoulder girdle in the dinosaurs as well as for the development of the forelimb girdle along
tetrapod evolution. As evolutionary changes in the musculoskeletal system were referred to be
locomotion related changes, its impact for the reconstruction of early tetrapod behaviour should
be investigated further on.
The formation of bone was determined to be influenced by mechanical signals as a result
of its function in the musculoskeletal system. To evaluate the relative impact on the remodelling
as well as on the modelling of bone these two pathways should be investigated by means of
mechanical, physiological and ontogenetical methods in an interdisciplinary approach. Such an
approch would clearly provide insight to the mechanisms of inidvidual bone formation and
vertebrate skeleton evolution.
136
Chapter 7
Summary
There is a lack of knowledge about the functional morphology of the shoulder girdle in
both extinct and extant tetrapods. This in particular holds true for the sauropod dinosaurs, which
were the biggest land living animals. Locomotor behaviour directly influences the biology of the
animal, and therefore provides the opportunity to investigate possible correlations between the
morphology of the forelimb girdle, its biomechanics and energy saving adaptations. The obtained
information could then provide further insight into the biology of these extinct giants, especially
with regard to the evolution of sauropod gigantism.
The current investigation is mainly based on Wolff´s law and Pauwels causal
morphogenesis, which predicts that the formation of bone is highly influenced by mechanical
stresses, as a result of mechanical loading conditions in the vertebrate body. The finite element
method (FEM) allows for the investigation of stresses and deformation in solid structures and is
applied in this thesis in two different approaches. In the first step mechanical stresses in
vertebrate tetrapod bodies were analyzed to reveal the basic functional conditions under
gravitational loading. The second approach is termed the Finite-Element Structure Synthesis
(FESS), in which the shape of the original bone can be virtually synthesized from a primary
generated bauraum according to the mechanical stresses under loading conditions. Both
approaches were applied to investigate the musculoskeletal system of extant and extinct
vertebrates under defined biomechanical conditions.
The current study starts with basic morphological and functional considerations about the
shoulder girdle of extant tetrapods to provide ground for the investigation of the sauropod
shoulder girdle.
Morphological comparison between extant lacertilians and mammals reveals a similar
relative position of the shoulder girdle in both groups. This includes a cranially positioned
shoulder girdle covering the first ribs, and a scapula, which is positioned nearly parallel to the
vertebral column in lateral and dorsal view of the ribcage. In addition, all observed reptiles exhibit
a ventrally positioned tongue-and-groove coracosternal joint, in which the sternal part acts as the
groove and the coracoid part acts as the tongue. In lacertilians both parts consist of compact bone
137
except for Basiliscus plumifrons, where the coracoid tongue is only partly ossified. In contrast, the
tongue-and-groove structure in all observed crocodilians is built of cartilage, which was examined
by observations of macerated skeletons and of histological sections.
For the investigation of the mechanical stress distribution in two different limb positions
under the influence of gravitational loading, two simplified solid 3-D FE models of tetrapod bodies
were analysed. Here, the sprawling limb position reflects the primitive condition, which is
predicted for early tetrapods and is observed in all extant lacertilians and crocodiles. The
extended limb position, in which the limbs are held under the body, refers to the situation in
extant mammalians. The analysis reveals stress patterns in regions, which represent the limb
girdles and the vertebral column and depend on the mass distribution in the body segments as
well as the position and posture of the supporting limbs. The major difference between both limb
conditions was in the orientation of the scapula, which is more inclined in an extended limb
position, and in the presence of a coracoid between both supporting forelimbs. The latter is only
required in a sprawling limb position in order to sustain the medially directed compressive forces.
Morphological observations of the shoulder girdle in extant tetrapods and the analysis of
mechanical stresses in the simplified 3-D FE models lead to the following conclusions. First, all
tetrapod bodies underlie the same basic mechanical conditions since the body is exposed to earth
acceleration, which require appropriate bony and muscular structures to sustain the occuring
stresses, which is precondition for a successful transmission of body weight from the trunk to the
supporting forelimbs. Further, the transmission of body weight and the resulting mechanical
stresses depend on the position and posture of the forelimbs, which is realized in mammals and
reptiles in different ways.
To evaluate the relative movements of the shoulder girdle elements in one-limb stance
and the occurring mechanical stress flow in the shoulder girdle region a simplified 2- D multi-body
FE model of the crocodilian shoulder girdle was built. Based on the obtained information a a
multi-body 3-D FE model of the crocodilian shoulder girdle was created and investigated with the
FESS method. The virtual synthesis of the crocodilian scapulocoracoid resembles basic features of
the original structure, like the position of the scapulocoracoid, the long slender shape of the
scapula and the connection of the coracoid to the sternal region. The results allow for the
evaluation of the FESS method, revea implications for the function and relative movements in the
shoulder girdle system, and finally lead to determine the basic shoulder girdle muscle function
during one limb stance.
138
Finally, the 3-D FESS method was applied to the shoulder girdle of a sauropod dinosaur,
Diplodocus longus. After establishing the static equilibrium in the shoulder girdle system by
means of muscle forces, the scapulocoracoid was synthesized according to the distribution of
physiological compressive stresses at the bauraum. The virtual synthesis of the scapulocoracoid
exhibits a variety of features, which can be observed in the original fossil, like e.g. the prominent
acromial region, a long and slender scapula, the overall convex shape of the scapulocoracoid and
the most compact structures at the level of the glenoid. According to the mechanical conditions in
the sauropod shoulder girdle system, which were determined by this FESS approach, a
mechanically reasonable reconstruction of the shoulder girdle of Diplodocus longus was
accomplished. This includes the position of the skeletal elements and the acting musculature. The
scapulocoracoid exhibits a lateral inclination of about 60° relative to the vertebral column. It is
positioned parallel to the ribcage, which results into a narrowed frontal aperture compared to the
original mount. This in consequence enables a direct contact of the coracoids to the sternal
plates, which were placed medio-caudally to the coracoids. Following the results of the FESS a
moderate extended posture of the forelimbs is reconstructed. This includes a slight flexion in the
shoulder and the elbow joint in lateral view, and a moderate abduction of the humerus in frontal
view. The position, function and force values of the musculature were determined according to
the requirements of static equilibrium in one limb stance. The reconstruction comprises all
muscles running between the trunk and the shoulder girdle, as well as from the shoulder girdle to
the humerus and the antebrachium.
The results reveal the adaptation of a reptile-like shoulder girdle in sauropods to an
extended limb posture. It therefore presents an alternative way of function compared to extant
mammals on the one hand and extant lacertilians and crocodiles on the other hand.
The present study indicates that mechanical stress patterns can be related to the position,
form and function of the elements in the musculoskeletal system of tetrapod vertebrates.
Furthermore, with the successful virtual synthesis of the scapulocoracoid of Diplodocus longus,
carried out by the FESS method, the close relationship of mechanical function and form in a
limbgirdle system was pointed out for the first time
In addition, bending moments in the virtual synthesis of the diplodocid scapulocoracoid
are minimized, while the structure is mainly loaded under compression. This in turn results into an
overall reduction of the material, in the virtual synthesis and in the original fossil. The
musculoskeletal system is predicted to be optimized as a light-weight construction, in which the
maximal function of bone is maintained by the investment of the minimum of material. These
139
considerations are supported by the results of the 3-D FESS of Diplodocus longus. According to the
situation in Diplodocus longus a light weight construction of the shoulder girdle skeleton is
assumed to be one precondition for an increase of body size in sauropod dinosaurs.
Finally the FE method in general and the FESS method in particular are appropriate tools
to investigate the mechanics of the musculoskeletal system in extinct and extant tetrapod
vertebrates under a functional point of view.
140
Chapter 8
Zusammenfassung
Die funktionelle Morphologie des Schultergürtels wurde bisher sowohl für fossile als auch
rezente Tetrapoden nur unzureichend untersucht. Dies gilt insbesondere für die größten jemals
lebenden Landlebewesen, die Sauropoden. Da die Fortbewegung eines Tieres einen direkten
Einfluss auf seine Lebensweise hat, bietet sich hier die Möglichkeit, die Morphologie und
Mechanik des Schultergürtels in Bezug auf geeignete Strategien zur Optimierung des
Energieverbrauchs hin zu untersuchen. Mit der vorliegenden Dissertation werden neue
Erkenntnisse in Bezug auf die Schultergürtelmechanik der Sauropoden aufgezeigt, welche dazu
beitragen die Biologie dieser fossilen Giganten zu untersuchen und mögliche Anpassungen des
Schultergürtels als Voraussetzung des sauropoden Gigantismus zu definieren.
Die Grundlage für die mechanischen Überlegungen dieser Untersuchung beruhen auf dem
Wolff`schen Transformationsgesetz und Pauwels `Kausaler Morphogenese`, welche besagen, dass
sowohl die Bildung als auch der Erhalt knöcherner Strukturen von mechanischen Spannungen
beeinflusst werden. Mit Hilfe der Finite- Element Methode (FEM) ist es möglich, mechanische
Spannungen und Verformungen in festen Körpern zu untersuchen und bildlich darzustellen. Die
FEM wurde in der vorliegenden Arbeit in zwei unterschiedlichen Ansätzen verwendet. Um den
Einfluss der Schwerkraft auf die Spannungsverteilung im Wirbeltierkörper zu untersuchen,
wurden zunächst Spannungsanalysen an vereinfachten Finite-Element Modellen vierbeiniger
Wirbeltierkörper durchgeführt. Mit Hilfe der Finiten-Element Struktur Synthese (FESS) ließen sich
anschließend virtuelle Synthesen knöcherner Strukturen auf der Basis der beobachteten
mechanischen Spannungen erzeugen. Diese Vorgehensweise erlaubte sowohl die Untersuchung
des Zusammenhangs zwischen der Form und Funktion von Knochen, als auch die Bestimmung
mechanischer Funktionen im Schultergürtel rezenter und fossiler tetrapoder Wirbeltiere.
Die Untersuchung begann mit einer Reihe funktionsmorphologischer Überlegungen zum
Schultergürtel rezenter und fossiler Tetrapoden, welche die Grundlage für die Untersuchung des
Schultergürtels von Diplodocus longus bildeten.
Im morphologischen Vergleich zwischen rezenten Lacertiliern und Säugetieren zeigen sich
große Übereinstimmungen in der relativen Position des Schultergürtels. In beiden Gruppen liegt
141
der Schultergürtel weit kopfwärts auf Höhe der ersten Rippen, und ist von dorsal gesehen parallel
zum Rippenkorb angeordnet. Der distale Rand der knöchernen Scapula beziehungsweise der
knorpeligen Suprascapula zeigt ebenfalls eine parallele Ausrichtung zur Wirbelsäule. Desweiteren
besitzen alle untersuchten Reptilien ein Coracosternalgelenk, welches aus einer sternalen
Gleitrinne und dem sich darin gleitenden Coracoid gebildet wird. Diese Nut-und Feder
Konstruktion zeigt sich bei allen untersuchten Lazertiliern vollständig verknöchert, mit Ausnahme
von Basiliskus plumifrons, bei dem der entsprechende Anteil des Coracoids teilweise knorpelig
ausgebildet ist. Im Gegensatz dazu wird bei den Krokodilen sowohl die sternale Gleitrinne als auch
der Anteil des Coracoids ausschließlich aus Knorpelgewebe gebildet, was an mazerierten
Skeletten und histologischen Querschnitten durch das Coracosternalgelenk verdeutlicht werden
konnte.
Um den Einfluss der Schwerkraft auf die Spannungsverteilung zwischen zwei
unterschiedlichen Gliedmaßenstellungen zu analysieren, wurden zwei vereinfachte 3-D FE
Tetrapodenmodelle erstellt. Die abgespreizte Gliedmaßenstellung stellt die ursprüngliche
Kondition dar, wie sie für basale Tetrapoden angenommen wird und auch bei allen rezenten
Lacertiliern und Krokodilen beobachtet werden kann. Eine gestreckte Gliedmaßenstellung
hingegen reflektiert die Situation bei rezenten Säugetieren, insbesondere großen Lauftieren. Die
größten Spannungen zeigen sich in den Regionen, die mit der Position der Gliedmaßengürtel und
der Wirbelsäule übereinstimmen. Die Position und Orientierung der Spannungsverläufe werden
von der Massenverteilung innerhalb des Körpers und der Stellung der Gliedmaßen beeinflusst. Die
größten Unterschiede zwischen beiden Gliedmaßenstellungen zeigen sich in der Orientierung der
Scapula, welche im Fall der gestreckten Gliedmaßen eine stärkere Neigung nach caudal aufweist,
und in der Präsenz eines ventral gelegenen Coracoids zwischen den Vorderextremitäten. Ein
Coracoid, welches in der Lage ist, hohe medial gerichtete Druckspannungen aufzunehmen, lässt
sich ausschließlich bei einer abgespreizten Gliedmaßenstellung annehmen.
Die morphologische Betrachtung rezenter Tetrapoden und die Spannungsanalysen
anhand der basalen 3-D FE Modelle lässt folgende Rückschlüsse zu. Unter dem Einfluss der
Schwerkraft lassen sich für alle Wirbeltierkörper grundsätzliche mechanische Belastungen
beobachten. Diese können nur durch die Ausbildung geeigneter knöcherner und muskulärer
Strukturen kompensiert werden, um somit eine Übertragung des Körpergewichts vom Rumpf auf
die unterstützenden Gliedmaßen zu gewährleisten. Die Kraftübertragung und die resultierenden
mechanischen Spannungen sind hierbei abhängig von der Stellung der Gliedmaßen, und werden
von Säugetieren als auch Reptilien in unterschiedlicher Weise realisiert.
142
Um die Bewegungen der einzelnen Elemente des Schultergürtels während des
Einbeinstandes zu bestimmen und die daraus resultierenden Spannungsverteilungen zu
analysieren,
wurde
zunächst
ein
vereinfachtes
2-D
FE
Mehrkörpermodel
eines
Krokodilschultergürtels modelliert. Basierend auf den gewonnenen Erkenntnissen wurde die
Untersuchung um ein 3-D FE Mehrkörpermodell erweitert und mit Hilfe der FESS Methode
untersucht. Die virtuelle Synthese des Scapulocoracoids zeigt in wesentlichen Merkmalen
Übereinstimmung mit dem Original. Dies betrifft die Position des Scapulocoracoids, die elongierte
Scapula und die Verbindung zwischen Coracoid und der Sternalregion. Die 3-D FESS des
Scapulocoracoids eines Krokodils erlaubt die Evaluation der verwendeten Methode und die
Bestimmung der grundlegenden mechanischen Bewegungen innerhalb des Schultergürtels unter
statischen Bedingungen im Einbeinstand.
Im letzten Teil der Dissertation wurde die FESS auf den Schultergürtel von Diplodocus
longus angewandt und ein 3-D FE Modell erstellt. Nach der Einstellung des statischen
Gleichgewichts im Einbeinstand mit Hilfe von Muskelkräften wurde der primäre Bauraum des
Scapulocoraocoids
unter
Berücksichtigung
der
beobachteten
physiologischen
Druckspannungsverteilung reduziert. Die virtuelle Synthese des Scapulocoracoids zeigt eine
Vielzahl von Merkmalen des originalen Fundes. Dies sind insbesondere die ausgeprägte
Acromialregion, eine lange schlanke Form der Scapula, die konvexe Form des Scapulocoraocoids
und die größte Knochendicke im Bereich des Glenoids.
Auf der Basis der ermittelten mechanischen Bedingungen, wurde eine Rekonstruktion des
Schultergürtels unter funktionellen Gesichtspunkten durchgeführt. Die Rekonstruktion beinhaltet
die Position der Skelettelemente und der im vorliegenden Lastfall wirkenden Muskulatur. Das
Scapulocoracoid zeigt eine laterale Neigung zur Wirbelsäule von 60° und ist parallel zum
Rippenkorb ausgerichtet, was in einer Verschmälerung der frontalen Öffnung im Vergleich zur
ursprünglichen Montage des Skeletts resultiert. Dies ermöglicht einen direkten Kontakt der
Coracoidea und der Sternalplatten, welche medio-caudal den Coracoidea positioniert werden.
Basierend auf den Ergebnissen der FESS wird für Diplodocus longus eine gestreckte
Extremitätenstellung angenommen, die jedoch in der Lateralen eine moderate Flexion im
Schulter- als auch im Ellenbogengelenk zeigt und in der Frontalen eine Abduktion des Humerus
um 20° zeigt.
Die Position und Funktion der Muskeln sowie der Muskelkräfte wurden ausschließlich
anhand der mechanischen Funktion bei der Einstellung des statischen Gleichgewichts im
Einbeinstand bestimmt. Die Rekonstruktion umfasst alle Muskeln, welche vom Rumpf zum
143
Schultergürtel, bzw. vom Schultergürtel zum Humerus und Antebrachium verlaufen. Die
Ergebnisse zeigen desweiteren eine Anpassungen des sauropoden Schultergürtels mit seiner
ursprünglichen Reptilienmorphologie an eine gestreckte Gliedmaßenstellung, welcher somit einen
funktionell alternativen Weg im Vergleich zu rezenten Reptilien und Säugtieren darstellt.
Es konnte gezeigt werden, dass mechanische Spannungen als Folge der Einwirkung von
Kräften auf den Wirbeltierkörper sowohl die Position, als auch die Form und Funktion der
Elemente des Skelettapparats reflektieren. Schließlich wurde mit Hilfe der virtuellen Synthese des
Scapulocoraocoids von Diplodocus longus erstmals der Zusammenhang zwischen mechanischen
Spannungen und der Form und Funktion am Beispiel eines Gliedmaßengürtels dargestellt. Zudem
konnten die in der Synthese des Scapulocoracoids auftretenden Biegespannungen minimiert
werden, wodurch die Struktur hauptsächlich unter Druckspannung belastet wird. Die Minimierung
von Biegespannungen gilt als die Grundvoraussetzung für die Leichtbauweise von Knochen,
welche in dieser Dissertation mit Hilfe der FESS für das Scapulocoracoid von Diplodocus longus
nachgewiesen werden konnte. Diese Optimierung im Schultergürtelskelett kann somit als eine
Präadaptation für die Größenzunahme bei sauropoden Dinosauriern interpretiert werden.
Abschließend soll festgehalten werden, dass die FE Methode und insbesondere die FESS
adäquate Untersuchungsmethoden darstellen, um den Zusammenhang zwischen Form und
Funktion innerhalb des Bewegungsapparats von rezenten und fossilen Tieren unter funktionellen
Gesichtspunkten zu untersuchen.
144
Chapter 9
Bibliography
Abel O (1910) Die Rekonstruktion des Diplodocus. Abhandlungen der zoologisch –
botanischen Gesellschaft Wien, 5 (3), 1-60.
Ahlberg PE, Milner AR (1994) The origin and early diversification of tetrapods. Nature 368,
507-514.
Ahlberg PE, Johanson Z (1998) Osteolepiforms and the ancestry of tetrapods. Nature 395,
792-794.
Ahlberg PE, Clack JA, Blom H (2005) The axial skeleton of the Devonian tetrapod
Ichthyostega. Nature 437, 137-140.
Ahlberg P, Clack J (2006) A firm step from water to land. Nature 440, 747-749.
Alexander RM (1976) Estimates of speeds of dinosaurs. Nature, 261, 129-130.
_____________(1985) Mechanics of posture and gait of some large dinosaurs. Zoological
Journal of the Linnean Society 83, 1-25.
_____________(1989) Dynamics of dinosaurs and other extinct giants. Columbia
University Press, New York, 167.
_____________(1998) All-time giants: The largest animals and their problems.
Palaeontology 41, 1231-1245.
_____________(2002) Principles of animal locomotion. Princeton University Press.
Alexander RM, Ker RF (1990). The architecture of leg muscles. In: Winters, J. M., Woo, S.
L. (Eds.), Multiple Muscle Systems. Springer, New York, 568–577.
Anderson, J. F., A. Hall-Martin, R. Njau, and A. S. Jayes. 1985. Long-bone circumference
and weight in mammals, birds and dinosaurs. Journal of Zoology, London (A)
207:53-61.
Bakker RT (1971a) The ecology of Brontosaurus. Nature, 229, 172-174.
Bakker RT (1972) Anatomical and ecological evidence of endothermy in dinosaurs.
Nature, 238, 81-85.
_________(1987) The return of the dancing dinosaurs; pp. 39-69 in Czerkas SJ, and Olson
EC (eds.), Dinosaurs Past and Present. Vol. 1. L.A.: Univ. Washington Press.
Barrett & Upchurch (2007) The evolution of feeding mechanisms in early
sauropodomorph dinosaurs. Special Papers in Palaeontology, 77, 91-112.
Benton MJ (1999). Scleromochlus taylori and the origin of dinosaurs and pterosaurs.
Philosophical Transactions of the Royal Society of London Series B Biological
Sciences, 354 (1388): 1423-1446.
Benton (2005) Vertebrate palaeontology. 3rd edition. Blackwell Publishing.
145
Biewener, A. A. (1989a): Scaling body support in mammals: limb posture and muscle
mechanics. Science, 245, 45–48.
_____________(1989b) Mammalian terrestrial locomotion and size. Mechanical design
principles define limits. BioScience, 39, 776-783.
_____________(1990) Biomechanics of mammalian terrestrial locomotion. Science, 250,
1097-1103.
_____________(1991). Musculoskeletal design in relation to body size. Journal of
Biomechanics, 24, (Supplement 1), 19-30.
Blob RW & Biewener AA (1999) In vivo locomotor strain in the hindlimb bones of alligator
mississipiensis and Iguana iguana: Implications of limb bone safety and nonsprawling limb posture. The Journal of Experimental Biology 202, 1023–1046.
_____________(2001) Mechanics of limb bone loading during terrestrial locomotion in
the green iguana (Iguana iguana) and american alligator (Alligator mississipiensis)
The Journal of Experimental Biology, 204, 1099–1122.
Bonaparte JF (1971) Los tetrapodos del sector Superior de la Formacion Los Colourados,
La Rioja, Argentina (Triásico Superior). I Parte. Opera Lilloana, 22, 1-183.
Bonnan MF (2001) The evolution and functional morphology of sauropod dinosaur
locomotion. PhD Dissertation, Northern Illinois University, 700 pp.
__________(2003) The evolution of manus shape in sauropod dinosaurs: Implications for
functional morphology, forelimb orientation, and phylogeny. Journal of Vertebrate
Paleontology, 23(3), 595-613.
_________(2004) Morphometric analysis of humerus and femur shape in Morrison
sauropods: Implications for functional morphology and paleobiology.
Paleobiology, 30 (3), 444-447.
Bonnan MF, Parrish MJ, Stevens KA, Graba J, Senter P (2005) Scapular position and
function in the Sauropodomorpha (Reptilia: Saurischia). Journal of Vertebrate
Paleontology 25, 38A.
Bonnan & Yates AM (2007) A new description of the forelimb of the basal
sauropodomorph Melanosaurus: implications for the evolution of pronation,
manus shape and quadrupedalism in sauropod dinosaurs. Special Papers in
Palaeontology 77, 157-168.
Bonnan MF & Senter P (2007) Were the basal sauropodomorph dinosaurs Plateosaurus
and Massospondylus habitual quadrupeds? Special Papers in Palaeontology, 77,
139-155.
Borsuk-Bialynicka M (1977) A new camarasaurid sauropod Opisthocoelicaudia scarzynski
gen. n., sp.,n. from the upper cretaceous of mongolia. Paleontologica Polonica 37,
5-63.
Brinkmann J. (2000) Die Muskulatur der Vorderextremität von Alligator mississippiensis
(Daudin, 1802). Diploma Thesis. University Tübingen.
146
Bryant HN & Russel AP (1992) The role of phylogenetic analysis in the inference of
unpreserved attributes of extinct taxa. Philosophical Transactions of the Royal
Society of London, Series B: Biological Sciences, 337, 405-418.
Brochu CA (2003) Osteology of Tyrannosaurus rex: Insights from a Nearly Complete
Skeleton and High-Resolution Computed Tomographic Analysis of the Skull. In:
Memoir (Society of Vertebrate Paleontology), Vol. 7, Osteology of Tyrannosaurus
rex: Insights from a Nearly Complete Skeleton and High-Resolution Computed
Tomographic Analysis of the Skull. 1-138
Brown B (1906) Tyrannosaurus, Upper Cretaceous carnivorous dinosaur". Bulletin of the
American Museum of Natural History 22 (16): 281–296.
Burness GP, Diamond J & Flannery T (2001) Dinosaurs, dragons, and dwarfs: the evolution
of maximal body size. Proceeding of the National Academy of Science
98(25):14518-14523.
Carrano MT (2000) Homoplasy and the evolution of dinosaur locomotion. Paleobiology,
26 (3), 489-512.
__________(2005) The evolution of sauropod locomotion: Morphological diversity of a
secondarily quadrupedal radiation. 229-251. In: Curry Rogers, K. and Wilson, J. A.
(eds). The Sauropods: Evolution and Paleobiology. University of California Press,
Berkeley.
Carroll RL (1987) Vertebrate paleontology and evolution. WH Freeman and Company.
Carpenter K (2002) Forelimb biomechanics of nonavian theropod dinosaurs in predation.
Senckenbergiana lethaea, 82 (1), 59-76.
Carlson KJ, Patel BA (2006) Habitual use of the primate forelimb is reflected in the
material properties of subchondral bone in the distal radius. Journal of Anatomy,
208(6), 659–670.
Carter DR, Orr TE, Fyhrie DP & Schurman DJ. (1987) Influences of mechanical stress on
prenatal and postnatal skeletal development. Clinical Orthopaedic 219:237–250.
Carter DR & Beauprè GS (2001) Skeletal function and form: mechanobiology of skeletal
development, aging and regeneration. Cambridge University Press.
Charig AJ (1972) The evolution of the archosaur pelvis and hind-limb. An explanation in
functional terms. In: K. D. Joysey and T. S. Kemp (eds.), Studies in vertebrate
evolution., 121-155 Winchester, New York.
Chiappe LM , Witmer LM (2002) Mezozoic birds: above the head of dinosaurs. University
of California Press.
Christian A (2002) Neck posture and overall body design in sauropods. Mitteilungen des .
Museums für Naturkunde Berlin Geowissenschaftliche Reihe 5, 271-281.
Christian A, Preuschoft H (1996) Deducing the body posture of extinct large vertebrates
from the shape of the vertebral column. Paleontology 39, 801-812.
Christian A, Heinrich WD, Golder W (1999 a) Posture and mechanics of the forelimbs in
Brachiosaurus brancai. Mitteilungen des . Museums für Naturkunde Berlin
Geowissenschaftliche Reihe 2, 63-73.s
147
Christian A, Müller RHG, Christian G, Preuschoft H (1999 b) Limb swinging in elephants
and giraffes as implications for the reconstruction of limb movements and speed
estimates in large dinosaurs. Mitteilungen des . Museums für Naturkunde Berlin
Geowissenschaftliche Reihe 2, 81-90.
Christiansen P (1997) Locomotion in sauropod dinosaurs. Gaia 14, 45-75.
Clack JA (2007) Devonian climate change, breathing, and the origin of the tetrapod stem
group. Integrative and Comparative biology, 1-14.
Clack JA, Coates MI (1995) Acanthostega- a primitive aquatic tetrapod? In: Arsenault M,
Lelièvre, H, Janvier P, eds. Studies on early vertebrates. Bulletin du museum
National d´Histoire naturelle 17, 373-388.
Clack JA, Bloom H, Ahlberg P (2003) New insights into the postcranial skeleton of
Ichtyostega. Journal of vertebrate morphology, 23, 41 A.
Clarke JA, Norell MA (2002) The Morphology and Phylogenetic Position of Apsaravis
ukhaana from the Late Cretaceous of Mongolia. American Museum Novitates
3387 :1-46.
Coates MI, Clack JA (1990) Polydactyly in the earliest known tetrapod limbs. Nature, 347:
66-69.
Colbert (1962) The weight of dinosaurs. Amer. Mus. Novitates 2076, American Museum
of Natural History, N.Y.
______(1989) The Triassic dinosaur Coelophysis. Museum of Northern Arizona Bulletin,
57, 1-160.
Coombs WP (1975) Dinosaur habits and habitats. Paleogeography, Paleoclimatology,
Paleoecology 17, 1-33.
___________(1978) Theoretical aspects of cursorial adaptations in dinosaurs. Quaternary
Review of Biology, 53, 393-418.
___________(1978b) Forelimb muscles of the Ankylosauria (Reptilia: Ornithischia).
Journal of Paleontology, 52, 642-657.
Cooper MR (1981) The prosauropod dinosaur Massospondylus carinatus Owen from
Zimbabwe: its biology, mode of life and phylogenetic significance. Occasional
Papers of the National Museum of Rhodesia, Series B Natural Sciences, 6 (10),
689-840.
Currey JD (1984) The mechanical adaptations of bones. Princeton University Press
________(2002): Bones. Structure and mechanics: Princeton University Press.
Curry-Rogers K, Forster CA (2001) The last of the dinosaur titans: a new sauropod from
Madagascar. Nature, 412, 530-534.
Daniel W, McHenry C (2001) Bite force to skull stress correlation - modelling the skull of
Alligator mississippiensis. In: Crocodilian Biology and Evolution.
Day JJ, Norman DB, Gale AS, Upchurch P, Powell HP (2004) A middle Jurassic trackway site
form Oxfordshire, UK. Palaeontology, 47, (2), 319–348.
148
Demes B, Qin JX, Stern JT Jr., Larson SG, Rubin CT (2001) Patterns of strain in the macaque
tibia during functional activity. American Journal of Physical Anthropology 116 (4)
257 - 265.
Dilkes DW (2000) Appendicular myology of the hadrosaurian dinosaur Maiasaura
peeblesorum from the Late Cretaceous (Campanian) of Montana. Transactions of
the Royal Society of Edinburgh Earth Sciences, 90 (2), 87-125.
Dumont ER, Piccirillo J, Grosse IR. 2005. Finite element analysis of biting behaviour and
bone stress in the facial skeletons of bats. Anatomical Record 283A:319–330.
Dzik J (2003) A beaked herbivorous archosaur with dinosaur affinities from the early Late
Triassic of Poland. Journal of Vertebrate Paleontology, 23 (3), 556-574.
Effenberger H, Witzel U, Lintner F, Rieger W (2001) Stress analysis of threaded cups.
International Orthopaedics 25, 228-235.
Ewer RF (1965) The anatomy of the thecodont reptile Euparkeria capensis Broom.
Philosophical Transactions of the Royal Society of London Series B Biological
Sciences, 248, 379-435.
Farlow (1981) Estimates of dinosaur speeds from a new trackway site in Texas. Nature,
294, 747 - 748
______(1992) Sauropod tracks and trackmakers. Integrating the ichnolgical and skeletal
records. Zubia, 89-138.
Farlow JO (1993) On the rareness of big, fierce animals: speculations about the body size,
population densities, and geographic ranges of predatory mammals and large
carnivorous dinosaurs. American Journal of Science 293A: 167-199.
Farlow JO, Smith MB & Robinson JM (1995) Body mass, bone ''strength indicator'', and
cursorial potential of Tyrannosaurus rex. Journal of Vertebrate Paleontology
15(4):713-725.
Fechner R (2005) Hindlimb osteology and myology of the dinosauromorph Lagerpeton
chanarensis. Pp. 105-106. In A. W. A. Kellner, D. D. R. Henriques, and T. Rodrigues,
eds. Boletim de Resumos, II Congresso Latino-Americano de Paleontologica des
Vertebratos, Rio de Janeiro.
_________(2006a) New information on the anatomy of basal dinosauromorphs. Tagung
der Paläontologischen Gesellschaft. Kiel.
_________(2006b) Evolution of bipedality in dinosaurs. Journal of Vertebrate
Paleontology 26(3):60A-60A.
_________(2007) Die frühe Entwicklung der Saurischia: Eine integrative Studie. In: J.
Erfurt, and C. Maul, eds. 34. Tagung des Arbeitskreises für Wirbeltierpaläontologie
der Paläontologischen Gesellschaft. Hallesches Jahrbuch für Geowissenschaften,
Freyburg/Unstrut, 55-56.
________(2009) Morphofunctional Evolution of the Pelvic Girdle and Hindlimb of
Dinosauromorpha on the Lineage to Sauropoda. Unpublished PhD thesis.
University of Munich.
Feduccia & McCrady (1991) Torrey`s morphogenesis of the vertebrates. Wiley.
149
Filla J, Redman PD (1994) Apatosaurus yahnahpin: preliminary description of a new
species of diplodocoid sauropod from the Late Jurassic Morrison Formation of
southern Wyoming, the first sauropod dinosaur found with a complete set of
„belly ribs". In: 44th Annual Field Conference, Wyoming Ecological Association
Guidebook. p 159-178.
Frost HM (1964) The laws of bone structure, The Henry Ford Hospital surgical
monographs, Thomas, Springfield.
Fürbringer M (1876) Zur vergleichenden Anatomie der Schultermuskeln - 3. Teil, Capitel
IV: Saurier und Crocodile. Gegenbaurs Morphologisches Jahrbuch, 1, 636-816.
___________(1900) Zur vergleichenden Anatomie des Brustschulterapparates und der
Schultermuskeln, IV. Teil: Reptilien. Jenaische Zeitschrift für Naturwissenschaft,
34, 215-712.
Fujiwara S, Kuwazuru O, Inuzuka N, Yoshikawa N (2009) Relationship Between Scapular
Position and Structural Strength of Rib Cage in Quadruped Animals. Journal of
Morphology, 1-11.
Galton PM (1990) Basal Sauropodomorpha - Prosauropoda. 320-344. In Weishampel DB,
Dodson P and Osmolska H (eds). The Dinosauria. University of California Press,
Berkeley.
_________(2001) The prosauropod dinosaur Plateosaurus Meyer, 1837 (Saurischia:
Sauropodomorpha; Upper Triassic). II. Notes on the referred species. Revue de
Paleobiologie 20 (2):435–502.
Galton PM, Upchurch P (2004) Prosauropoda. 232-258. In Weishampel DB, Dodson P and
Osmolska H (eds). The Dinosauria, second edition. University of California Press,
Berkeley.
Gilmore CW (1932) On a newly mounted skeleton of Diplodocus in the United States
national museum. Proceeding of the U.S. National Museums, Vol. LXXXI, 1-21.
Gomez C, David V, Peet NM, Vico L, Chenu C, Malaval L & Timothy MS (2007) Absence of
mechanical loading in utero influences bone mass and architecture but not
innervation in Myod-Myf5-deficient mice. Journal of Anatomy, 210, 259–271.
Gomani EM (2005) Sauropod Dinosaurs from the Early Cretaceous of Malawi, Africa,
Palaeontologia Electronica Vol. 8, 27A ,37.
Gower DJ & Wilkinson M (1996) Is there any consensus on basal archosaur phylogeny?
Proceedings of the Royal Society of London, B, 263, 1399-1406.
Gössling RM, Witzel U, Distler C (2008) Biomechanical investigation on the long snout in
Plateosaurus. Journal of vertebrate paleontology, 36 A.
Gunga HC (1995) New Data on the dimensions of Brachiosaurus brancai and their
physiological implications. Naturwissenschaften, 82, 190-192.
__________Kirsch K, Rittweger J, Röcker L, Clarke A, Albertz J (1999) Body size and body
volume distribution in two sauropods from the upper Jurassic of tendaguru
(Tanzania). Mitteilungen des Museums für Naturkunde Berlin,
Geowissenschaftliche Reihe. 2, 91-102.
150
__________Suthau T, Bellmann A, Stoinski S, Friedrich A, Trippel T, Kirsch K, Hellwich O
(2008) A new body mass estimation of Brachiosaurus brancai Janensch, 1914
mounted and exhibited at the Museum of Natural History (Berlin, Germany)
Hatcher JB (1901) Diplodocus (Marsh): its osteology , taxonomy and probable habits with
a restoration of the skeleton. Memoirs of the Carnegie Museum Pittsburgh, Vol. I,
1-61.
Harris JD & Dodson P (2004) A new diplodocoid sauropod dinosaur from the Upper
Jurassic Morrison Formation of Montana, USA. Acta Palaeontologica Polonica, 49
(2): 197–210.
Hay OP (1908) On the habits and the pose of sauropodus dinosaurs, especially
Diplodocus. American Naturalist N.Y. Vol. XLII, 672-681.
Henderson DM (1999) Estimating the masses and centers of mass of extinct animals by 3D mathematical slicing. Paleobiology 25 (1), 88-106
____________ (2006) Burly gaits: Centres of mass, stability, and the trackways of
sauropod dinosaurs. Journal of Vertebrate Paleontology 26 (4), 907-921.
Hert (1994) A new attempt at the interpretation of the functional architecture of the
cancellous bone. Journal of Biomechanics, 27(2), 239-242.
Hildebrand M, Goslow GE (2004) Vergleichende und Funktionelle Anatomie der
Wirbeltiere. Springer, Berlin.
_________und Gatesy (2000) Adductors, abductors, and the evolution of
archosaurlocomotion. Paleobiology, 26 (4), 734-751.
Holland WJ. (1906) The osteology of Diplodocus Marsh. Memoirs of the Carnegie
Museum, 2, 225- 278.
Holtz (1994) The phylogenetic position of the Tyrannosauridae: implications for theropod
systematics. Journal of Paleontology, 68, 1100-1117.
Horner JR, Weishampel DB, Forster CA (2004). Hadrosauridae. In: Weishampel, Dodson,
and Osmolska (eds.), The Dinosauria. 2nd edition, University of California Press,
438-463.
Huene (1926) Vollständige Osteologie eines Plateosauriden aus dem schwäbischen
Keuper. Geologische und Paläontologische Abhandlungen,15, 129-179.
______(1932) Die fossile Reptilordnung Saurischier. Ihre Entwicklung und Geschichte.
Monographie geologische Paläontologie, 1 (4), 1-163.
______(1935-1942) Die fossilen Reptilien des südamerikanischen Gondwanalandes. C. H.
Beck'sche Verlagsbuchhandlung, München
Hutchinson JR (2006) The evolution of locomotion in archosaurs. Comptes Rendus
Palevol, 5 (3-4), 519-530.
Hutchinson JR, Schwerda D, Famini D, Dale RHI, Fischer M, Kram M (2006) The locomotor
kinematics of African and Asian elephants: changes with speed and size. Journal of
Experimental Biology, 209, 3812-3827.
151
Ikegame M, Ishibashi O, Yoshizawa T, Shimomura J, Komoro T, Ozawa H, Kawashima H
(2001) Tensile Stress Induces Bone Morphogenetic Protein 4 in Preosteoblastic
and Fibroblastic Cells, Which Later Differentiate into Osteoblasts Leading to
Osteogenesis in the Mouse Calvariae in Organ Culture. Journal of bone and
mineral research. 16, 24-32.
Janis CM & Carrano MT (1992) Scaling of reproductive turnover in archosaurs and
mammals: why are large terrestrial mammals so rare. Annals of Zoology. 28:201216.
Jarvik E (1996) The Devonian tetrapod Ichthyostega. Fossils Strata, 40, 1-213.
_______(1998) Forerunners on four legs. Nature, 395, 748-749.
Jenkins FAJ, Goslow GEJ (1983) The functional anatomy of the shoulder of the savannah
monitor lizard (Varanus exanthematicus). Journal of Morphology 175, 195-216.
Jones DC, Zelditch ML, Lightfoot Peake P, German RZ (2007) The effects of muscular
dystrophy on the craniofacial shape of Mus musculus. Journal of Anatomy. 210 ,
723–730.
Johnson RE & Ostrom JH (1995) The forelimb of Torosaurus and an analysis of the posture
and gait of ceratopsian dinosaurs. In: Thomason JJ (eds). Functional Morphology in
Vertebrate Paleontology. Cambridge University Press, Cambridge. 205-218.
Kaspar et al. (2000) Dynamic cell stretching increases human osteoblast proliferation and
CICP synthesis but decreases ostocalcin synthesis and alkaline phophatase activity.
Journal of Biomechanics 33, 45-51.
Kummer B (1959) Bauprinzipien des Säugerskeletts. Thieme, Stuttgart.
_________(1962) Funktioneller Bau und funktionelle Anpassung des Knochens.
Anatomischer Anzeiger, 110, 261–293.
_________(1985) Die funktionelle Anpassung des Bewegungsapparates in der
Phylogenese der Wirbeltiere. Verhandlungen der deutschen zoologischen
Gesellschaft, 78, 23-44.
Klima (1987) Early development of the shoulder girdle and sternum in marsupials
(Mammalia: Metatheria) Advances in Antomy and Cell Biology, 109, Springer.
Kranenbarg S, Cleynenbreugel T v., Schipper H, van Leeuwen J (2005) Adaptive bone
formation in acellular vertebrae of sea bass (Dicentrarchus labrax L.) Journal of
Experimental Biology, 208, 1-10.
Larsen CS (1997) Bioarcheology. Interpreting behavior from the human skeleton.
Cambridge University Press, Cambridge.
Liebermann et al., (2003) Lieberman, D. E., Pearson, O. M., Polk, J. D., Demes, B. &
Crompton, A. W. 2003 Optimization of bone growth and remodelling in response
to loading in tapered mammalian limbs. Journal of. Experimental Biology, 206,
2135–3138.
Lightfoot P, German RZ (1998) The Effects of Muscular Dystrophy on Craniofacial Growth
in Mice: A Study of Heterochrony and Ontogenetic Allometry. Journal of Anatomy
235:1–16.
152
Lilje KE (2007) Arboreale Lokomotion beim Chamäleon (Chamaeleo calyptratus). PhD
Thesis. Friedrich-Schiller-Universität Jena.
Lockley MG, Farlow JO, Meyer CA. (1994) Brontopodus and Parabrontopodus Ichnogen.
nov. and the significance of wide- and narrow-gauge sauropod trackways. Gaia,
10, 135-146.
Lovelace DM, Hartman SA, Wahl WR (2007) Morphology of a specimen of Supersaurus
(Dinosauria, Sauropoda) from the Morrison formation of Wyoming, and a reevaluation of diplodocoid phylogeny. Arquivos do Museu Nacional, Rio de Janeiro,
65, 4, 527-544.
Main RP (2007) Ontogenetic relationships between in vivo strain environment, bone
histomorphometry and growth in the goat radius. Journal of Anatomy, 210, 272–
293.
Main RP & Biewener AA (2004) Ontogenetic patterns of limb loading, in vivo bone strains
and growth in the goat radius. The Journal of Experimental Biology, 207, 25772588.
Marsh OC (1878) Principle characters of American Jurassic dinosaurs, Part I., American
Journal of Science & arts, 16, 413-416.
Matthew WD (1901) The pose of sauropod dinosaurs. American Naturalist, Vol. XLIV, 547560.
McGowan, Chris (1994): Diatoms to Dinosaurs. The size and scale of living things.
Washington D.C.: Island Press.
McHenry CR, Clausen PD, Daniel WJT, Meers MB, Pendharkar A (2006) Biomechanics of
the rostrum in crocodilians: a comparative analysis using finite element analysis.
Anatomical Record Part A 288A:827–49.
McIntosh JS, Brett-Surman MK, Farlow JO (1997) Sauropods. In: Farlow JO, Brett- Surman
MK (eds.). The Complete Dinosaur. Bloomington: Indiana
McIntosh JS & Williams ME (1988) A new species of sauropod dinosaur,
Haplocanthosaurus delfsi sp. nov., from the Upper Jurassic Morrison Fm. of
Colorado. Kirtlandia, 43, 3-26.
Meers MB (2003) Crocodilian Forelimb Musculature and Its Relevance to Archosauria.
Anatomical Record Part A, 274, 891–916.
Michelis I (2004) Taphonomie des Howe Quarry’s (Morrison-Formation, Oberer Jura),
Bighorn County, Wyoming, USA. unpublished PhD thesis. University of Bonn.
Moser K, Rauhut O, Witzel U (2008) Morphology and function of the neck of Plateosaurus,
Journal of vertebrate paleontology, 36 A.
Möser M & Hein W (1992) The bone as a compression member in a cable tensioning
device: The example of the hip. Wolff`s law and connective tissue regulation.
Regling G (ed.), Walter de Gruyter, Berlin New York, 81-92.
Müller GB, Streicher J (1989) Ontogeny of the syndesmosis tibiofibularis and the evolution
of the bird hindlimb: a caenogenetic feature triggers phenotypic novelty.
Anatomical Embryology 179:327–339.
153
Newman SA, Müller GB (2005) Origination and Innovation in the Vertebrate Limb
Skeleton: An Epigenetic Perspective. Journal of Experimental Zoology (Molecules,
Development, Evolution) 304 B:593–609.
Nicholls EL & Russell AP (1985) Structure and function of the pectoral girdle and forelimb
of Struthiomimus altus (Theropoda: Ornithomimidae). Palaeontology, 28, 643-677.
Nigg, Benno M.; Herzog, Walter (Hg.) (2007): Biomechanics of the musculoskeletalsystem.
3. Aufl. Chichester (GB): John Wiley & Sons Inc.
Norman, DB, Witmer LM & Weishampel DB (2004) Basal Ornithischier. 325-334. In:
Weishampel DB, Dodson P & Osmólska H, eds. The Dinosauria. University Press of
California, Berkeley.
Orr, C. M. 2005 Knuckle-walking anteater: a convergence test of adaptation for
purported knuckle-walking features of African Hominidae. Am. J. Phys. Anthropol.
128, 639–658.
Osborn HF (1899) A skeleton of Diplodocus. Mem. American Museum of Natural History
N.Y. Vol. I, 168-214.
Osmolska H (Eds.) The Dinosauria. University of California Press 259-322.
Patel BA, Carlson KJ (2007) Bone density spatial patterns in the distal radius reflect
habitual hand postures adopted by quadrupedal primates. Journal of Human
Evolution 52 (2): 130-141.
Parish JM (1986) Locomotor adaptations in the hindlimb and pelvis of the Thecodontia.
Hunteria, 1, 1-35.
Parish JM & STEVENS K A (2002). Rib angulation, scapular position, and body profiles in
sauropod dinosaurs. Journal of Vertebrate Paleontology, 22 (3), 95A.
Paul GS (1987) The science and art of restoring the life appearance of dinosaurs and their
relatives: a rigorous how-to guide. In: Czerkas SJ & Olson EC (eds.), Dinosaurs past
and present. Vo. II, Nature Historical Museum L.A., Univ. Washington Press.
Pauwels F (1965) Gesammelte Abhandlungen zur funktionellen Anatomie des
Bewegungsapparates. Berlin: Springer.
Peterson JA (1973) Adaptation for arboreal locomotion in the shoulder region of lizards.
Ph.D.Thesis, University of Chicago.
Polk JD, Blumenfeld J & Ahluwalia K (2008) Knee posture predicted from subchondral
apparent density in the distal femur: an experimental validation. Anat. Rec. 291,
293–302.
Pontzer H (2007): Effective limb length and the scaling of locomotor cost in terrestrial
animals. In: J Exp Biol, Jg. 210, H. 10, S. 1752–1761.
Preuschoft H (1970) Functional anatomy of the lower extremity. In: G. H. Bourne. The
Chimpanzee, Volume 3. 221-294. Karger, Basel.
Preuschoft H & Christian A (1999) Statik und Dynamik bei Tetrapoden. - In: U. Gansloßer
(Hrsg.): Spitzenleistungen: 89-130. - Filander Verlag, Fürth.
154
Preuschoft H, Gudo M (2005) Der Schultergürtel von Wirbeltieren: Biomechanische
Überlegungen zu Bauprinzipien des Wirbeltierkörpers und zur Fortbewegung von
Tetrapoden. Zentralblatt für Geologie und Paläontologie, II., 339-361.
Preuschoft H, Schmidt M (2003) The influence of three dimensional movements of the
forelimb on the shape of the thorax and its importance to erect body posture.
Courier Forschungsinstitut Senckenberg 243: 9-24.
Preuschoft H, Witte H, Witzel U. (2002) Pneumatized spaces, sinuses and spongy bones in
the skull of primates. Anthropologischer Anzeiger, 60, .
Preuschoft H & Witzel U (2002) Biomechanical investigation of the skulls in reptiles and
mammals, Senckenbergiana lethaea, Concepts of Engineering and Functional Morphology
______________(2004) Functional structures of the skull in hominoidea. Folia
Primatologica 75, 219-259.
______________(2005) The functional shape of the skull in vertebrates: Which forces
determine morphology? With a comparison of lower primates and ancestral
synapsids. Pp. 402-413. In C. Ross, ed. Anatomical Record, Special Issue: Finite
Element Analysis in Vertebrate Biomechanics, Part A.
Preuschoft H, Witzel U, Hohn B, Schulte D, Distler C (2007) Biomechanics of locomotion
and body structure in varanids with special emphasis on the forelimb.
Mertensiella, Suppl. Salamandra, Advances in monitor research III. …
Preuschoft H, Witte H, Christian A & Recknagel ST (1994) Körpergestalt und Lokomotion
bei großen Säugetieren. Body shape and locomotion in large (cursorial) mammals.
Verband der Deutschen Zoologischen Gesellschaft 87(2):147-163.
Rauhut OW (2003) The interrelationships and evolution of basal theropod dinosaurs.
Special Papers in Palaeontology, 69, 1-213.
Rayfield EJ (2004) Cranial mechanics and feeding in Tyrannosaurus rex. Proceedings of the
Royal Society London B 271, 1451–59.
_________(2005) Aspects of comparative cranial mechanics in the theropod dinosaurs
Coelophysis, Allosaurus and Tyrannosaurus. Zoological Journal of the Linnean
Society 144, 309–16.
_________(2007) Finite Element Analysis and Understanding the Biomechanics and
Evolution of Living and Fossil Organisms. Annual Review of Earth Planetary Science
35:541–76.
Remes K (2006) Revision of the Tendaguru sauropos dinosaurs Torniera africana
(Fraas)and its relevance for sauropod paleobiography. Journal of Vertebrate
Paleontology 26, 651–669.
Remes K (2007) Evolution of the pectoral girdle and forelimb in sauropodomorpha
(Dinosauria, Saurischier). Osteology, myology and function. PhD Thesis, Universität
München.
Ren, Lei; Butler, Melanie; Miller, Charlotte; Paxton, Heather; Schwerda, Delf; Fischer,
Martin S.; Hutchinson, John R. (2008): The movements of limb segments and joints
155
during locomotion in African and Asian elephants. In: Journal of Experimental
Biology, 211, 2735–2751.
Reilly SM& Elias AJ (1998) Locomotion in Alligator mississipiensis: Kinematic effect of
speed and posture and their relavanvce to the sprawling to errect paradigm,
Jounrnal of experimental Biology, 201, 2559-2574.
Reilly SM, McElroy EJ & Biknevicius AR (2007) Posture, gait and the ecological relevance of
energy-saving mechanisms in tetrapods. Zoology 110, 271-289.
Romer AS (1966) Vertebrate paleontology. 3rd edition, University of Chicago Press.
_________(1956) Osteology of the Reptiles. University of Chicago Press, Chicago.
Romer AS, Parsons TS (1977) The vertebrate body. WB Saunders, Philadelphia.
Rose PJ ( 2007) A New Titanosauriform Sauropod (Dinosauria: Saurischia) from the Early
Cretaceous of Central Texas and its Phylogenetic Relationships. Palaeontologia
Electronica Vol. 10, 8A, 65.
Ross CF, Patel BA, Slice DE, Strait DS & Dechow PC (2005) Modelling masticatory muscle
force in finite element analysis: sensitivity analysis using principal coordinates
analysis. Anatomical Record 283A:288–99
Rossmann T, Witzel U, Welmann J (2001) Biomechanics of the skull in proterosuchus.
Journal of morphology. 248,
Roux W (1893) Beiträge zur Entwickelungsmechanik des Embryo : Nr 7 ; Über
Mosaikarbeit und neuere Entwickelungshypothesen. S. 279-333.
Runestad JA, Ruff CB, Nieh JC, Thorington JR, and Teaford MF (1999) Radiografic
Estimation of Long Bone Cross-Sectional Geometric Properties. American Journal
of Physical Anthropology, 90, 207-213.
Santa Luca AP (1980) The postcranial skeleton of Heterodontosaurus tucki (Reptilia,
Ornithischia) from the Stormberg of South Africa. Annals of the South Africa
Museum, 79, 159-211.
Sato M, Ochi T, Nakase T, Hirota S, Kitamura Y, Nomura S, Yasui N (1999) Mechanical
tension-stress induces expression of bone morphogenetic protein (BMP)-2 and
BMP-4, but not BMP-6, BMP-7, and GDF-5 mRNA, during distraction osteogenesis.
Journal of Bone and Mineral Research, 14, 1084–1095.
Schwarz D, Frey E, Meyer CA (2007) Novel reconstruction of the orientation of the
pectoral girdle in Sauropods. Anatomical Record, 290, 32-47.
Seebacher F (2001). A new method to calculate allometric length-mass relationships of
dinosaurs. Journal of Vertebrate Paleontology 21:51-60.
P. Senter (2007): Analysis of forelimb function in basal ceratopsians. Journal of Zoology,
Jg. 273, H. 3, S. 305–314.
Sereno PC (1991) Basal archosaurs: phylogenetic relationships and functional
implications. Society of Vertebrate Paleontology Memoirs, 2, 1-53.
______(1993) The pectoral girdle and forelimb of the basal theropod Herrerasaurus
ischigualastensis. Journal of Vertebrate Paleontology, 13, 425-450.
156
_______(1999)The evolution of dinosaurs. Science, 284, 2137-47.
Shubin NH, Daeschler EB, Jenkins Jr. FA (2006) On the pectoral fin of Tiktaalik roseae.
Nature 440, 764-771.
Starck D (1979) Vergleichende Anatomie der Wirbeltiere. Band II. Das Skelettsystem.
Springer. Berlin, Heidelberg, New York.
Stevens Parrish (2005) Digital reconstructions of sauropod dinosaurs and implications for
feeding. 178-200. In Curry Rogers K & Wilson JA (eds). The Sauropods: Evolution
and Paleobiology. University of California Press, Berkeley.
Sues, H. D. 1997. European Dinosaur Hunters. In J. O. Farlow and M. K. Brett-Surman
Sverdlova N, Witzel U (2010) Principles of determination and verification of muscle forces
in the human musculoskeletal system: muscle forces to minimise bending stress. Journal
of Biomechanics, 43, 387-396
Swinton WE (1970) The dinosaurs. G. Allen & Unwin Ltd., London.
Takahashi I, Nuckolls GH, Takahashi K, Tanaka O, Semba I, Dashner R, Shum L, Slavkin HC.
(1998) Compressive force promotes sox9, type II collagen and aggrecan and
inhibits IL-1beta expression resulting in chondrogenesis in mouse embryonic limb
bud mesenchymal cells. Journal of Cellular Science, 111, 2067–2076.
Therrien F & Henderson DM (2007) My theropod is bigger than your…or not: estimating
body size from skull length in theropods 27:108-125.
Thulborn (1990) Dinosaur tracks. Chapmann & Hall, London.
Trinkaus E, Churchill SE, and Ruff B. (1994) Postcranial Robusticity in Homo. II: Humeral
Bilateral Asymmetry and Bone Plasticity. American Journal of Physical
Anthropology 93:1-34.
Tornier T (1908) Wie war der Diplodocus carnegii wirklich gebaut? Sitzung der Berliner
Gesellschaft der naturforschenden Freude Berlin, 4, 193-209.
Upchurch P, Barrett PM, Dodson P (2004) Sauropoda. In: Weishampel DB, Dodson P,
Wedel MJ, Cifelli RL & Sanders RK (2003) Sauroposeidon proteles, a new sauropod from
the Early Cretaceous of Oklahoma. Journal of Vertebrate Paleontology 20(1):109114.
_________(2003) The evolution of vertebral pneumaticity in sauropod dinosaurs. Journal
of Vertebrate Palaeontology 23:324-328.
Wilson JA, Sereno PS (1998) Early evolution and higher-level phylogeny of sauropod
dinosaurs. Journal of Vertebrate Paleontology Memoirs 5, 18:1-68.
Wilson JA & Carrano MT (1999) Titanosaurs and the origin of "wide-gauge" trackways: A
biomechanical and systematic perspective on sauropod locomotion. Paleobiology,
25 (2), 252-267.
Witmer LM (1995) The extant phylogenetic bracket and the importance of reconstructing
soft tissues in fossils. P. 277. In J.J. Thomason, ed. Functional morphology in
vertebrate paleontology. Cambridge University Press, Cambridge.
157
Witzel U (1985) Zur Biomechanik der ossären Schafteinbettung von Hüftendoprothesen,
5. Symposium über isoelastische RM-Hüftprothesen, Mathys, CH.
___________(1996) Mechanische Integration von Schraubpfannen. G. Thieme Verlag,
Stuttgart.
___________(2000) Die knöcherne Integration von Schraubpfannen. In: Perka C & Zippel
H (Eds.) Pfannenrevisionseingriffe nach Hüft-TEP. Standards und Alternativen.
Reinbeck: Einhorn-Press, 11-16.
___________(2006) Virtual synthesis of the skull in Neanderthals by FESS. 150 Years of
Neanderthal Discoveries. Terra Nostra.
___________(2007) A case study of Camarasaurus as an illustrative example for a new
method for virtual synthesis of skulls in vertebrates. Hallesches Jahrbuch für
Geowissenschaften, 23:73-78.
Witzel U & Preuschoft H (2002) Function dependent shape characteristics of the human
skull. Anthroplogischer Anzeiger. 60, .
____________(2004) Finite Element model construction for the virtual synthesis of the
skulls in primates. Journal of morphology. 260, .
____________(2005) Finite-Element Model Construction for the Virtual Synthesis of the
Skulls in Vertebrates: Case Study of Diplodocus. Anatomical Record 283A:391–401.
Witzel U, Preuschoft H, Sick, H (2004) The role of the zygomatic arch for the statics of the
skull and its adaptive shape. Folia primatologica. 75, 202-218.
Witzel U., Goessling R.M. (2006) Deductive Virtual Synthesis of a Sauropod skull, e.g.
Camarasaurus by FESS. Journal of vertebrate paleontology.
Witzel U, Mannhardt J, Goessling RM, Micheli d.P, Preuschoft H (2010) in press. Finite
element analyses and virtual syntheses of biological structures and their
application to sauropod skulls. In: Biology of the Sauropod Dinosaurs:
Understanding the life of giants. Klein N, Remes K, Gee C (eds.) James Farlow,
Indiana University Press.
Weishampel DB, Dodson P & Osmolska H (eds.) (1990) The dinosauria. University of
California Press, Berkeley.
Wilhite (2003) Biomechanical reconstruction of the appendicular skeleton of three north
american Jurassic sauropods. PhD thesis. University of Alabama at Birmingham.
Wolff J (1892) Das Gesetz der Transformation der Knochen. Hirschwald.
Wollfson DM (1950) Scapula shape and muscle function, with special reference to the
vertebral border. American Journal of Physical Anthropology 8:331-341.
Wroe D, HuberR, Lowry M, McHenry C, Moreno K, Clausen P, Ferrara TL, Cunningham E,
Dean MN, Summers AP (2008) 3-D computer analysis of white shark jaw
mechanics: how hard can a great white bite. Journal of Zoology, 1469-7998.
Yates AM & Vasconcelos CC (2005) Furcula like clavicles in the prosauropod
Massospondylus, Journal of vertebrate morphology, 25, 466-468.
158
Young CC (1941a) A complete osteology of Lufengosaurus huenei Young (gen. et sp. nov.)
from the Lufeng, Yunnan, China. Palaeontologica Sinica, Series C 7:1-53.
________(1942) Yunnanosaurus huangi Young (gen. et sp. nov.), a new prosauropoda
from the Red Beds at Lufeng, Yunnan. Bulletin of the Geological Society of China
22:63-104.
Zhang YH & Yang ZL (1994) A new complete osteology of Prosauropoda in Lufeng Basin,
Yunnan, China: Jingshanosaurus]. Yunnan Publishing House of Science and
Technology, Kunming: 1-100. (In Chinese).
Zienkiewicz OC, Taylor RL, Zhu JZ (2005) The Finite Element Method: Its Basis and
Fundamentals. Elsevier Butterworth-Heinemann, Amsterdam.
Webb GW & Gans C (1982) Galloping in Crocodylus johnstoni- a reflection of terrestrial
activity? Records of the Australian Museum 34: 607-618.
Zug GR (1974) Crocodilian galloping: an unique gait for reptiles. Copeia: 550-552.
159
Chapter 10
Supplement
10.1 Laboratory manual for histological sections of the
coracosternal joint in Caiman spec.
Dehydration and embedding

Ethanol 70%, as long as deserved, minimum 24h; Change 2 times

Ethanol 95% 24h; Change 2 times

Isopropanol 24h; Change 2 times; last 2h at 60° in a heated chamber.

Mix the warm Isopropanol 1:1 with Paraplast

Leave overnight Isopropanol will vaporize.

2h Paraplast 1 at 60°

2h Paraplast 2

2-4h Paraplast 3

Embedding in fresh Paraplast and cool down.

Deparaffinise

After drying and before dying, paraffin should be removed from the sections.

10 min Rotihistol I at 45°

10 min Rotihistol II at 45°

10 min Isopropanol

10 min Ethanol 95%

10 min Ethanol 70%

4 times 2 min. distilled water (change with high frequency)

Dyeing

Dyeing process

Sections were dyed using Aldehydfuchsin Goldner (AFG) solution, resulting in red-green
dyeing for bony tissue and green-violet dyeing for cartilaginous tissue.

Deparaffinise and rinse with distilled water

min. oxidation preparation until sections turn brown

Reductional preparation until section is colourless

Rinse with water.
160

times in Ethanol 70%

Leave 5 min. in Aldehydfuchsin

2 times in Ethanol 70%

2 times in distilled water

2-5 min. Hämatooxylin-Weigert

10 min. under flowing water

10 min. Säurefuchsin-Ponceau-Azophloxin

times vinegar acid 1% (change frequently)

30 sec. Orange G (moving)

3 times vinegar acid 1% (change frequently

min. in Lightgreen

3 times vinegar acid 1% (short)

2 times Isopropanol

Leave in Isopropanol

Cover the sections in Euparal
161
10.2 Muscles of the crocodilian shoulder girdle
Table 3 Muscles of the shoulder girdle in crocodiles. Description of the anatomical position, origin and insertion sites and the muscle function during locomotion are described (Brinkmann 2000;
Meers 2002).
Muscle
Position
Origin
Insertion
M. serratus
superficialis
(Fürbringer 1902)
Deep at the posterior part of the
shoulder flank
Last cervical rib and processus uncinati of the first
thoracic ribs.
Caudal border
scapula.
the
Transmitting
body
weight from the trunk to
the shoulder girdle.
M.
serratus
profundus
(Fürbringer 1876)
Completely covered by the
scapula
and
the
adjacent
musculature, between scapula
and trunk.
Ribs of the last four cervical vertebrae.
Dorso-medial part of the
scapular head.
Supports transmission of
body weight.
M. trapezius
(Buttmann 1826)
Superficial at the anterior part of
the neck and thorax.
From the thoraco-dorsal fascia ventro-lateral along
the surface of the trunk.
Along the cranial edge of
the scapula, dorsal to the
acromion.
Rotates the scapula
cranially and assists
during protraction.
M. levator scapulae
(Fürbringer 1876)
Superficial at the lateral side of
the neck, completely covered by
m. trapezius.
First two cervical ribs and processi transversi of the
rd
th
3 -6 cervical ribs.
Almost along the whole
cranial margin of the
scapula.
Cranial rotator of the
scapula.
M. rhomboideus
(Fürbringer 1876)
Deep under the distal end of the
scapula and suprascapula.
Dorso-lateral to the thoracic fascia at the last cervical
and first thoracic ribs.
Medial distal part of the
scapular head and part of
the suprascapula; ventral
to m. serratus profundus.
Pulls
the
medially.
162
Function
of
scapula
Muscle
Position
Origin
Insertion
Function
M. costo-coracoideus
(Fürbringer 1976)
Deep to m. pectoralis, between
coracoid plate and rib cage.
At the last cervical rib and the first lower thoracic
ribs.
Dorsal area and caudal
margin of the coracoid
plate.
Pulls
the
coracoid
caudally relative to the
sternum. Acts as a
caudal rotator of the
coracoid.
M. deltoideus
clavicularis (Romer
1977)
Lateral part of the scapula and
upper humerus, running over the
glenoid joint.
at the ventral two third of the frontal part of the
scapula.
At the processus lateralis
humeri and the most
proximal third of the
humerus shaft.
Protracts the humerus
and
stabilises
the
glenoid joint.
M. deltoideus
scapularis
(Fürbringer 1876)
Positioned below m. trapezius at
the lateral side of the scapula
Lateral surface of the scapula, cranial to m. teres
major.
at the proximal processus
lateralis humeri.
Stabiliser of the glenoid
joint, adductor of the
humerus.
M. subscapularis
(Fürbringer 1876)
Deep at the medial side of the
scapula.
Originates fleshy from the second third part of the
medial side of the scapula, whereby it covers most of
the cranial to caudal area.
Has a tendinous insertion
at processus medialis
humeri.
Adductor
of
humerus draws
scapula medially.
M. scapulo-humeralis
(Fürbringer 1900)
Deep muscle, above the glenoid
joint.
Takes it origin from the last third of the caudal ridge
of the scapula.
Dorsal side of the most
proximal end of the
humerus and glenoid joint
capsule.
Draws the scapulacoracoid caudally.
Mm. supracoracoidea
(Fürbringer 1876)
Muscle complex divided into three
distinct
parts
(M.
supracoracoideus
longus,
intermedius and brevis) cover
nearly the whole area of the
coracoid, deep at the glenoid
joint.
Cranial edge and medial surface of the coracoid near
to the glenoid joint. M. supracoracoideus
intermedius: Lateral side at the apex of the most
ventral part of the scapular body cranial to the
glenoid joint; M. supracoracoideus brevis runs from
below m. supracoracoideus to the most dorsal part of
the coracoid body.
Insertion:
portions
processus
humeri.
Function:
Primary
protractor
of
the
humerus,
secondly
acting as an adductor of
the humerus.
163
All
three
insert
at
lateralis
the
the
Muscle
Position
Origin
Insertion
Function
M. coracobrachialis
brevis dorsalis
(Fürbringer 1876)
Deep at the glenoid joint.
Dorso-lateral cranial portion of the scapula, ventral
to the acromion and caudal to the m.
supracoracoideus brevis.
Cranio-ventral aspect of
the joint capsule and the
deltopectoral crest.
Stabilizer of the humerus
head; may assists in
protraction and flexion
of the humerus.
Mm. triceps brachii
(Starck 1978-82)
Large complex consists of five
muscle independent heads, lying
superficial at the dorsal side of the
humerus drawing over the
olecranon to the ulna.
s
All three parts insert in a
common tendon at the
olecranon of the ulna.
Flexors of the humerus
as well as extensors of
the
antebrachium,
serving as stabilisers of
the glenoid joint.
M. triceps caput
scapulare
See above
Lower portion of the lower third of the scapula,
dorsal to the glenoid joint.
See above
See above
M. triceps caput
coracoideum
See above
Consists of a tendinous arc. First originates from the
caudal edge of the scapula, dorsal to the glenoid rim.
Second from the caudal border of the coracoid,
immediately dorsal to m. costo-coracoideus.
See above
See above
Mm. triceps caput
humerale lateralis,
medialis & posticum
Position: Surrounding the middle
and distal part of the humerus
shaft.
Humerus shaft
See above
Serves as extensors of
the antebrachium.
M. brachialis
(Fürbringer 1876)
Superficial at the ventral side of
the humerus and the elbow joint.
Cranio-dorsal surface of the proximal humerus
distal to the deltopectoral crest.
At the most proximal
end of ulna and radius,
just below the elbow
joint.
Strong flexor of the
antebrachium.
M. spiralis
(Fürbringer, 1876)
On the ventral side of the elbow
joint, running between humerus
and antebrachium.
At the ventral side of the most distal part of the
humerus.
Proximal
radius.
Flexor
of
antebrachium.
164
end
of
the
the
M. teres major
(Fürbringer 1876)
Superficial at the shoulder blade.
Caudal to m. deltoideus scapularis at the lateral side
of the shoulder blade.
165
In a common tendon with
m. latissimus dorsi dorsolateral at the proximal
end of the humerus.
Serves together with m.
latissimus dorsi as an
elevator of the humerus
and flexor of the glenoid
joint.
10.3 3-D FESS of a crocodilian shoulder girdle
10.3.1 Adjustment of static equilibrium
1. M. serratus superficialis transmits the body weight forces from the trunk to the
shoulder girdle; m. pectoralis prevents the humerus from sliding aside. The trunk is free to move
up and downwards, the scapulocoracoid is restricted by applied bearings in z and x direction. Both
the humerus and the antebrachium exhibit fixed-bearings preventing any movement. The
indicator bearings applied to the scapulocoracoid show reaction forces indicating a caudal
rotation and downward movements, which are prevented by pushing the humerus into the
glenoid joint. The humerus is tending to rotate cranially within the ulnar joint, pushing the
antebrachium caudally. Humerus is exhibiting high bending forces in the shaft. For pretension and
reaction forces see Table 4.
Figure 59 3-D model in the initial situation in frontal (above) and lateral view (below). In addition to the initial body
weight force (red arrows) m. serratus superficialis and m. pectoralis are applied to the system.
166
2. Scapulocoracoid is only restricted in the cranio-caudally direction, by bearings at the
right distal edge and left proximal corner, able to move in vertical directions. Bearings in x,y,zdirection are applied to humerus at the distal end, while the proximal part is now free from
restrictions, thus allow free movement in the glenoid joint and rotation in the elbow joint. Applied
muscles in this step are m. trapezius and levator scapulae acting in opposite direction to m.
serratus superior rotating the scapulocoracoid cranially. Further applied muscles are m.
rhomboideus, exerting forces medial directed forces towards the trunk, and m. costo-coracoideus
to support the m. serratus and m. trapezius in cranial rotation of the scapula. The trunk element
and scapulocoracoid again move downwards, while bending in the humerus shaft is reduced.
Reaction forces in the scapulocoracoid bearings still indicate a caudalward rotation. Note the
increase of maximum stress value compared to the first calculation step. Distribution of stresses
has modified, resulting from the reduction of applied bearings. Stresses concentrate mainly at the
coracoid position and at the caudal edge of the scapulocoracoid. For pretension and reaction
forces see Table 5.
167
Figure 60 3-D model of the object in frontal (above) and lateral view (below). Muscles added in this step are m.
trapezius, m. levator scapulae, m. costo-coracoideus and m. rhomboideus. The bearing at the left distal end of the
scapulocoracoid was removed.
3. In the following step, distribution of the applied links and restriction of movements is
modified. M. deltoideus scapularis und clavicularis are applied to the system. The scapulocoracoid
is restricted by one bearing, which is positioned at the proximo-cranial corner. M. costocoracoideus prevents caudal rotation of the scapula by pulling the coracoid region backward and
counter wise rotation. Reaction force in the scapular bearing indicates a shift from caudal to
cranial rotation, which indicates that m. costo-coracoideus overcompensates the movement of
the scapulocoracoid. M. deltoideus act in the opposite direction, but has no visible impact on the
results. Total stress value is decreases in comparison to the former calculation step. With regard
168
to the stress distribution, no major modifications can be observed. For pretension and reaction
forces see Table 6
Figure 61 One bearing at the distal corner of the scapula was removed. M. deltoideus scapularis and clavicularis are
placed in the system to counter cranial rotation of the scapulocoracoid.
4. In the present calculation m. triceps caput coracoideum, caput scapulare, caput
humerale and m. supracoracoideus were applied to the FE model, in order to counter the cranial
rotation of the scapula. To support this action, m. costo-coracoideus was removed. Reaction
forces value in the bearing at the scapulocoracoid is increased. The additionally applied muscles
are not sufficient to keep the system in equilibrium. The maximum stress value is increases, while
the overall distribution of stresses is not affected. Reaction forces within the bearings applied to
the antebrachium element and the distal part of the humerus indicate posterior movement of the
antebrachium. To investigate static equilibrium in the elbow joint as well, the model would
require the entire set of muscles and structures reaching from the limb to the forefoot. This is not
169
in the focus of this work. Distribution of stresses within the antebrachium or humerus and the
equilibrium within the ulnar joint will not be taken into account. Muscle related links and
pretension of link elements can be obtained from Table 7.
Figure 62 Boundary conditions are equal to the former step. Here m. triceps caput coracoideum, caput scapulare and m.
supracoracoideus were added to the model.
170
10.3.2 Reduction of the bauraum
10.3.2.1
First reduction step
st
Figure 63 1 reduction step of the FE model in lateral (A) and frontal (C) view. Single reduction of the scapulocoracoid in
lateral (B) and frontal (D) view. Boundary conditions (links, bearings and weight forces) are equal to the final calculation
step. Insertion sites of the links is maintained. Reduction of the initial bauraum is most present in lateral view. In frontal
view the outlines of the model are more rounded than in the original bauraum. The overall structure is still massive.
Figure 64 Calculation of the 1st reduction step with all applied bearings and link elements, which equal to the boundary
condition in the final calculation step. The distribution of compressive stresses in lateral (A) and frontal (C) view.
Compressive stress pattern in the single reduced bauraum of the scapulocoracoid, in lateral (B) and frontal (D) view.
Black outlines in (B) and (D) indicate the selected contours for the next reduction step, according to the selected stress
bearing areas.
171
Figure 65 Anterio-posterior cross sections of the first reduction step. Sections are at the same level compared to the
final calculation. Selected frontal cross sections show thecompressive stressed regions of the scapulocoracoid bauraum.
Compressive stresses spread from the glenoid joint medio-dorsally and are undistrubed throughout the bauraum.
10.3.2.2
Second reduction step
Figure 66 Second reduction step of the FE model in lateral (A) and frontal (C) view. Single bauraum, which refers to the
scapulocoracoid in lateral (B) and frontal (D) view. Reduction at this point is showing a heavy decrease of the primary
structure compared to the first reduction. The shape of the bauraum is in approximation to the original
scapulocoracoid. The outlines of the bauraum are still edged.
172
10.3.3 Technical settings
Table 4 Applied links for the actual calculation step, which refer to muscles or muscle groups in crocodiles. Link
diameter and pretension of the link element. F x,y,z are reaction forces, detected at the bearings after calculation. The
first three bearings are the right and left distal bearings, 21921 the proximal bearing at the scapula, respectively.
No.
Muscle
Link diameter
(mm)
Pretension
(mm)
NODE
FX
1a
M. pectoralis major
0,1
0,01
21074
13.627
-9.905
1b
M. pectoralis minor
0,1
0,01
21132
15.574
12.372
2
M. rhomboideus
21921
12.023
-0.86006E01
3a
M. serratus
superficialis 1
0,1
0,01
24336
20.593
-35.828
3b
M. serratus
superficialis 2
0,1
0,01
3c
M. serratus
superficialis 3
0,1
0,01
3d
M. serratus
superficialis 4
0,1
0,01
VALUE
49.553
4
M. serratus profundus
5
M. subscapularis
6
M. coracobrachialis
brevis
7
M. scapulo-humeralis
8
M. biceps brachii
9
M. brachialis
10
M. latissimus dorsi
11
M. teres major
12
M. trapezius
13
M. levator scapulae
14
M. deltoideus
clavicularis
173
FY
0.0000
FZ
-32.632
15
M. deltoideus
scapularis
16
M. triceps caput
scapulare
17
M. triceps caput
coracoideum
18
M. triceps caput
humerale
19
M. costocoracoideus
20
M. spiralis
21
M. supracoracoideus
Table 5 Applied links for the actual calculation step, which refer to muscles or muscle groups in crocodiles. Link
diameter and pretension of the link element. F x,y,z are reaction forces, detected at the bearings after calculation. Node
number 21074 is the distal bearing at the scapula, node 21921 is the proximal bearing.
No.
Muscle
Link
diameter
(mm)
Pretension
(mm)
NODE
1a
M. pectoralis major
0,1
0,01
21074
-117.47
1b
M. pectoralis minor
0,1
0,01
21921
-24.933
2
M. rhomboideus
0,1
0,01
49327
-61.131
18.213
3a
M. serratus
superficialis 1
0,1
0,01
49328
30.040
11.062
3b
M. serratus
superficialis 2
0,1
0,01
49815
35.367
-22.453
3c
M. serratus
superficialis 3
0,1
0,01
72120
-59.656
-97.608
53.934
3d
M. serratus
superficialis 4
0,1
0,01
72128
-80.453
14.425
-62.266
4
M. serratus
profundus
72212
-22.452
0.45032
-0.12602E01
5
M. subscapularis
72220
-11.304
22.276
10.653
6
M. coracobrachialis
brevis
76204
-31.395
80.842
140.51
174
FX
FY
FZ
7
M. scapulo-humeralis
8
M. biceps brachii
9
M. brachialis
10
M. latissimus dorsi
11
M. teres major
12
M. trapezius
0,1
0,01
13
M. levator scapulae
0,1
0,01
14
M. deltoideus
clavicularis
15
M. deltoideus
scapularis
16
M. triceps caput
scapulare
17
M. triceps caput
coracoideum
18
M. triceps caput
humerale
19
M. costocoracoideus
0,1
0,01
20
M. spiralis
21
M. supracoracoideus
76533
-0.33880
TOTAL
VALUES
0.71294E05
-73.323
-14.737
100.00
0.47307E05
Table 6 Applied links for the actual calculation step, which refer to muscles or muscle groups in crocodiles. Link
diameter and pretension of the link element. F x,y,z are reaction forces, detected at the bearings after calculation.
No.
Muskel
Link diameter
(mm)
Pretension
(mm)
NODE
1a
M. pectoralis major
0,1
0,01
21074
1b
M. pectoralis minor
0,1
0,01
2
M. rhomboideus
0,1
0,01
TOTAL
VALUES
3a
M. serratus
superficialis 1
0,1
0,01
VALUE
0.0
175
FX
FY
FZ
113.22
0
0.0000
113.22
3b
M. serratus
superficialis 2
0,1
0,01
3c
M. serratus
superficialis 3
0,1
0,01
3d
M. serratus
superficialis 4
0,1
0,01
4
M. serratus profundus
0,1
0,01
5
M. subscapularis
6
M. coracobrachialis
brevis
7
M. scapulo-humeralis
8
M. biceps brachii
9
M. brachialis
10
M. latissimus dorsi
11
M. teres major
12
M. trapezius
0,1
0,01
13
M. levator scapulae
0,1
0,01
14
M. deltoideus
clavicularis
0,1
0,01
15
M. deltoideus
scapularis
0,1
0,01
16
M. triceps caput
scapulare
17
M. triceps caput
coracoideum
18
M. triceps caput
humerale
19
M. costocoracoideus
0,1
0,01
20
M. spiralis
21
M. supracoracoideus
176
Table 7 Applied links for the actual calculation step, which refer to muscles or muscle groups in crocodiles. Link
diameter and pretension of the link element. F x,y,z are reaction forces, detected at the bearings after calculation.
No.
Muskel
Link diameter
(mm)
Pretension
(mm)
NODE
1a
M. pectoralis major
0,1
0,01
21074
1b
M. pectoralis minor
0,1
0,01
2
M. rhomboideus
0,1
0,01
TOTAL
V
ALUES
3a
M. serratus
superficialis 1
0,1
0,01
VALUE
0.0
3b
M. serratus
superficialis 2
0,1
0,01
3c
M. serratus
superficialis 3
0,1
0,01
3d
M. serratus
superficialis 4
0,1
0,01
4
M. serratus profundus
0,1
0,01
5
M. subscapularis
6
M. coracobrachialis
brevis
7
M. scapulo-humeralis
8
M. biceps brachii
9
M. brachialis
10
M. latissimus dorsi
11
M. teres major
12
M. trapezius
0,1
0,01
13
M. levator scapulae
0,1
0.01
14
M. deltoideus
clavicularis
15
M. deltoideus
scapularis
0,1
0.01
177
FX
FY
FZ
130.53
0
0.0000
130.53
16
M. triceps caput
scapulare
0,1
0.01
17
M. triceps caput
coracoideum
0,1
0.01
18
M. triceps caput
humerale
19
M. costocoracoideus
20
M. spiralis
21
M. supracoracoideus
0,1
0.01
Table 8 Link diameters, pretension, acting forces and stresses within the link elements according to the shoulder girdle
musculature in the final calculation step.
No.
Muskel
M. pectoralis major
Link diameter
(mm)
0,1
Pretension
(mm)
0.01
Force
(N)
10
Link stress (N/mm²) after
calculation
150
1a
1b
M. pectoralis minor
0,1
0.01
10
129
2
M. rhomboideus
Not defined
3a
M. serratus
superficialis 1
M. serratus
superficialis 2
M. serratus
superficialis 3
M. serratus
superficialis 4
M. serratus
profundus
M. subscapularis
0,1
0.01
10
137
0,1
0.01
10
196
0,1
0.01
10
290
0,1
0.01
10
315
0,1
0.01
10
0
0,1
0.01
10
92
Not defined
8
M. coracobrachialis
brevis
M. scapulohumeralis
M. biceps brachii
0,1
0.01
10
130
9
M. brachialis
0,1
0.01
10
169
10
M. latissimus dorsi
Not defined
11
M. teres major
Not defined
12
M. trapezius
1
0.01
100
69
13
M. levator scapulae
0,1
0.01
10
0
14
M. deltoideus
clavicularis
0,1
0.01
10
157
3b
3c
3d
4
5
6
7
Not defined
178
15
16
17
18
19
20
21
M. deltoideus
scapularis
M. triceps caput
scapulare
M. triceps caput
coracoideum
M. triceps caput
humerale
M.
costocoracoideus
M. spiralis
0,1
0.01
10
70
0,1
0.01
10
128
0,2
0.01
20
135
0,1
0.01
10
205
1
0.02
200
147
0,1
0.01
10
0
M.
supracoracoideus
Antebrachium
0,1
0.01
10
196
0,1
0,01
10
803
179
10.4 3-D FESS of the shoulder girdle in Diplodocus longus
10.4.1 Adjustment of static equilibrium in forelimb position I.
In the following, the outlines of the FE-model in frontal and lateral view demonstrate the
applied forces representing the shoulder girdle muscles (above). The contour plot in frontal and
lateral view shows the distribution and magnitude of compressive stresses after calculation; black
outlines indicate deformation after calculation showing the initial situation. Contact elements are
applied to the adjacent areas of the trunk and the scapula, at the region of the assumed
coracosternal joint and at the glenoid and elbow joint. Values of muscle forces applied to the
model and reaction forces within the bearings can be obtained from the supplement. Forces are
shown as arrows. The coordinate system is placed right at the bottom. Compressive stress values
can be obtained from the contour legend (colour coded) or from the heading legend (maximum
and minimum). High compressive stresses are blue, low stresses are red. Areas with stresses
beyond the spectrum are grey. The elements of the shoulder girdle system can be held in position
by first, bearings (indicated as triangles) and second, by forces preventing the movement.
Therefore, two settings with different boundary conditions can be determined. The first is the
initial situation, whereas the trunk is restricted in x- and z-direction to prevent medial and
forward movements. In living animals, this function is adopted by the right supporting forelimb
and hindlimbs, respectively. A fixed bearing restricts the antebrachium. The scapulocoracoid and
the humerus require indicator bearings from which reaction forces can be read-out. These
indicator bearings should be near zero if the elements are balanced out and provide stability in
non-balanced calculations. They will be removed iteratively, leading to the second setting in
which the trunk and antebrachium are allowed to move in physiological directions. The
scapulocoracoid and the humerus are free from restrictions and held in position only by the
applied forces representing the shoulder girdle muscles. To determine the function of the
required forces (muscles), they are applied to the model in iterative steps, considering their
impact to the system, which again can be detected at the indicator bearings. Changes concerning
the boundary conditions (applied forces and bearings) will be described in the figure legends.
1. Basic movements of the multi-body model can be shown. The trunk is sinking
downwards, while the scapulocoracoid is tending to rotate caudally around the pivot of the
glenoid joint under the influence of the forces transmitted by m. serratus superficialis.
Concurrently the scapulocoracoid is sinking downwards and is translated in cranialward direction,
while the humerus rotates around the pivot of the elbow joint cranially as well. Stress distribution
180
within the scapulocoracoid shows a non-physiological pattern. The antebrachium is restricted for
movements in all directions. For boundary conditions see Table 9.
Figure 67 Distribution and deformation in the FE model after calculation. Applied muscles are m. serratus
superficialis, m. serratus profundus and m. trapezius. Bearings are applied to the distal end of the
scapulocoracoid in z-direction and to both ends of the humerus in x,y,z-direction.
2. Equilibrium between the scapulocoracoid and the trunk is reached. The
scapulocoracoid is held in position only by the applied forces, while sinking downwards but shows
no rotational movements. The trunk is sinking downwards and the humerus is subject to bending.
Stress pattern is non-physiological, spreading from the glenoid joint to the caudal edge of the
scapula. Transmission of forces can be detected between the coracoid and sternal elements and
at the glenoid joint. In the next step, the equilibrium should be reached between scapulocoracoid
and humerus by removing bearings from the humerus. For boundary conditions see Table 10.
181
Figure 68 Show the results after calculation, while reaching equilibrium without any restrictions at the
scapulocoracoid. Muscles added to the model are, m. levator scapulae, m. costo-coracoideus and m.
rhomboideus. No bearings at the scapulocoracoid are present.
182
3. The present figure is showing the calculated model adding m. triceps caput
coracoideum and caput scapulae in order to attain balance of the humerus by pushing it into the
glenoid joint. Balance in the scapulocoracoid should not be affected, as both part of the triceps
complex act as functional antagonists, producing the same rotational moments referring to their
line of action. One indicator bearing is applied to the humerus shaft in z-direction, not affecting
the balance of the scapulocoracoid. Reaction force is high, indicating a strong forward direction,
even with the actual applied forces in this step. Relative movement of the humerus and
antebrachium to the medial side is shown. Stress distribution within the scapulocoracoid is
slightly modified compared to former calculation steps. For boundary conditions see Table 11.
Figure 69 Bearings are removed from the distal and proximal humerus joints, leaving one z-directed bearing at the
humerus shaft. Further m. triceps caput coracoideum and caput scapulae were added to the model.
183
4. As the former calculation did not prevent the humerus from forward movement, the m.
pectoralis complex is added to the model, instead of mm. triceps. One bearing is again placed at
the distal end of the scapula, as without any calculation can be realized. Reaction force at the
bearing is high pointing out that addition of m. pectoralis has a destabilizing effect to the
scapulocoracoid in this arrangement. The humerus is free from restrictions, but shows high
bending in medial direction, as a result of applied m. pectoralis. Stress distribution shows a
decrease of compressive stresses in the cranial part of the scapulocoracoid and is therefore not
physiological. For boundary conditions see Table 12.
Figure 70 Position of bearings equals to the former step. M. pectoralis is applied to the model instead of m. triceps
caput scapulare and caput coracoideum.
5. In a number of iterative steps equilibrium was reached. No bearings are required to
keep scapula and humerus in balance. Maximum compressive stress value is -25 MPa.
184
Compressive stresses spread from the glenoid joint to the caudal border of the scapula, to the
coracoid and sternal element. There still is no fully accordance to the original. Stress distribution
within the acromial region has to be enhanced. In a next step modification of stress distribution is
aimed due to repositioning of m. serratus superficialis. For boundary conditions see Table 13.
Figure 71 No bearing are applied to the scapulocoracoid and humerus. Equilibrium is maintained only by means of
forces. All muscles of the shoulder girdle system are applied, except for m. latissimus and m. teres major.
6. Boundary conditions in this calculation step are equal to the final calculation step
except for m. serratus superficialis (Table 13).The upper picture shows the origin of m. serratus
superficialis at the most caudal ribs. In the lower picture m. serratus superficialis is shifted
cranially. As rotational moments result from the forces, they should increase in the upper
configuration, because of the greater lever arm they exhibit in this arrangement. Concurrently the
stability of equilibrium within the model is not affected by these modifications. The distribution of
stresses within the scapulocoracoid bauraum cannot be modified essentially.
185
Figure 72 Two different calculation steps of the FE model. In the above shown calculation m. serratus
superficialis originates most caudally from the trunk; In the calculation below the origin of m. serratus
superficialis is placed more cranially.
7. The concentration of compressive stresses at the caudal border of the scapula superior
to the glenoid joint is a result of strong bending moments at this area. To reduce bending in the
scapulocoracoid, the forces values of m. trapezius and m. triceps scapularis are increased. Note
the unusual deformation of the humerus shaft. Although strong forces are applied to diminish
bending in the scapula, no reduction can be observed. Instead a heavy increase of compressive
stresses at the proximal part of the scapulocoracoid can be observed. Further the scapula was
aimed to rotate cranially, as a result only -y-directed translation can be detected.
186
Figure 73 Boundary conditions equal to the former calculations. The forces values of m. trapezius and m. triceps
scapulae were increased.
8. The meshed scapulocoracoid bauraum and the humerus are refined to reduce canting
in between these elements. This is not leading to a sufficient result, although stress distribution is
approximating to the predicted distribution. The humerus exhibit great deformation in the
proximal section, resulting from entrainment by the scapulocoracoid during movement. Minor
rotation is visible around the restricted distal scapular edge. Further, the equilibrium is affected
and no longer maintained.
To enable calculation one bearing is applied to the distal edge of the bauraum. Lowering
the friction coefficient from 1 to 0.1, diminish the unusual deformation in the humerus shaft, but
neither rotation of the scapulocoracoid is visible nor free rotation within the glenoid joint is
achieved. It can be concluded that a refinement of the elements and an increased friction do not
account for free rotational movements between areas provided with contact elements. Boundary
conditions can be obtained from Table 13.
187
Figure 74 Ground model with refined elements, calculated under the same boundary conditions as in Fig. 57 (A).
Calculation of (A) shows alteration in stress distribution and equilibrium in the system, while the humerus shows no
undisturbed movement within the glenoid joint (B). Friction coefficient for contact elements is lowered from .1 to 0.1,
again no detecTable changes towards the desired status in the glenoid region were achieved (C).
188
10.4.2 Adjustment of static equilibrium in forelimb position II.
The model shows a slight abduction of the humerus, whereas all other parts refer to the
former defined model. Friction coefficient for contact elements has been kept. Convergence
criteria for calculation is 0.001. The refined model is calculated under the same boundary
conditions as in forelimb position I. Again adjusting of static equilibrium starts by forces running
between the trunk and the scapulocoracoid. The scapulocoracoid is restricted to translational
movements by one single indicator bearing at the distal corner. The humerus is restricted by two
fixed bearings applied to both distal and proximal end of the shaft. Out of this initial situation,
equilibrium is reached in iterative steps.
1. M. serratus superficialis here is the initial force rotating the scapulocoracoid caudally.
M. trapezius and m. levator scapulae are counteracting the rotation in cranial direction, but at the
same time adding high z-directional forces to the system, moving the scapulocoracoid cranially.
M. costo-coracoideus produces z-directional forces into the opposite direction, but not to a great
extent. Forces value of m. serratus profundus is here first calculated in order to transfer the +y
dues of the acting vector forces resulting from m. trapezius and m.levator scapulae. In the
following calculation steps forces value of m. serratus profundus will be calculated out of the
summation of all y-dues of the acting vector forces running between the trunk and the
scapulocoracoid, in order to keep the sum of forces=0, which is a precondition for equilibrium.
Indicator bearing at the distal end of the scapula shows a moderate cranial rotation. Sum of ydirceted forces within the bearing located at the glenoid is near the initial forces value of about
18956N, denotes that the sum of weight force is actually completely transferred to this point.
Distribution of compressive stresses is showing high values within the former aimed regions, like
the acromial region, the scapula and within the coracoid region. For boundary conditions and
reaction forces see Table 14.
Compared to the first FE model, the forces in the present steps were subdivided and
placed along the predicted insertion sites at the scapulocoracoid to obtain a more distributed
compressive stress pattern and to minimize stress peaks, which occur especially at insertions of
high forces, while the aabsolute force values are kept constant. Calculation of the initial situation
resulted in an increase of reaction force at the z-directed bearing at the distal edge of the
scapulocoracoid, thus indicating strong forward translation, which results into imbalance of the
trunk and the scapulocoracoid. Although forces and rotational moments were calculated
precisely, the trunk was entraining the scapulocoraocid along its direction. Friction, even with a
189
lowered friction coefficient of about 0.1 is beyond tolerable limits, because of the large number of
contact elements between the trunk and scapulocoracoid.
Figure 75 First calculation of the refined model. The applied force were reduced to an initial situaltion. Single forces are
subdivided and distributed along the assumed insertion sites. For boundary conditions see Table 14.
2. To avoid misleading results from uncertain reactions between the trunk and
scapulocoracoid, the trunk was removed, and the scapulocoracoid along with the humerus and
the antebrachium were calculated separately. The focus in this investigation is on stress
distributions within the scapulocoracoid element. Areas of the trunk adjacent to the
scapulocoraoid preventing medial directed movement were replaced by adequate bearings to
fulfil the function. Assumed regions of main force transmission are defined at the predicted
190
coracosternal joint and the acromial region, where the claviculae presumably contact the scapula
and where forces can be transmitted. Further applied bearing should serve as indicator bearing
and to provide stability until equilibrium was reached. Forces directed to the trunk remain their
line of action positioned towards the trunk. (for reaction forces and boundary condition see Table
15).
Figure 76 Calculation without the trunk element under the same boundary conditions as in the former steps, which
includes the trunk (Table 15). Lateral movements of the scapulocoracoid, which formerly were prevented by the
adjacent trunk element are transferred by applied x-dirceted bearings. The bearing at the distal end of the
scapulocoracoid shows indicates low translational movements of the scapulocoracoid.
3. In the next calculations, views of the medial and caudal side are presented in addition
to the lateral and medial view to gain necessary information. In the present calculation m.
serratus superficialis, m. serratus profundus, m trapezius, m. levator scapulae and m. costocoracoideus are applied to the system. They run between the scapulocoracoid and the trunk,
while m. triceps caput scapulare and m. triceps caput coracoideum run between the antebrachium
and the scapulocoracoid. Medially directed movements of the scapulocoracoid are inhibited by
application of x-directed bearings at the coracoid and the acromial region. To support stability
one x-directed bearing is placed at the caudal edge near the glenoid joint. Indicator bearing in zdirection is applied to the distal edge of the scapulocoracoid. The humerus is restricted in all
191
directions at both the proximal and distal end. One fixed bearing is placed at the antebrachium
fixed bearing to inhibit any movement. For reaction forces and forces values see Table 16.
Figure 77 M. serratus superficialis, m. serratus profundus, m trapezius, m. levator scapulae and m. costo-coracoideus are
applied to the FE model. M. triceps caput scapulare and caput coracoideum run between the antebrachium and the
scapulocoracoid. One x-directed bearing is placed at the coracoid and the acromial region. One x-directed bearing is
placed at the caudal edge near the glenoid joint. One indicator bearing in z-direction is applied to the distal edge of the
scapulocoracoid. The humerus is restricted in all directions at both the proximal and distal end. A fixed bearing is placed
at the antebrachium.
4. No restriction is required at the distal end of the scapula, because z- directed reaction
forces within this bearing are low. Reaction force within the bearing at the glenoid joint indicates
translation of the scapulocoracoid in cranial direction, which cannot be diminished by the applied
forces of m. supracoracoideus, m. biceps brachii, m. coraco-brachialis brevis, m. deltoideus pars
scapularis et clavicularis, m. subscapularis and m. scapulo-humeralis. Another force is required to
192
counter the high z-directed forces of m. trapezius and m. levator scapulae. Compressive stresses
are increasing at the glenoid region, resulting from m. deltoideus clavicularis. High stress
magnitudes can be observed at the medial side of the scapula as result of medial bending. For
reaction forces and boundary conditions see Table 17.
Figure 78 Forces added to the system are m. supracoracoideus, m. biceps brachii, m. coraco-brachialis brevis, m.
deltoideus pars scapularis et clavicularis, m. subscapularis and m. scapulo-humeralis. No bearing are placed at thedistal
end of the scapulocoraoid. Medial bending of the scapulocoracoid is indicated by black outlines
5. In order to counteract the cranialward translation of the scapulocoracoid, m. pectoralis
profundus is applied to the system, originating from the ventral side of the trunk inserting profund
at the medial side of the glenoid joint. To eliminate influence of further forces, m. triceps
scapulare, m. deltoideus scapularis, m. deltoideus clavicularis, m. subscapularis and m. scapulo193
humeralis are temporary removed. Reaction force within the bearing at the distal end of the
scapula is increased, while reaction force value of the bearing located at the glenoid is highly
decreased, showing the great benefit of m. pectoralis profundus in preventing cranialward
translation. For reaction forces and boundary conditions see Table 18.
Figure 79 M. triceps scapulare, m. deltoideus scapularis, m. deltoideus clavicularis, m. subscapularis and m. scapulohumeralis are temporary removed. M. pectoralis profundus was applied to the system to counter caudally directed
translational movements.
6. M. triceps scapulare, m. deltoideus clavicularis and m. subscapularis are applied to the
system including an enhanced m. pectoralis major. M. levator scapulae has been divided into a
superior and inferior portion for a refinement of the resulting stresses. The occuring reaction
forces within the proximal bearing of the scapulocoracoid indicate high translational forces in
194
directed caudally, thus pushing the humerus into the glenoid joint. Although reaction forces at the
proximal part as well as at the distal end are not eliminated, calculation could be accomplished
without any restriction at the scapulocoracoid and proximal end of the humerus. For reaction
forces values and boundary conditions for this calculation see Table 19.
Figure 80 M. triceps scapulare, m. deltoideus clavicularis and m. subscapularis are applied to the system including an
enhanced m. pectoralis major. M. levator scapulae has been divided into a superior and inferior portion for a
refinement of the resulting stresses.
7. Medially directed movements of the scapulocoracoid are indicated at the bearing,
which is located at the caudal border of the scapula. In order to minimize these deformations, m.
costo-coracoideus pars profundus is applied. It originates from the ventral surface of the coracoid
to the medial part of the sternal region. Further m. deltoideus scapularis and m. deltoideus
clavicularis are slightly enhanced. Medial directed forces are now flowing medially from the
195
coracoid to the sternal region, while the indicator bearing at the caudal border is released (see
Table 20). Stress distribution has not been modified essentially.
Figure 81 M. costo-coracoideus pars profundus is applied to the system. It originates from the ventral surface of the
coracoid to the medial part of the sternal region. Further forces for m. deltoideus scapularis and m. deltoideus
clavicularis are increased.
8. Moderate modifications are executed in the present calculation step. Forces value in m.
trapezius and m. levator scapulae are decreased. The calculation is accomplished without any
restrictions at the scapulocoracoid and proximal humerus. The stress distribution in lateral view is
sufficient, while the medial side is still affected by bending and shows high stresses and an
unphysiological stress pattern. For boundary conditions and reaction forces see Table 21.
196
Figure 82 Forces value in m. trapezius and m. levator scapulae has been decreased. No restrictions at the
scapulocoracoid and proximal humerus are necessary for calculation.
9. In order to modify the occurring compressive stresses m. serratus superficialis is
positioned more laterally at the scapula compared to the former steps (see frontal view). The
boundary conditions and reaction forces can be obtained from Table 22. As a result the
compressive stress flow is modified. The compressive stresses are lowered at the medial side and
concurrently increased at the lateral side. The static equilibrium is not affected, but the
197
caudalmost bearing at the scapula shows highly increased reaction forces, which indicate a
medialward direction at this location.
Figure 83 M. serratus superficialis is positioned more laterally at the scapula (frontal view) compared to the former
steps.
The boundary conditions are consistent. M. serratus superficialis are shifted to the middle
of the scapulocoracoid bauraum. M. deltoideus clavicularis, m. deltoideus scapularis and m.
subscapularis are repositioned. Their vector forces are directed more laterally. This results into
higher laterally x-directional components of vector force in m. deltoideus clavicularis and m.
deltoideus scapularis and medially x-directed components of force in m. subscapularis. The stress
198
distribution is modified. The medial side shows a decline of compressive stresses. In lateral view
stresses are within the acromial region, the coracoid part and the scapula. The reaction forces are
low at the indicator bearing, which is positioned at the distal end of the scapula. The calculation
could be accomplished in addition without any restrictions at the scapulocorcoid and the proximal
end of the humerus (not shown here). For reaction forces and boundary conditons see Table 23.
Figure 84 M. serratus superficialis has been shifted to the middle of the scapulocoracoid bauraum. The boundary
conditions are consistent. The medial side of the scapulocoracoid shows a decrease of bending , indicated by a
reduction of compressive stresses.
10.4.3 Reduction of the “bauraum”
After the first reduction the new bauraum still displays only the rough dimensions of the
original, but is yet clearly reduced compared to the former bauraum. The bauraum shows
compressive stresses ate the prominent acromial region at the cranial border of the
scapulocoraoid. The scapula is more slender and the coracoid region has been rounded compared
199
to the original bauraum. A reduction of the material can also be viewed in frontal view as the
breadth of the scapula is reduced. The contour legend is set to range from 0 to -2.7, which
indicate an increase of compressive stresses in the bauraum. Maximum compressive stress
increase (-111 N), but can only be detected at a sinlge location at the caudal border of the glenoid
joint. The internal distribution of compressive stresses is shown in frontal cross sections. The
stresses flow mainly undisturbed through the structure of the bauraum (Fig.).
Fi
gure 85 Pictures A-D FE model after the first reduction step, recalculated under the same boundary
conditions as in the final calculation step.
200
Figure 86 Figure A-H are showing cross sections at the same location as in the final calculation step. The stressed areas
are accumulating in the bauraum, while the overall structure is reduced. This leads to a thinning of the former bauraum.
201
In the second reduction step the main features of the scapulocoracoid again are visible,
whereas the whole structure is refined compared to the original bauraum. The element is now
exhibiting a more slender appearance, with a medially dished scapula. The acromial region of the
scapula is clearly visible. At the distal end of the scapula the stresses are accumulating at the
medial side. The contour plot is set to range from 0 to-4.5. The main part of the reduced bauraum
now shows compressive stresses, which are within the range of physiological stress value (-2 to 20 MPa). Again, the maximum of compressive stresses is increased (-280 N), but still present at
one distinct location at the caudal border of the glenoid joint. For a more detailed examination
see cross sections of this reduction step (Fig. 82)
Figure 87 FE model after the second reduction step. Compressive stress distribution after recalculation under constant
boundary conditions.
202
Figure 88 Frontal cross sections of the second reduction step after calculation. Again sections were taken at the same
position as in figure 71. Distribution of stresses along the bauraum are in accordance to the original. Features do not
alter, but shows an ongoing refinement of the structure.
203
After the third and last reduction step the thinning of the whole structure is clearly visible
and most pronounced at the most cranial border of the scapula. The prominent acromial region is
present and in accordance to the orginal. The reduced bauraum is medially concave (Fig. B). The
most compact structures can be obsereved in the coracoid at the position of the assumed
coracosternal joint and dorso-caudally to the glenoid joint.
Figure 89 Figure FE model after the third reduction step and before recalculation. Lateral view (A), frontal view (B),
medial view (C), back view (D). The forces required under statical loading are present (arrows). The scapulocoracoid and
humerus are free from restrictions. Bearings at the position of the coracosternal joint are inidicated as triangles.
In the recalculation of the third and last reduction step the stressed regions are still in
accordance to the original and exhibit physiological stress values in the structure,which range
from -2 to -9 MPa. The amount of areas with show stresses below the physiological value is
reduced. The compressive stresses concentrate within the aimed regions of the scapulocoraoid. A
further reduction step could lead to a more defined structures, but is not considered at this point
204
of investigation. No additional information are expected. Contour legend has been set ranging
from 0 to -9. Maximum compressive stress value again can be detected only at one distinct area
at
the
caudal
border
of
the
glenoid
joint.
Figure 90 FE model after the third reduction step, recalculated under the former boundary conditions. is pointing out
the stressed regions in accordance to the original, exhibit mostly physiological stress value in the structure. Contour
legend has been set to -9. Maximum compressive stress value again can be detected at the caudal border of the glenoid
joint and at the coracosternal joint.
205
In order to simulate acceleration under locomotor related conditions, the third reduction
step was recalculated with all forces scaled two times as high as in the initial load case, according
to 2 times earth acceleration. Compressive stress values in the relevant part of the structure are
between -2 and -20 MPa.
Figure 91 Calculation of the 3rd reduction step according to 2 times earth acceleration. The amount of compressive
stresses in the relavent parts of the scapulocoracoid are between -3 and -20 MPa..
206
10.4.4 Technical settings Forelimb position I.
Table 9 Applied forces, which refer to muscles or muscle groups.
Muscle
Muscle force (N)
Fg body weight
18956
M. serratus superficialis
15800
M. serratus profundus
3160
M. trapezius
10354
M. levator scapulae
M. costo-coracoideus
M. rhomboideus
M. pectoralis posterior
M. pectoralis anterior
M. triceps caput coracoideum
M. triceps caput scapulare
M. supracoracoideus
M. biceps brachii
M. coracobrachialis brevis ventralis
M. deltoideus scapularis
M. deltoideus clavicularis
M. subscapularis
M. scapulo-humeralis
M. triceps caput humerale laterale
M. triceps capita humerale mediale
M. triceps capita humerale posticum
M. spiralis
M. brachialis
M. latissimus dorsi
M. teres major
207
Table 10 Showing the applied forces referring to muscles or muscle groups. F x,y,z are reaction forces ,which can be
detected at supported nodes after calculation.
Muscle
Muscle force
(N)
NODE
FX
FY
FZ
Fg body weight
18956
228
-100.49
286.89
130.21
M. serratus superficialis
15800
344
-60.062
126.09
-53.008
M. serratus profundus
3160
462
32.473
85.347
-56.963
M. trapezius
6902
73725
-3675.2
-942.35
M. levator scapulae
6360
78115
3216.0
-166.78
M. costo-coracoideus
13516
83886
1713.1
-49.955
M. rhomboideus
1400
141302
43.812
1370.5
728.29
141424
-1169.7
17087.
410.56
Total value
-0.37996E-02
18956.
-0.92033E-05
M. pectoralis posterior
M. pectoralis anterior
M. triceps caput coracoideum
M. triceps caput scapulare
M. supracoracoideus
M. biceps brachii
M. coracobrachialis brevis ventralis
M. deltoideus scapularis
M. deltoideus clavicularis
M. subscapularis
M. scapulo-humeralis
M. triceps caput humerale laterale
M. triceps capita humerale mediale
M. triceps capita humerale posticum
M. spiralis
M. brachialis
M. latissimus dorsi
M. teres major
208
Table 11 Values of applied forces refer to muscles or muscle groups. F z is reaction force, which can be detected at
supported nodes located at the humerus after calculation.
Muscle
Muscle force (N)
NODE
Fg body weight
18956
141623
M. serratus superficialis
15800
M. serratus profundus
3160
M. trapezius
6902
M. levator scapulae
6360
M. costo-coracoideus
13516
M. rhomboideus
1400
VALUE
M. pectoralis posterior
M. pectoralis anterior
M. triceps caput coracoideum
9332
M. triceps caput scapulare
7466
M. supracoracoideus
M. biceps brachii
M. coracobrachialis brevis ventralis
M. deltoideus scapularis
M. deltoideus clavicularis
M. subscapularis
M. scapulo-humeralis
M. triceps caput humerale laterale
M. triceps capita humerale mediale
M. triceps capita humerale posticum
M. spiralis
M. brachialis
M. latissimus dorsi
M. teres major
209
FX
FY
FZ
-3173.1
0.0000
0.0000
-3173.1
Table 12 Values of applied forces refer to muscles or muscle groups.
Muscle
Muscle force (N)
Fg body weight
18956
M. serratus superficialis
15800
M. serratus profundus
3160
M. trapezius
3543,77
M. levator scapulae
7633,56
M. costo-coracoideus
13904
M. rhomboideus
1400
M. pectoralis anterior
M. pectoralis posterior
14964
M. triceps caput coracoideum
M. triceps caput scapulare
M. supracoracoideus
M. biceps brachii
M. coracobrachialis brevis ventralis
M. deltoideus scapularis
M. deltoideus clavicularis
M. subscapularis
M. scapulo-humeralis
M. triceps caput humerale laterale
M. triceps capita humerale mediale
M. triceps capita humerale posticum
M. spiralis
M. brachialis
M. latissimus dorsi
M. teres major
210
Table 13 Values of applied forces refer to muscles or muscle groups.
Muscle
Muscle force (N)
Fg body weight
18956
M. serratus superficialis
15800
M. serratus profundus
3160
M. trapezius
6902
M. levator scapulae
6360
M. costo-coracoideus
13644
M. rhomboideus
1400
M. pectoralis anterior
1800
M. pectoralis posterior
3640
M. triceps caput coracoideum
1500
M. triceps caput scapulare
2000
M. supracoracoideus
1000
M. biceps brachii
1500
M. coracobrachialis brevis ventralis
2000
M. deltoideus scapularis
100
M. deltoideus clavicularis
2000
M. subscapularis
5000
M. scapulo-humeralis
2000
M. triceps caput humerale laterale
500
M. triceps capita humerale mediale
500
M. triceps capita humerale posticum
500
M. spiralis
500
M. brachialis
500
M. latissimus dorsi
M. teres major
211
10.4.5 Technical settings forelimb position II.
Table 14 Values of applied forces refer to muscles or muscle groups. F x,y,z are reaction forces, which can be detected
at supported nodes after calculation.
Muskel
Fr
NODE
FX
FY
FZ
M. serratus superficialis
19058
12551
-606.35
15226.
-6577.1
M. serratus profundus
5707
122396
-3120.4
970.99
M. trapezius
8550
122708
3735.3
1447.6
M. levator scapulae
6464
122730
1026.0
3620.8
M. costo-coracoideus
7880
123301
-48.179
-249.01
186985
165.12
M. triceps caput coracoideum
0
186991
-11.561
190.52
M. triceps caput scapulare
0
187047
80.003
507.73
305.28
M. supracoracoideus
0
187145
439.19
192.74
M. biceps brachii
0
193046
-1054.8
2427.2
878.78
M. coracobrachialis brevis ventralis
0
M. deltoideus scapularis
0
M. deltoideus clavicularis
0
Total
values
0.28684E04
18956.
0.60038E04
M. subscapularis
0
M. scapulo-humeralis
0
M. rhomboideus
M. pectoralis 1
M. pectoralis 2
M. pectoralis 3
M. pectoralis 4
M. latissimus dorsi
M. triceps caput humerale laterale
M. triceps capita humerale mediale
M. triceps capita humerale posticum
M. spiralis
M. brachialis
212
Table 15 Values of applied forces refer to muscles or muscle groups. F x,y,z are reaction forces, which can be detected
at supported nodes after calculation.
Muskel
Fr
NODE
FX
M. serratus superficialis
19058
1287
M. serratus profundus
5707
1312
6420.5
M. trapezius
8550
5541
2976.5
M. levator scapulae
6464
12585
1124.1
M. costo-coracoideus
7880
186985
FY
FZ
60.484
15143.
6155.8
167.02
186991
86.684
200.41
93.814
529.01
304.21
444.85
189.63
M. triceps caput coracoideum
0
187047
M. triceps caput scapulare
0
187145
M. supracoracoideus
0
193046
991.00
2004.8
774.34
M. biceps brachii
0
M. coracobrachialis brevis ventralis
0
M. deltoideus scapularis
0
TOTAL
VALUES
8866.1
18489.
4948.1
M. deltoideus clavicularis
0
M. subscapularis
0
M. scapulo-humeralis
0
M. rhomboideus
M. pectoralis 1
M. pectoralis 2
M. pectoralis 3
M. pectoralis 4
M. latissimus dorsi
M. triceps caput humerale laterale
M. triceps capita humerale mediale
M. triceps capita humerale posticum
M. spiralis
M. brachialis
213
Table 16 Values of applied forces refer to muscles or muscle groups. F x,y,z are reaction forces, which can be detected
at supported nodes after calculation.
Muskel
Fr
x/y/z
NODE
FX
M. serratus superficialis
19058
1287
M. serratus profundus
5707
1312
8479.9
M. trapezius
8550
5541
4563.3
M. levator scapulae
6464
12551
-3343.2
M. costo-coracoideus
7880
186985
FY
FZ
-432.77
18855.
-4966.8
-110.84
186991
100.57
-197.29
184.68
42.130
-95.471
147.49
-91.053
M. triceps caput coracoideum
30000
187047
M. triceps caput scapulare
2000
187145
M. supracoracoideus
0
193046
-754.57
1545.0
488.34
TOTAL VALUES
6253.4
20282.
-5097.8
M. biceps brachii
M. coracobrachialis brevis ventralis
0
M. deltoideus scapularis
0
M. deltoideus clavicularis
0
M. subscapularis
0
M. scapulo-humeralis
0
M. rhomboideus
0
M. pectoralis 1
0
M. pectoralis 2
0
M. pectoralis 3
0
M. pectoralis 4
0
M. latissimus dorsi
0
M. triceps capita humerale mediale
0
M. triceps capita humerale posticum
0
M. spiralis
0
M. brachialis
0
214
Table 17 Values of applied forces refer to muscles or muscle groups. F x,y,z are reaction forces, which can be detected
at supported nodes after calculation.
Muskel
Fr
x/y/z
NODE
FX
FY
FZ
M. serratus superficialis
19058
2
2567.4
M. serratus profundus
4040
1287
M. trapezius
7650
1312
8666.5
M. levator scapulae
6464
5541
3962.4
M. costo-coracoideus
7800
12551
5156.2
M. triceps caput coracoideum
2000
186985
M. triceps caput scapulare
3000
186991
177.97
331.34
M. supracoracoideus
2000
187047
303.40
534.91
455.66
M. biceps brachii
3000
187145
391.00
256.81
M. coracobrachialis brevis ventralis
2000
193046
399.42
634.16
436.64
M. deltoideus scapularis
3500
M. deltoideus clavicularis
3500
M. subscapularis
3500
TOTAL
VALUE
5786.1
20275.
4267.1
M. scapulo-humeralis
4000
M. rhomboideus
0
M. pectoralis 1
0
M. pectoralis 2
0
M. pectoralis 3
0
M. pectoralis 4
0
M. latissimus dorsi
0
M. triceps capita humerale mediale
0
M. triceps capita humerale posticum
0
M. spiralis
0
M. brachialis
0
15.515
215
22393
3119.5
226.73
Table 18 Values of applied forces refer to muscles or muscle groups. F x,y,z are reaction forces, which can be detected
at supported nodes after calculation.
Muskel
Fr
x/y/z
NODE
FX
M. serratus superficialis
1958
2
2379.5
M. serratus profundus
2123
1287
M. trapezius
6000
1312
7887.0
M. levator scapulae
7000
5541
3833.4
M. costo-coracoideus
5000
12551
3299.3
M. pectoralis profundus
5000
186985
M. triceps caput coracoideum
2000
187047
M. triceps caput scapulare
0
187145
M. supracoracoideus
2500
193046
M. biceps brachii
2000
M. coracobrachialis brevis ventralis
2500
M. deltoideus scapularis
0
TOTAL
VALUE
M. deltoideus clavicularis
0
M. subscapularis
0
M. scapulo-humeralis
0
M. rhomboideus
0
M. pectoralis 1
0
M. pectoralis 2
0
M. pectoralis 3
0
M. pectoralis 4
0
M. latissimus dorsi
0
M. triceps capita humerale mediale
0
M. triceps capita humerale posticum
0
M. spiralis
0
FY
FZ
597.16
216
18150.
847.52
85.090
91.135
10.622
79.529
119.14
99.542
30.515
2542.1
194.76
6245.1
20565.
624.19
Table 19 Values of applied forces refer to muscles or muscle groups. F x,y,z are reaction forces, which can be detected
at supported nodes after calculation.
Muskel
Fr
x/y/z
Fg Summe
NODE
FX
444
3451.7
FY
FZ
M. serratus superficialis
19058
1287
M. serratus profundus
446
5565
4342.4
M. trapezius
5500
10124
8752.2
M. levator scapulae
3500
12551
5160.0
M. levator scapulae superior
1500
186985
M. rhomboideus
1000
186991
195.27
290.65
M. costo-coracoideus
5000
187047
152.24
625.67
1008.6
M. pectoralis profundus
6000
187145
237.44
691.05
M. triceps caput coracoideum
1900
M. triceps caput scapulare
1000
M. supracoracoideus
3500
19132.
6266.8
M. biceps brachii
2000
M. coracobrachialis brevis ventralis
4000
M. deltoideus scapularis
0
M. deltoideus clavicularis
2660
M. subscapularis
2000
M. scapulo-humeralis
0
M. pectoralis 1
0
M. pectoralis 2
0
M. pectoralis 3
0
M. pectoralis 4
3300
M. latissimus dorsi
0
200.05
TOTAL
VALUE
M. triceps caput humerale laterale
M. triceps capita humerale mediale
M. triceps capita humerale posticum
M. spiralis
M. brachialis
217
17409.
6400.3
459.46
5077.0
Table 20 Values of applied forces refer to muscles or muscle groups. F x,y,z are reaction forces, which can be detected
at supported nodes after calculation.
Muskel
Fr
x/y/z
NODE
FX
FY
FZ
M. serratus superficialis
19058
444
4879.7
M. serratus profundus
542
1287
M. trapezius
5500
5565
2039.2
M. levator scapulae
3500
5764
2152.7
M. levator scapulae superior
1500
5830
581.46
M. rhomboideus
1000
10124
4132.7
M. costo-coracoideus
4000
186985
M. costo-coracoideus pars prof.
3500
186991
35.459
761.37
M. pectoralis profundus
4500
187047
16.848
1858.9
937.35
M. pectoralis 1
500
187145
1931.9
474.33
M. pectoralis 2
500
193046
4161.8
13910.
5248.3
M. pectoralis 3
500
M. pectoralis 4
3500
M. latissimus dorsi
0
TOTAL
VALUE
9676.3
19323.
6501.5
M. triceps caput coracoideum
1900
M. triceps caput scapulare
1000
M. supracoracoideus
4500
M. biceps brachii
2000
M. coracobrachialis brevis ventralis
4000
M. deltoideus scapularis
0
M. deltoideus clavicularis inferior
0
M. deltoideus clavicularis
2000
M. subscapularis
2000
M. scapulo-humeralis
0
158.48
M. triceps caput humerale laterale
M. triceps capita humerale mediale
M. triceps capita humerale posticum
M. spiralis
218
860.58
M. brachialis
Table 21 Values of applied forces refer to muscles or muscle groups. F x,y,z are reaction forces, which can be detected
at supported nodes after calculation.
Muskel
Fr
x/y/z
NODE
FX
FY
FZ
M. serratus superficialis
19058
444
4879.7
M. serratus profundus
542
1287
M. trapezius
5500
5565
2039.2
M. levator scapulae
3500
5764
2152.7
M. levator scapulae superior
1500
5830
581.46
M. rhomboideus
1000
10124
4132.7
M. costo-coracoideus
4000
186985
M. costo-coracoideus pars prof.
3500
186991
35.459
761.37
M. pectoralis profundus
4500
187047
16.848
1858.9
937.35
M. pectoralis 1
500
187145
1931.9
474.33
M. pectoralis 2
500
193046
4161.8
13910.
5248.3
M. pectoralis 3
500
M. pectoralis 4
3500
M. latissimus dorsi
0
TOTAL
VALUE
9676.3
19323.
6501.5
M. triceps caput coracoideum
1900
M. triceps caput scapulare
1000
M. supracoracoideus
4500
M. biceps brachii
2000
M. coracobrachialis brevis ventralis
4000
M. deltoideus scapularis
0
M. deltoideus clavicularis inferior
0
M. deltoideus clavicularis
2000
M. subscapularis
2000
M. scapulo-humeralis
0
158.48
M. triceps caput humerale laterale
M. triceps capita humerale mediale
M. triceps capita humerale posticum
219
860.58
M. spiralis
M. brachialis
Table 22 Values of applied forces refer to muscles or muscle groups. F x,y,z are reaction forces, which can be detected
at supported nodes after calculation.
Muskel
Fr
x/y/z
NODE
FX
FY
FZ
M. serratus superficialis
19057
444
181.93
M. serratus profundus
354
1287
M. trapezius
5100
5565
4569.6
M. levator scapulae
3100
10124
8638.2
M. levator scapulae superior
1500
186985
M. rhomboideus
40
186991
46.329
686.25
M. costo-coracoideus
4000
187047
109.57
1770.8
1087.8
M. costo-coracoideus pars prof.
3000
187145
1869.6
718.24
M. pectoralis profundus
4500
193046
4118.2
14028.
4323.8
M. pectoralis 1
0
M. pectoralis 2
0
M. pectoralis 3
0
TOTAL
VALUE
8970.9
19187.
6116.4
M. pectoralis 4
3500
M. latissimus dorsi
0
M. triceps caput coracoideum
1900
M. triceps caput scapulare
1000
M. supracoracoideus
4500
M. biceps brachii
2000
M. coracobrachialis brevis ventralis
2000
M. deltoideus scapularis
0
M. deltoideus clavicularis inferior
0
M. deltoideus clavicularis
6000
M. subscapularis
2000
M. scapulo-humeralis
0
13.439
M. triceps caput humerale laterale
M. triceps capita humerale mediale
220
832.29
M. triceps capita humerale posticum
M. spiralis
M. brachialis
Table 23 Values of applied forces refer to muscles or muscle groups. F x,y,z are reaction forces, which can be detected
at supported nodes after calculation.
Muskel
Fr
x/y/z
NODE
FX
M. serratus superficialis
19057
444
2506.1
M. serratus profundus
354
Scapula
M. trapezius
5100
5565
5822.5
M. levator scapulae
3100
10124
4531.4
M. levator scapulae superior
1500
186985
M. rhomboideus
40
186991
-65.951
655.40
M. costo-coracoideus
4000
187047
104.03
1761.0
1096.9
M. costo-coracoideus pars prof.
3000
187145
1872.5
737.45
M. pectoralis profundus
4500
193046
-4093.5
14098.
4279.1
M. pectoralis 1
0
M. pectoralis 2
0
M. pectoralis 3
0
TOTAL VALUE
8804.6
19204.
6050.4
M. pectoralis 4
3500
M. latissimus dorsi
0
M. triceps caput coracoideum
1900
M. triceps caput scapulare
1000
M. supracoracoideus
4500
M. biceps brachii
2000
M. coracobrachialis brevis ventralis
2000
M. deltoideus scapularis
0
M. deltoideus clavicularis inferior
0
M. deltoideus clavicularis
6000
M. subscapularis
2000
M. scapulo-humeralis
0
M. triceps caput humerale laterale
221
FY
FZ
-63.120
816.65
M. triceps capita humerale mediale
M. triceps capita humerale posticum
M. spiralis
M. brachialis
Table 24 Final calculation. Values of applied forces refer to muscles or muscle groups. F x,y,z are reaction forces, which
can be detected at supported nodes after calculation.
Muscle
Fr
x/y/z
NODE
M. serratus superficialis
17603
Sternocoracoid
joint
M. serratus profundus 1
FX
FY
FZ
2
683.14
301
12
-19.819
M. serratus profundus 2
2140
18
1519.2
M. trapezius
4300
25
1016.7
M. levator scapulae
3000
32
445.13
M. levator scapulae superior
1175
425
464.48
M. rhomboideus
1000
447
1233.9
M. costo-coracoideus
4800
553
-507.90
M. costo-coracoideus pars prof.
3000
557
-712.87
M. pectoralis profundus
4500
565
735.69
M. pectoralis
3500
570
314.88
M. triceps caput coracoideum
1900
576
-179.77
M. triceps caput scapulare
1000
Acromion
10124
1171.8
M. supracoracoideus
4500
Antebrachium
186985
M. biceps brachii
2000
186991
-2325.6
1090.8
M. coracobrachialis brevis ventralis
4000
187047
-2709.8
3730.7
2518.0
M. deltoideus scapularis
5000
187145
8529.8
3329.2
M. deltoideus clavicularis
6000
M. subscapularis
2000
M. scapulo-humeralis
500
18387.
5847.1
TOTAL
VALUE
222
5035.7
1129.2
Curriculum Vitae
Date of birth
14.02.1974, in Dortmund
Marital status
married, one child
Education
Since 02/05
PhD student at the Ruhr-University; Faculty of Biology and
Biotechnology, Department of Zoology and Neurobiology,
Bochum; Thesis: Biomechanics of the shoulder girdle in
sauropod dinosaurs
10/00 bis 12/03
Biology at the University of Mainz, Diploma
10/95 bis 07/00
Biology at the Ruhr-University Bochum
10/94 bis 07/95
Biology at the University Göttingen
1993
Abitur
Professional Experience
Since 15/02/10
Project coordinator at the Ruhr-University Research School,
Bochum
10/05-10/08
Scientific assistant of Prof. Dr. Dr. h.c. H. Preuschoft, formerly
Institute of Anatomy, Medical Faculty, Ruhr-University Bochum
01/04-06/04
Institute of Anthropology,
University Mainz
Publications
Hohn-Schulte B, Preuschoft H & Witzel U (submitted) Biomechanics of the tetrapod shoulder
girdle with special emphasis on the early tetrapod Tiktaalik roseae. PLoS ONE.
Hohn-Schulte B (in press) Walking with the Shoulder of Giants: Biomechanical Conditions in the
Tetrapod Shoulder Girdle as a Basis for Sauropod Shoulder Reconstruction In: Biology of
the Sauropod Dinosaurs: Understanding the life of giants. Klein N, Remes K, Gee C (eds.)
James Farlow, Indiana University Press, 513-551.
Hohn B, Witzel U & Preuschoft H (2008) 3-D Finite-Element Structure Synthesis of the shoulder
girdle in Diplodocus longus. Journal of Vertebrate Palaeontology 28 (3) 88A.
223
Hohn B, Witzel U & Preuschoft H (2006) Functional morphology of the shoulder girdle and the
forelimbs in sauropod dinosaurs under consideration of 3-D finite-element structure
synthesis (FESS). Journal of Vertebrate Palaeontology 26(3): 77A.
Preuschoft H, Hohn-Schulte B, Stoinski S & Witzel U (in press) Why so huge? Biomechanical
reasons for the acquisition of large size in sauropod and theropod dinosaurs. In: Biology of
the Sauropod Dinosaurs: Understanding the life of giants. Klein N, Remes K, Gee C (eds.)
James Farlow, Indiana University Press.
Preuschoft H, Hohn-Schulte B, Scherf H, Schmidt M, Krause C & Witzel U (2010) Functional
analysis of the primate shoulder. International Journal of Primatology 31: 301-320.
Preuschoft H, Witzel U, Hohn B, Distler C & Schulte D (2007) Biomechanics of locomotion and
body structure in varanids with special emphasis on the forelimbs. Mertensiella 16, 59-78.
Congresses and Workshops
2008
1st International Workshop” The Evolution of Gigantism in Sauropod
dinosaurs”, Bonn
2008
68th Annual meeting of the “Society of Vertebrate Palaeontologists” (SVP),
Cleveland (USA)
2007
„International Congress of Vertebrate Morphologists“(ICVM),
Paris (F)
2006
66th Annual meeting of the “Society of Vertebrate Palaeontologists” (SVP),
Ottawa (CA)
2006
8th Congress of the “Society of Vertebrate Palaeontology and
Comparative Anatomy” (SVPCA), Paris (F)
Grants
10/08-03/09
PhD Grant Wilhelm und Günter Esser-Stiftung, Gesellschaft der Freunde
der Ruhr-Universität Bochum (GdF)
10/2008
Womens support grant of the Faculty of Biology and Biotechnology, RuhrUniversity Bochum
Memberships in professional associations
DFG-Forschergruppe 533 „Biologie der Sauropoden- Evolution des
Gigantismus“
Society of Vertebrate Palaeontology (SVP)
224
Acknowledgements
I first express my gratitude to Dr. C. Distler-Hoffmann for her guidance, support and
encouragement throughout the thesis. I wish to thank Prof. Dr. H. Preuschoft for assignment of
this work, all the beneficial discussions and advices and for the opportunity to join the Research
group “Biology of the Sauropod Dinosaurs: The Evolution of Gigantism. For his extensive support
and advice my thanks go to Prof. Dr. U. Witzel. I have greatly benefited from discussions, guidance
and overall support from R. Gössling and Dr. N. Sverdlova. Further thanks go to all members of
the LMK. I am also indebted to Prof. M. Sander, Dr. C. Gee, Dr. A. Christian, N. Pajor, T. Bräuer, K.
Moser, Dr. K. Remes, Dr. O. Rauhut, Dr. H. Mallison, Dr. N. Klein, J.-T. Möller, Dr. T. Suthau, S.
Stoinski, Dr. T. Tütken, and the remaining members of the Research group for discussions on
sauropods and locomotion. Special thanks goes to K. Heitplatz for organization and help,
whenever it was needed. This thesis could never have been completed without Dr. R. Fechner,
and I am grateful for her support and encouragement. For access to the wonderful fossil of
Tiktaalik roseae many thanks go to Dr. Shubin and K. Monoyios, University of Chicago. For access
to the specimens and assistance I thank the staff of the following institutions: Institute
d´Anatomie Normale, Strassbourg, Senckenberg museum of natural history, Frankfurt; Natural
museum of history, Berlin; Saurier museum Aathal; Bayrische Staatsammlung, München. Many
thanks go to my colleagues at the Ruhr-University Research School. I am overall grateful to my
family and friends who never stopped supporting me through the long years to finish this project.
This research project was funded by the DFG and a PhD grant of the Wilhelm & Günther Esser
Stiftung.
225
Erklärung
Hiermit erkläre ich, dass ich die Arbeit selbstständig verfasst und bei keiner anderen Fakultät
eingereicht und dass ich keine anderen als die angegebenen Hilfsmittel verwendet habe. Es handelt
sich bei der heute von mir eingereichen Dissertation um fünf in Wort und Bild völlig
übereinstimmende Exemplare.
Weiterhin erkläre ich, dass digitale Abbildungen nur die orginalen Daten enthalten und in
keinem Fall inhaltsverändernde Bidbearbeitung vorgenommen wurde.
Bochum, im April 2010
_______________________________
(Unterschrift)
226